Evaluating Australian Electoral Reforms: 1983–2007

Norm Kelly

December 2008

A thesis submitted for the degree of Doctor of Philosophy of the Australian National University

i

Declaration

This thesis contains no material which has been accepted for the award of any other degree or diploma in any university. To the best of my knowledge and belief, it contains no material previously published or written by another person, except where due reference is made in the text.

Norm Kelly

18 December, 2008

ii

iii

Acknowledgements

Many people have contributed to this thesis. Foremost in my gratitude is Professor Marian Sawer, my principal supervisor, who saw the value in my work. Thank you for asking the tough questions when they needed asking, and being supportive over these past few years. To my other panel members, Professor Ian McAllister and Professor John Warhurst, my discussions with you were infrequent, but quite valuable in helping me to decide on the direction the thesis would take, and what was best left aside.

Some parts of this thesis (particularly Chapter 5) were presented in draft form at conferences in Australia and New Zealand. To those people with whom I’ve discussed my work, I have appreciated your input and encouragement. You are probably unaware of the sometimes small, but useful remarks you have made, and hopefully I have provided something in return.

The Australian National University has a talented political science community, and I have appreciated being able to learn and gain inspiration from colleagues who have become friends. Thank you Ben Reilly, Kim Huynh, Peter Brent, Kate Lee-Koo, David Adams and Norman Abjorensen.

A large part of the data for this study came from my interview participants –– commissioners, ministers, and others –– all people with important responsibilities and time pressures. I was pleasantly surprised and I am grateful to these participants for their willingness to be interviewed. Their candour was important to this research, and without their co-operation, this study would have taken a very different direction.

Special thanks go my friend and colleague, Merrindahl Andrew, who provided support and friendship, as well as taking on the massive task of reading my draft chapters –– your suggestions were invaluable.

Finally, I give my biggest thanks and gratitude to Gail, my partner in life, for your enduring love, support and encouragement over the past years. And to top it off, a damn good social scientist who read all my chapters and assisted me greatly in working through the conceptual stuff! iv

v

Abstract

Australia has an international reputation for electoral innovation and the professionalism and independence of its electoral administration. Significant reforms over the past 25 years, however, demand a re-appraisal of Australia’s standing in terms of electoral democracy.

Applying and adapting international measures of ‘fairness’, this doctoral thesis assesses electoral reforms since 1983, considers their impacts on voters, candidates, and parties, and also examines the drivers of reform. While governments tend to introduce electoral reforms that reap partisan benefits, there may be unintended consequences of such changes. Moreover, and consistent with the party cartelisation thesis, reforms may benefit major parties as a group at the expense of other electoral competitors. At the same time, impacts on voters in terms of fairness may be mixed, depending on the partisan effect.

The thesis investigates specific reforms in Australia’s nine electoral jurisdictions using both a comparative and case study approach. The thesis applies a model for evaluating the fairness of electoral governance that draws on the work of Elkit and Reynolds (2005) but extends it to make it more capable of distinguishing between good and bad practice in developed democracies. In evaluating the reforms, the study draws on comparative measures such as Elklit and Reynolds’ (2005) framework for measuring fairness and partisan bias, and Gallagher’s Least Squares Index. The Elklit and Reynolds model is modified to make it more applicable to developed democracies such as Australia. In addition to specific reforms, the independence of electoral management bodies is interrogated and measured, building on existing assessment criteria.

The thesis draws on a variety of sources, including personal interviews with current electoral commissioners and electoral administrators, ministers and shadow ministers, and parliamentary committee members. In addition to these interviews with key players, the thesis uses legislative and parliamentary data to inform its analysis and assessment.

vi

Electoral reforms in the past 25 years have largely benefited the major parties (Labor and the Liberal/National Coalition), to the detriment of other parties and candidates. While benefits to minor parties such as the Greens and Democrats can be identified, the major parties have gained more significant electoral advantages. The thesis finds that the fairness of the Australian electoral system, with some notable exceptions, has diminished over the study period. Voters are often significantly disadvantaged by reforms. In addition, when compared to international norms of electoral management and fair elections, administrative ‘best practice’ is often sacrificed for partisan advantage.

vii

Contents

Declaration i

Acknowledgments iii

Abstract v

Tables xii

Figures xiv

Abbreviations xv

Chapter 1: Introduction 1 Research Questions and Framework 7 Thesis Structure 9

Chapter 2: Literature Review 11 Theory of Electoral Systems 13 Voting and Electoral System Design Theory Party Cartelisation Theory Fairness 19 Defining Fairness Measuring Electoral Outcomes Studies of Electoral Institutions 24 Comparative Studies of Voting Systems and Electoral Regulation International Norms for Electoral Management Australian Electoral Institutions 27 Gaps in the Literature 33

Chapter 3: Methodology 35 Research Design 35 Assessment Frameworks 36 Discussion-based model –– Democratic Audit Framework Comparative models Ratings-based model –– the Elklit and Reynolds Framework

viii

Scope of the Study 42 Data Collection 45 Analysis of Data 47 Personal Interviews 49 The Role of the Researcher 52 Summary 53

Chapter 4: Institutional Structure 55 Australian Electoral Management Bodies 56 NSW Election Funding Authority 57 Parliamentary Oversight of Electoral Matters 60 Commonwealth 2004 Federal Election Inquiry –– the Richmond Result New South Wales Joint Standing Committee on Electoral Matters Victorian Electoral Matters Committee Queensland Legal, Constitutional and Administrative Review Committee Senate Finance and Public Administration Standing Committees NSW Select Committee on Electoral and Political Party Funding Engagement with Citizens and Civil Society Organisations 69 Summary 72

Chapter 5: The Independence of Electoral Management Bodies 75 Models of Independence 76 Commissioner Experience 81 The Size of Commissions 83 Commissioner Selection and Appointment Processes 84 The Requirement to Consult Political Affiliations 88 Length of Tenure 89 Security of Tenure 92 Reporting Mechanisms 93 Budget Processes 94 Budgetary Independence

ix

Determining Best Practice for Independence 98 Summary 101

Chapter 6: Reform Processes I –– The Commonwealth 103 Win Government, Implement Electoral Reform 103 Labor’s 1983–84 Reforms 104 The Size of Parliament –– The House of Representatives The Size of Parliament –– The Senate Public Funding Senate Ticket Voting National Voters Voting Liberal Reform Attempts –– 1996–2004 120 The ATSIEEIS Case The Coalition’s 2006 Reforms 122 Earlier Closure of the Electoral Roll Proof of Identity Provisions Donation Disclosure Thresholds Tax-deductibility of Political Donations Postal Voting Applications –– Self-Interest vs Fair Elections 131 Postal Voting Processes –– The 2007 Federal Election Labor Liberal A Change of Government –– More Reform 139 Summary 140

Chapter 7: Reform Processes II –– States and Territories 143 New South Wales 144 Too Many Parties –– the 1999 Legislative Council Election Victoria 147 The Move to Proportional Representation Queensland 150 Western Australia 153 The Burke Government Reforms The Gallop Government Reforms South Australia 156

x

Challenges in Implementing Reform Truth in Political Advertising Tasmania 159 Reform for Political Advantage Reform for Efficient Electoral Administration Australian Capital Territory 163 Electronic Voting Northern Territory 169 Determining the Northern Territory’s Party Registration Regime Summary 174

Chapter 8: Reform Processes III –– Cross-Jurisdictional Issues 177 Voter Turnout 177 The Northern Territory –– Indigenous Enrolment and Turnout Administrative Response to Indigenous Enrolment and Turnout Northern Territory Assembly Elections –– Timeframe The Prisoner Franchise 184 Changes to Commonwealth Laws The 2007 High Court Challenge Prisoner Franchise at the State and Territory Levels Partisan Impacts of Prisoner Voting Party Registration 190 Registration Criteria Parliamentary Representative Alternative for Party Registration South Australian Registration Regime Party Registration –– Achieving a Balance Deregistration of the Liberals for Forests Informal Voting 200 Informal Voting at Federal Elections Informal Voting at State and Territory Levels Analysis of Cross-Jurisdictional Issues 204 Summary 208

xi

Chapter 9: Concluding Discussion 209 Summary of the Research Process 210 Summary of Findings 211 Significance of the Research 216 Future Research Directions 217 Conclusion 218

References 221

Websites 237

Appendices 239

Appendix A List of Interview Participants 239 Appendix B JSCER and JSCEM (Commonwealth) Reports 241 Appendix C Gallagher’s Least Squares Index –– Calculations 243 Appendix D Gallagher’s Least Squares Index –– Federal Elections 245 Appendix E Gallagher’s Least Squares Index –– NSW Legislative Council 253 Appendix F Gallagher’s Least Squares Index –– Victorian Legislative Council 255 Appendix G Gallagher’s Least Squares Index –– Queensland Legislative Assembly 257 Appendix H Gallagher’s Least Squares Index –– Western Australian Parliament 259 Appendix I Gallagher’s Least Squares Index –– Tasmanian Legislative Assembly 263 Appendix J Northern Territory Turnout and Informality –– 2005 General Election 265

xii

Tables

Table 1.1 Principal Electoral Legislation and Major Reforms 1983–2007

Table 2.1 Support at Federal Elections (%) –– House of Representatives v. Senate

Table 3.1 Democratic Audit Assessment Framework Table 3.2 Democratic Audit Assessment Questions Table 3.3 Three Levels of Electoral Governance Table 3.4 Establishing the Rules of the Game Table 3.5 Election Assessment Framework Table 3.6 Indicators used for Comparative Analysis

Table 4.1 Australia’s Electoral Management Bodies Table 4.2 Parliamentary Oversight Committees

Table 5.1 IDEA’s Model of Independent Electoral Management Bodies Table 5.2 Electoral Experience of Current Australian Electoral Commissioners Table 5.3 Appointment Processes for Australian Electoral Commissioners Table 5.4 Prohibitions on Appointment of Party Members or Parliamentarians Table 5.5 Length of tenure for Australian Electoral Commissioners Table 5.6 Dismissal Processes for Australian Electoral Commissioners Table 5.7 Reporting Mechanisms for Australian Electoral Administrations Table 5.8 Best Practice Analysis –– Independence of Electoral Administration

Table 6.1 Proportionality in the House of Representatives Table 6.2 Proportionality in the Senate Table 6.3 Public Funding payments ($m) Table 6.4 Public funding payments –– selected parties and candidates ($m) Table 6.5 Number of Parties and Independents Receiving Public Funding 1993–2007 Table 6.6 Senate Informal Voting Rate (%) Table 6.7 Senate Group Voting Ticket Usage and Number of Candidates, 1984-2007 Table 6.8 2004 Federal Election Results for the Coalition –– McMillan, Victoria Table 6.9 PVAs received for the 2007 election –– period between witness signature and AEC receipt of PVA

xiii

Table 7.1 NSW Legislative Council election results 1984–2007 Table 7.2 Victorian Election Results 1999–2006 Table 7.3 Queensland Election Results 1983–2006 Table 7.4 Western Australia –– Gallagher Least Squares Index Table 7.5 Reduction in Size of Tasmanian Parliament –– Reform Process Table 7.6 Tasmania –– Gallagher’s Least Squares Index, Legislative Assembly Table 7.7 Australian Capital Territory –– Significant Electoral Events Table 7.8 Participation in ACT elections Table 7.9 Use of electronic voting and informality rates in the ACT Table 7.10 Candidates and parties contesting Northern Territory elections

Table 8.1 Turnout Rates, Australian Elections 1983–2007 Table 8.2 Northern Territory Voter Turnout (%) 1983–2007 Table 8.3 Prisoner Voting at Federal Elections, Two Candidate Preferred Table 8.4 Party Membership Requirements for Registration Table 8.5 South Australia –– Registered Parties Table 8.6 Senate and House of Representatives Informal Voting Rates (%) Table 8.7 Informal Voting Rates –– 2004 House of Representatives Election Table 8.8 2004 House of Representatives High Informal Rates –– Ethnicity Table 8.9 State and Territory Informal Rates –– Lower/Single Houses

xiv

Figures

Figure 2.1 Hierarchy of Incumbency

Figure 3.1 Research and Reporting Model Figure 3.2 Elklit and Reynolds Framework –– Scoring Methodology (2005) Figure 3.3 Kelly Framework –– Modified Elklit and Reynolds

Figure 6.1 Liberal Party How-to-Vote Card, 1998 Figure 6.2 Australian Greens How-to-Vote Card, 1996 Figure 6.3 Greens (WA) How-to-Vote Card, 1998 Figure 6.4 Labor Postal Voting Information Leaflet (excerpt) Figure 6.5 Liberal Postal Voting Information Leaflet (excerpt)

Figure 7.1 Tasmanian Reform Process (1999–2004) Figure 7.2 Northern Territory Ballot Paper (from Katherine by-election, 2003)

Figure 8.1 Turnout at the 2005 Northern Territory Assembly general election

xv

Abbreviations

2PP Two party preferred AATA Administrative Appeals Tribunal of Australia ABS Australian Bureau of Statistics ACE Administration and Cost of Elections Project ACT Australian Capital Territory ACTEC Australian Capital Territory Electoral Commission ADI Australian Defence Industries AEC Australian Electoral Commission AGPD Australian Government and Politics Database ALP Australian Labor Party ATL Above the line ATSIEEIS Aboriginal and Torres Strait Islander Election Education and Information Service BTL Below the line CLP Country Liberal Party DLP Democratic Labor Party DoFA Department of Finance and Administration (Commonwealth) ECQ Electoral Commission of Queensland EFA Election Funding Authority (NSW) HCA High Court of Australia HoR House of Representatives ICCPR International Covenant on Civil and Political Rights IDEA International Institute for Democracy and Electoral Assistance IFES International Foundation for Electoral Systems JSCEM Joint Standing Committee on Electoral Matters (Commonwealth) JSCER Joint Select Committee on Electoral Reform (Commonwealth) LSI Least Squares Index MEC Minter Ellison Consulting MHA Member of the House of Assembly MLA Member of the Legislative Assembly

xvi

MLC Member of the Legislative Council MP Member of Parliament NSW New South Wales NSWEC New South Wales Electoral Commission NSWLC New South Wales Legislative Council NK Norm Kelly NT Northern Territory NTEC Northern Territory Electoral Commission NTLA Northern Territory Legislative Assembly PAEC Public Accounts and Estimates Committee (Victoria) PRSA Proportional Representation Society of Australia PVA Postal voting application SA South Australia SAS Save the ADI Site Party SCEPPF Select Committee on Electoral and Political Party Funding (NSW) SCETYP Standing Committee on Education, Training and Young People (ACT) SCLA Standing Committee on Legal Affairs (ACT) SCRRG Select Committee on the Report of the Review of Governance (ACT) SEO State Electoral Office (South Australia) SFPALC Senate Finance and Public Administration Legislation Committee SSCFPA Senate Standing Committee on Finance and Public Administration TEC Tasmanian Electoral Commission TEO Tasmanian Electoral Office UNDP United Nations Development Programme VEC Victorian Electoral Commission WA Western Australia WAEC Western Australian Electoral Commission

Spacer page

Spacer page

1

Chapter 1

Introduction

Well, it’s interesting you describe it as a reform. Some people might regard it as –– you think of a reform as actually improving things. 1 – an Australian electoral commissioner

A democracy’s electoral system is fundamental to its legitimacy. 2 From the electoral system flows the form and style of representation, the relative strength of political parties, the formation of government, and the development of policy positions. In a representative democracy, the structure of a state’s electoral system plays a critical role in determining the nature and form of political discourse and parliamentary representation. The electoral system establishes who can vote, how many representatives are to be chosen from which areas, who is in charge of the conduct of elections, and how votes are counted. Because adjustment or manipulation of these elements can have severe positive or negative consequences for the viability of political parties, attempts to make changes are often fiercely debated. Electoral system reforms can also impact strongly on the ability of citizens to participate in a state’s general political discourse and democratic processes –– as voters, candidates or as members of political parties.

There are two main schools of thought on the relationship between electoral and party systems. The first is represented by Maurice Duverger who argues that parties and

1 Unless otherwise cited, all italicised quotes in this thesis are drawn from interview data. 2 In this thesis, the term ‘electoral system’ is used in a holistic sense to refer to all aspects of an electoral regime, including the voting system, electoral regulation, administration and management. The term ‘voting system’ is used to refer to the method of voting, such as proportional representation or majoritarian single-member electorates.

2 party systems are formed by the electoral environment in which they operate (Duverger, 1951). That is, the party system is a product of the electoral system. The second is represented by Stein Rokkan, Josep Colomer and others who argue that while this may be the case in the early development of a democracy, over time, it is the governing political parties that shape the design of the electoral system. In doing so, it is argued, governing political parties reinforce the electoral system that has created the power relationships between the parties (Rokkan, 1970; Benoit, 2004; Colomer, 2005). Using evidence from the Australian system, this thesis will argue in support of the second view.

Australia has historically been at the forefront of electoral design and innovation over the past 150 years, and with nine legislative jurisdictions, there is ample opportunity for experimentation and diversity. Australia is one of the oldest continuing democracies in the world, and the first to introduce the modern form of the secret ballot, the first country to vote itself into existence through popular referenda, and the first country in which women could both vote and stand for Parliament. 3 In the 19 th century, Australian colonies introduced democratic innovations such as the use of government-printed ballot papers and written nominations for elections. By the late 19 th century, Tasmania had pioneered the single-transferable-vote form of proportional representation for its lower house (Kelly and Sawer, 2008).

Since the 1980s, significant electoral reforms have continued at the national and sub- national levels. However, while 19 th century innovations such as the secret ballot, independent electoral officials and standardised ballot papers were seen as measures to improve the fairness of elections, more recent reforms have often been viewed as providing a partisan advantage for the political parties in power. For example, Elkins (1992: 68) argues that recent electoral experimentation, while promoted as improving democratic outcomes, is primarily designed to provide partisan advantage.

Academics and practitioners alike have identified the need for a systematic approach to assessing equity, and free and fair access for participants, in the conduct of elections (see Elklit and Reynolds, 2005: 147). As Massicotte, Blais and Yoshinaka (2004: x)

3 Australian women (except Indigenous women in some States) were entitled to vote and stand for parliament from 1902. While New Zealand was the first country to give women the vote in 1893, women in that country were not entitled to stand for parliament until 1919.

3 point out, there is a particular dearth of information in the area of international comparative analysis of electoral governance issues, with the literature even more limited at the sub-national level. Australia, with the rich diversity of electoral systems in its nine legislative jurisdictions (comprising the national Commonwealth level, six States and two Territories), provides excellent material for comparative analysis. Most of these jurisdictions have long-established representative democracies of more than 100 years. It is only the two Territories, the Northern Territory and the Australian Capital Territory (ACT), that have achieved self-government comparatively recently. Their first legislative elections under self-government were in 1980 and 1989, respectively. While there are many similarities between these nine jurisdictions, there is also significant diversity in their electoral systems, histories of reform, and administrative procedures. This variation warrants detailed analysis. Australia is often held up as a shining example of electoral best practice, so it is worthwhile examining whether that remains the case. This can be done by comparing Australian systems to each other, and assessing them against internationally accepted norms for electoral management and fair elections.

Some scholars argue that the reason for the lack of academic interest in electoral governance in established democracies is that electoral events are seen as routine matters, with outcomes generally within accepted margins of error (for example, Mozaffar and Schedler, 2002: 5), and that governance issues are only raised when the conduct or outcome of an election falls outside such margins (Elklit and Reynolds, 2005: 148). When established democracies make reforms however, there is a danger in being complacent. This is so whether they appear to be minor or major in nature, as these actions invariably have disparate impacts on individuals or groups. Reforms consistently result in either advantage or detriment for competing political parties, and therefore arguments about the merits of reform are often based on partisanship, rather than broader issues of equity and fairness. A systematic study of such reforms and their impacts will identify who has benefited or been disadvantaged; alongside this, an overall assessment of the equity and access of a system, for voters, candidates and parties can be made. Electoral systems inherently contain numerous factors that can impact on equity and access, such as voter qualification and registration, party registration, and ballot design. A study incorporating factors such as these can identify the strengths and weaknesses of a system and from this, areas in need of further work or reform.

4

At the Australian national level, there have been two significant instances of electoral reform in the past 25 years. The first occasion followed the 1983 general election, when the incoming Hawke Labor Government initiated sweeping changes. These included the introduction of ticket voting for the Senate; party identification on ballot papers, ballot order decided by lot (replacing alphabetical listing); the adoption of a modified Gregory system for the Senate count; public funding; party registration; and the establishment of a statutory electoral management body –– the Australian Electoral Commission –– which replaced the Australian Electoral Office.

The second major set of reforms occurred in 2006 under the Howard Liberal/National Coalition Government, which was finally able to implement significant changes that it had pursued from the time it first won office in 1996. Having won a Senate majority at the 2004 election (taking effect from July 2005), the Howard Government enacted reforms in 2006 that impacted on many facets of electoral administration. These reforms included the disenfranchisement of all prisoners;4 the earlier closure of the electoral roll; more stringent proof of identity requirements; increases in donation disclosure thresholds; and an increase in the level of tax-deductible donations.

Electoral systems at the sub-national level have similarly been subject to significant reform and amendment during this 25-year period. Queensland and Western Australia have moved to reduce malapportionment, Western Australia and Victoria have replaced Legislative Council single-member elections with regional proportional representation, Queensland and New South Wales have moved to optional preferential voting, and all jurisdictions have introduced and/or tightened party registration criteria. Tasmania has significantly reduced district magnitude in its House of Assembly electorates –– changing from seven-member to five-member seats. As a result, the Assembly has been reduced in size from 35 to 25 members. Even in Australia’s newest legislative jurisdiction, the ACT, the modified d’Hondt voting system of a single Territory-wide electorate that was originally used, has been replaced with a system based on three multi-member electorates using the Hare-Clark form of proportional representation. The ACT’s parliamentary terms have also been extended from three to four years.

4 The disenfranchisement of all prisoners was subsequently found to be unconstitutional by the High Court.

5

This thesis investigates several of Australia’s major electoral reforms over the past 25 years. In Table 1.1, the principal legislation governing the electoral administration and conduct of elections for each jurisdiction is shown, along with a summary of some of the major electoral reforms that have occurred during this period.

Table 1.1 Principal Electoral Legislation and Major Reforms 1983–2007 Commonwealth –– Commonwealth Electoral Act 1918 Commonwealth Electoral • Established the Australian Electoral Commission Legislation Amendment Act • Increased the size of Parliament 1983 • Independent redistribution process and • Registration of political parties Commonwealth Electoral • Public funding Legislation Amendment Act • Introduced above-the-line Senate voting and group 1984 voting tickets Electoral and Referendum • Extended prisoner disenfranchisement Amendment (Electoral • Increased donation disclosure threshold Integrity and Other • Earlier roll closure Measures) Act 2006 • Increased tax deductibility of donations • New proof of identity requirements New South Wales –– Parliamentary Electorates and Elections Act 1912 Constitution (Legislative • Reduced the size of the Legislative Council from 45 to Council) Amendment Act 42 seats 1991 Constitution (Fixed Term • Fixed four-year terms for parliament Parliaments) Amendment Act 1993 Constitution and • Reduced the size of the Legislative Assembly from Parliamentary Electorates 99 to 93 seats and Elections Amendment Act 1997 Parliamentary Electorates • Abolished group voting tickets for the Legislative and Elections Amendment Act Council 1999 • Tightened party registration criteria Parliamentary Electorates • Established the NSW Electoral Commission and Elections Amendment Act • Updated general administrative arrangements for the 2006 conduct of elections Victoria –– Electoral Act 2002 Electoral Act 2002 • Major re-write of earlier Act • Reformed the Victorian Electoral Commission • Registration of political parties Constitution (Parliamentary • Introduced proportional representation for the Reform) Act 2003 Legislative Council • Fixed, four-year terms for parliament Queensland –– Electoral Act 1992 Elections Act 1983 • Re-write of earlier Act Electoral Districts Act 1991 • Removed the zonal system, and established electorates based on ‘one vote, one value’, with exceptions for geographically large electorates • Introduced optional preferential voting Electoral Act 1992 • Re-write of the earlier Act • Established the Electoral Commission of Queensland Electoral and Other Acts • Reformed party registration rules Amendment Act 2002 • Introduced internal party democracy requirements

6

Western Australia –– Electoral Act 1907 Acts Amendment (Electoral • Introduced proportional representation for the Reform) Act 1987 Legislative Council • Introduced ticket voting for both houses (ticket voting for the Assembly was removed in 1988) • Four-year terms for parliament • Western Australian Electoral Commission created Electoral Amendment Act • Party registration 2000 Electoral Amendment and • Introduced ‘one vote, one value’ for the Legislative Repeal Act 2005 Assembly • Increased the size of the Legislative Council from 34 to 36 seats • Increased the size of the Legislative Assembly from 57 to 59 seats Electoral Reform (Electoral • Introduced public funding Funding) Act 2006 South Australia –– Electoral Act 1985 Electoral Act 1985 • Major re-write of earlier Act • Introduced compulsory and ticket voting for the Legislative Council • Party registration and identification on ballot papers Constitution (Electoral • Introduced a ‘fairness’ clause for House of Assembly Redistribution) Amendment boundary redistributions Act 1991 Tasmania –– Electoral Act 2004 Electoral Act 1985 • Major re-write of earlier Act • Introduced party registration Parliamentary Reform Act • Reduced the size of the House of Assembly from 35 to 1998 25 members • Reduced the size of the Legislative Council from 19 to 15 members Electoral Act 2004 • Major re-write of earlier Act • Established the Tasmanian Electoral Commission Australian Capital Territory –– Electoral Act 1992 Australian Capital Territory • Provided for House of Assembly elections to be (Self-Government) Act 1988 conducted using a modified d’Hondt system (Commonwealth) Electoral Act 1992 • Established the ACT Electoral Commission • Formed an electoral boundary redistribution process for electorates using the Hare-Clark system Electoral (Amendment) Act • Created provisions for elections using the Hare-Clark 1994 system Proportional Representation • Entrenched the following requirements of electoral law (Hare-Clark) Entrenchment by mandating that a referendum be held for any Act 1994 alteration to be made: odd number of members for each electorate (with a minimum of five members); compulsory voting; full preferential voting; Robson Rotation; and no ticket voting Northern Territory –– Electoral Act 2004 Electoral Act 2004 • Major re-write of earlier Act • Introduced party registration Source: Australasian Legal Information Institute (Austlii) –– www.austlii.edu.au

7

It can be seen from Table 1.1 that six jurisdictions have replaced their principal legislation with new Acts in the past 25 years. However, the remaining three jurisdictions –– the Commonwealth, New South Wales and Western Australia –– have also made significant amendments to their principal Acts during the period. Taken together, these reforms amount to a substantial level of legislative activity in the electoral policy area. Many of these amendments have had differential impacts and consequences for voters, candidates, and parties. For this reason, the fairness of the various reforms needs to be assessed.

Research Questions and Framework

This thesis undertakes a comparative assessment of electoral reforms, based on principles of fairness and equity for election participants. The central question that the thesis seeks to answer is:

• Have electoral reforms over the past 25 years enhanced or diminished principles of fairness for Australian voters, candidates and parties?

In answering the central question, the thesis addresses the following questions about Australian electoral reforms in the study period: • Are these reforms consistent with internationally accepted norms for electoral management and fair elections? • To what extent does the institutional design of Australian electoral management provide protection against governing party control? • What are the drivers for reform –– that is, how and why do electoral reforms occur?

These questions will be addressed by a comparative study of the significant electoral reforms that have occurred between 1983 and 2007 in Australia’s nine legislative jurisdictions: the national (also referred to as the Commonwealth, or Federal) jurisdiction; the six States (in order of electoral population –– New South Wales, Victoria, Queensland, Western Australia, South Australia, and Tasmania); and the two mainland Territories –– the Australian Capital Territory and the Northern Territory).

8

Although the Australian offshore Territory of Norfolk Island has its own Legislative Assembly, it is not included in this study because its political and electoral systems are significantly different from those of other Australian jurisdictions.5

The 1983–2007 study period was selected for a number of reasons. First, it enables the assessment to capture the major Commonwealth reforms that took place in 1983–84. These reforms, which occurred soon after New South Wales introduced Australia’s first public funding and party registration regime in 1981, provided the impetus for similar reforms in State and Territory jurisdictions in the following years. Second, the timeframe allows for the inclusion of the most significant reform agenda at the Commonwealth level since 1984 –– that is, the Howard Government changes in 2006 to the Commonwealth Electoral Act 1918 . Finally, the 25-year timeframe enables a relevant assessment leading into the next wave of electoral reform proposed in 2008 by the Rudd Labor Government and some State governments.

In order to answer the research questions, the thesis first undertakes an analysis of the institutional structure of Australian electoral systems, including the roles of parliamentary committees and electoral management bodies. As part of this analysis, the independence of electoral management bodies is assessed. Overall, this institutional analysis provides an account of the legislative and administrative environments in which electoral laws are amended and administered, including an explanation of the drivers at play in the reform process. Building on this foundation, specific electoral issues and reforms are then examined to determine how the reform process operates. By measuring the impacts of these reforms, assessments are made of which groups are advantaged and disadvantaged. This part of the thesis investigates specific issues and reforms in every jurisdiction to provide a comprehensive case study assessment of each. The jurisdiction-by-jurisdiction analysis identifies electoral issues worthy of comparative treatment. The thesis then culminates in an examination of several issues that have relevance across a range of jurisdictions.

Based on an initial review of the available literature on Australian electoral and party systems, it is hypothesised that:

5 For example, Norfolk Island, with a population of about 2,000, lacks political parties, has no separate electoral management body, and uses a rare voting system of first-past-the-post voting in multi-member electorates, where voters can lodge multiple votes for one candidate (once referred to as ‘plumping’).

9

• The introduction of the 1983–84 Commonwealth electoral reforms has had mixed results in terms of fairness; • At the Australian sub-national level, while reforms have mostly improved equity in electoral outcomes, access to political participation has diminished for candidates and parties; • Due to the domination of the Australian Labor Party and the Liberal/National Coalition in Australian parliamentary politics, party cartelisation has intensified during the study period, benefiting major parties at the expense of minor parties and independent candidates in electoral system design and election outcomes; and • Cartelisation also acts against the adoption of ‘best practice’ international norms of electoral management and fair elections.

Thesis Structure

The literature relevant to this research is discussed in Chapter 2, while the methodological approach, upon which the research is based, is explained in Chapter 3. In Chapter 4, the institutional framework for the development of Australian electoral legislation and the administration and conduct of elections, is described and assessed. Chapter 5 addresses the question of how electoral systems are controlled by analysing the levels of independence exercised by Australia’s electoral management bodies.

Chapters 6 and 7 concentrate on selected reform processes that have occurred over the past 25 years: Chapter 6 details reforms at the Commonwealth level, and Chapter 7 examines reforms in the State and Territory jurisdictions. Chapter 8 presents a comparative analysis of selected aspects of electoral systems across the various jurisdictions, and the impacts of reforms in these areas on electoral competition and election participants. The final chapter, Chapter 9, evaluates the research findings and assesses current Australian practice against international standards for electoral management and fair elections.

10

11

Chapter 2

Literature Review

This chapter discusses the body of scholarship to which the thesis contributes. Detailed references to literature specific to chapter themes are also included in the relevant chapters.

A key issue in the fairness of electoral systems is the extent to which they support equitable competition between a range of candidates and parties. There is a rich international literature theorising the relationship between voting systems and party systems, with Maurice Duverger, Stein Rokkan and Josep Colomer being three of the more notable theorists in this area. Others, such as Arend Lijphart and Kenneth Benoit, explore the relationship between voting systems and broader patterns of majoritarian or consensual political institutions. These bodies of literature form the backdrop for a more direct engagement with the comparative literature on electoral regulation and management, which is the focus of this thesis. In particular, I draw on research on the democratic assessment of electoral management, as well as the literature directed towards electoral management practitioners, such as guides produced by the International Institute for Democracy and Electoral Assistance (IDEA) and the Administration and Cost of Elections (ACE) Project.

The major topic of this thesis is the comparative assessment of sub-national levels of electoral management, as well as the national system. Few other studies have attempted such an assessment and there is little literature on which to draw. While there are useful historical studies of Australian electoral institutions, these generally concentrate on only one jurisdiction, and are often produced or commissioned by the institutions themselves.

12

One example is Dean Jaensch’s study of South Australian electoral history (Jaensch, 2002). Studies of particular reforms, such as the history of malapportionment in Western Australia (Phillips and Robinson, 2006), or of particular issues, such as political finance (Young and Tham, 2006), are also relatively common. There has, however, been no previous attempt to focus on the particular reform period analysed in this study (1983–2007), or to undertake a comparative assessment of the reform processes in all nine Australian legislative jurisdictions. While Graeme Orr’s report on electoral systems for the Democratic Audit of Australia (Orr, 2004) provides a useful cross-jurisdictional analysis, Orr did not conduct a comparative analysis of the electoral impact of the issues discussed, nor did this commendable work adopt a longitudinal approach.

The relevant literature for this thesis can be grouped according to four primary themes: • Theory of electoral systems; • Fairness measures; • Studies of electoral institutions; and • Australian electoral institutions.

Literature in the first category, the theory of electoral systems, incorporates research on voting system design and attempts to develop propositions that apply to voting systems world-wide. This research encompasses issues such as the relationships between voting systems and political party systems, institutional design theory, path dependency and the differing applications of design theory in established or developing democracies, and the development of party cartelisation theory. Second, the review considers definitions and measures of electoral fairness, both qualitative and using quantitative measures of electoral outcomes. Third, the literature on electoral institutions is reviewed. This literature includes comparative studies of voting systems and electoral regulation, at the international and national levels, as well as the recent growth in practitioner-targeted publications that develop international standards for electoral management. Finally, the review assesses the literature on Australian electoral institutions, which addresses both the design of electoral systems and electoral reforms. This literature analyses the underlying motives for Australian reforms, and measures of the subsequent impacts of reforms.

13

Theory of Electoral Systems

To properly assess electoral systems and the processes of reforming electoral laws, an appreciation of the relevant political science theories is required. For the purposes of this study, voting and electoral system design theory and party cartelisation theory are the most relevant. These theories are used in the analysis that is undertaken in the following chapters.

Voting and Electoral System Design Theory

During the past half-century, political scientists such as Duverger, Rokkan, Colomer, and Lijphart have developed theories about whether electoral institutions shape political party systems or whether it is actually the party system that determines the design of electoral institutions. In the 1950s, Duverger contended that party systems are shaped by the electoral institutions in which they act. He argued, accordingly, that majoritarian, single-member voting systems create two-party systems, while proportional voting systems with multi-member electorates produce multi-party systems (Duverger, 1951).

Duverger’s theory appeared to be confirmed by Rae’s 1967 study of 20 western democracies, which showed a very strong correlation between majoritarian voting and two-party systems, and between proportional voting and multi-party systems (Harrop and Miller, 1987: 69). It is logical that a majoritarian system which disproportionately favours larger parties will limit the ability of a multiplicity of parties to exist. Likewise, proportional systems that provide a lower threshold for entry will allow more parties a share of the representation that is denied them in majoritarian systems. Rae’s study does not, however, address questions of causation in a definitive manner –– that is, it does not show whether party systems are a determinant or an effect of voting systems (Harrop and Miller, 1987: 70). Rae argues that both voting systems and party systems apply determining force upon each other, among numerous other forces, and that it is impossible to quantify the relative importance of each (Rae, 1967: 141).

In contrast to Duverger’s theory, Rokkan (1970) puts forward the hypothesis that governing political parties shape voting systems. In support of this hypothesis, Rokkan cites the example of Western Europe in the early 20th century, where established

14 political parties and groups introduced proportional representation to ensure their own survival against the rise of left-wing labour groups, which was brought about by the extension of the franchise (Rokkan, 1970; Benoit, 2004: 364). In his review of Rokkan’s work, Benoit describes the ongoing tension that exists between voting systems and parties, with each attempting to shape the other, until a level of equilibrium is achieved (2004: 364). In Benoit’s scenario, the stability of a democracy is determined by the manner in which this tension is managed. Like Rokkan, Colomer argues that it is the political parties that choose voting systems, and that over time this results in a trend towards proportional representation and away from majoritarian systems (2005: 2). Colomer further concludes that voting systems simply reinforce the pre-existing party relationships.

American political scientist George Tsebelis (1990) presents the French experience as an example of governing parties manipulating the voting system for their own benefit. In the 1980s, the French Socialist government introduced proportional representation to counteract the majoritarian system that disadvantaged the Socialists. However, this change was not sufficient to ensure a leftist coalition government. The succeeding right-wing coalition subsequently reinstated the two-round majority system (Tsebelis, 1990: 223–24).

Overall, it can be argued that in electoral democracies there is an ongoing tension between voting systems and political parties, with each attempting to exert dominance over the other. This tension is explained by Benoit (2004) as the mechanical functioning of voting systems against the psychological behaviour of political parties. Benoit argues that voting systems place mechanical restrictions on how votes are distributed and seats are awarded, while political parties use psychology to shape campaign strategies and thereby maximise results within the voting system’s constraints (2004: 364). Lijphart (1994, 1999) assesses the institutional consequences of this tension, arguing that proportional voting systems produce stronger electoral institutions due to the tendency towards non-majority governments in such systems.

Unlike many countries, Australia has a largely homogenous political landscape, with few significant differences in party support or structure across the various regions. In all nine Australian jurisdictions, a strong two-party system exists, comprising Labor and the Liberal/National Coalition. Queensland is the only jurisdiction where there is a

15 significant difference in the balance of the Coalition parties, with the National Party being the dominant coalition partner in that State.6 Duverger’s theory would suggest that this party system is a result of single-member electorates, which exist for the lower of the two houses (in bicameral systems) or the only house (in unicameral systems) of seven of the nine Australian jurisdictions. The two exceptions are Tasmania and the Australian Capital Territory, whose electoral systems are based on regional proportional representation (and where the Greens have played important roles as a third party).7

Duverger’s and Rokkan’s theories cannot be straightforwardly applied to the Australian context because of the pressure placed on governments in bicameral parliaments (which exist in six of the nine jurisdictions) to have distinctly different voting systems for each house of parliament. Such pressure was a contributing factor in the introduction by Labor governments of regional proportional representation for the upper houses of Western Australia (in 1987) and Victoria (in 2002). In the Western Australian case, however, Labor was successful in passing the reform legislation by obtaining the support of the National Party –– an uncommon instance of the parliament breaking with the traditional two-party conflict. Another example of the shift away from the traditional party system is the recent development in which the sole National Party member in the South Australian parliament, Karlene Maywald, entered into a coalition arrangement with the Labor Government, where she sits as a Cabinet Minister.

The genesis of Australian electoral system design took place in the mid 19 th century, when the British parliament began granting self-government to Australian colonies. Electoral system design began well before the development of a party system, which occurred in the late 19 th and early 20 th centuries. This meant that once they were formed, parties that achieved government had either to accept the existing nature of the system, or be seen as acting in self-interest to amend the system for their own advantage. While current Australian electoral systems retain some aspects of the colonial systems (such as secret voting and a degree of administrative independence), it can be expected that the findings of this study will be consistent with Tsebelis’ (1990) theory, that parties use their position to seek electoral advantage.

6 In July 2008, the two parties agreed to merge into one party, the Queensland Liberal National Party. 7 Tasmania uses single-member electorates for its upper house, the Legislative Council.

16

The ongoing tension referred to by Lijphart and Benoit is described by Offe (1996) as competing forces of institutional change and stability (1996: 207). Offe argues that there is an inherent institutional tendency to remain stable. Importantly, however, he also identifies the capacity for institutions to change from within, through legislated powers for institutions to alter their own lower-level operational and administrative mechanisms (1996: 209). This is pertinent for Australian electoral system design, where there can be a tension between electoral commissions seeking to act independently in accordance with professional norms, and governing political parties attempting to assert a preferred electoral regime for identified or perceived partisan advantage. Administrative changes that do not require legislative approval may have significant partisan impacts and the Australian experience has generally been that the governing political parties keep tight controls over this area of electoral regulation.

Goodin identifies three ways that institutional change occurs –– by accident, evolution, and intention (Goodin, 1996: 24). While Goodin applies this model to institutions in general, it is the latter two ways that change occurs that are important in examining electoral system design. Change based on evolution is closely related to path dependency. That is, the design of a system reflects its origins, when critical early decisions have a major impact on the future direction and shape of an institution (Pierson, 2000; Schwartz, 2004). Through these early decisions, institutions can become ‘locked in’ to a particular path of development (Liebowitz and Margolis, 1995).

The history of Australian electoral institutions indicates that path dependency plays a critical role in electoral design and as a consequence, impacts significantly on the representative nature of the democracy. For example, the Northern Territory’s self- governing legislation –– the Northern Territory (Self-Government) Act 1978 –– stipulates single-member electorates for a unicameral Assembly (MEC, 2003: 9). As a result, there has been only one change of government in more than 35 years, with the government always having an Assembly majority. While the Northern Territory Labor Government’s electoral reforms in 2004 were extensive, the Commonwealth legislation prevented it from considering changes to the single-member majoritarian voting system, demonstrating that electoral reform is restricted to a particular path.

In comparison, the Australian Capital Territory’s (ACT’s) voting system, which was designed less than 15 years after the Northern Territory’s first election under self-

17 government, is regulated by the Commonwealth’s Australian Capital Territory (Electoral) Act 1988. This Act specifies a proportional representation system for a unicameral Assembly. In the twelve years from 1989 there were four changes of government (including two changes of government between elections). In addition, only one government has held an ACT Assembly majority since 1989 (the 2005–08 Stanhope Labor Government). When the ACT considered electoral reform in the early 1990s, a referendum was held to determine voters’ preferences for single-member electorates or proportional representation, with two-thirds of electors supporting the latter (ACTEC, 2006: 56). The original Commonwealth legislation had set the ACT on a path of proportional representation. The reasons for such different approaches to the voting system design for the two Territories can be found in the timing of the legislation –– for the Northern Territory legislation, the Fraser Coalition Government had control of the Senate. For the ACT legislation, in contrast, the Hawke Labor Government had a Senate minority, and therefore needed to negotiate with the Australian Democrats, a minor party that benefited from proportional representation. These brief examples are indicative of the importance of path dependency in the evolution of electoral systems.

The third type of change identified by Goodin –– intentional change –– is the most common way in which electoral institutions are altered. As the thesis will show, intentional changes to electoral systems are typified by the ongoing tension of political parties attempting to assert dominance over institutions. However, as the two cases above illustrate, intentional electoral reform is shaped by the original electoral system design and by the subsequent evolution of the system based on path dependency.

Tsebelis (1990) and Benoit (2004) argue that the design characteristics of electoral institutions can be differentiated from traditional rational choice theory, which suggests that institutions will evolve into more efficient and cooperative entities in which all players benefit. In contrast, electoral institutions work in a competitive environment, where there is a zero sum gain, as wins for one set of groups and individuals are redistributed as losses to other groups and individuals (Tsebelis, 1990: 104; Benoit, 2004: 366). Benoit further develops a model of electoral system change based on the assumption that political parties select voting systems that maximise their own seat- winning potential. He identifies this overriding self-interest as the primary motivation for electoral system reform (Benoit, 2004: 373). Conversely, it can be expected that reforms will not occur if the governing political parties do not see any electoral

18 advantage in changing the system. As the beneficiaries of an electoral system, it is the motivator of self-interest that encourages parties to resist any changes (Cox, 1997: 18). As Bowler, Donovan and Karp (2006) note (citing Benoit and Rokkan, as well as Boix, 1999; and Grofman, 1990), an assumption of electoral system change theory is that governing parties act to protect self-interest, requiring a strategic reason to change the electoral laws under which they came to power (2006: 435).

Party Cartelisation Theory

Party cartelisation theory supports Rokkan’s and Colomer’s view that governing parties collude to use state resources to assist their own political and electoral survival (Katz and Mair, 1995: 5). Australian political scientist Ian Marsh presents an extensive assessment of possible party cartelisation in Australian politics (Marsh, 2006). He points out that the major parties can use their control of government to limit the ability for new, smaller parties to participate, particularly where public funds and resources are allocated to incumbent parties to assist their participation in elections (Marsh, 2006: 1). Interestingly, Marsh identifies the election of the Hawke Labor Government in 1983 as the beginning of a new phase of major party cartelisation, with the convergence of Labor and Coalition economic policy (2006: 6–7). Simultaneously, Labor was introducing electoral reforms for party registration and public funding regimes that created thresholds for competing in elections and accessing resources.

Ian Ward (2006: 89) and Sally Young (2003) both support Marsh’s assertion that party cartelisation is present in the Australian context, arguing that Labor and the Coalition parties coalesce to use state resources, both directly to subsidise their own parties, and to provide administrative and campaigning resources to incumbent members of parliament. In contrast, as Rodney Smith and John O’Mahony argue, the growth of minor parties, such as the Democrats, Greens and One Nation, and the lowering of effective thresholds for some houses of parliament, indicates a lack of strong party cartelisation (2006: 110–4).

Smith and O’Mahony do note an exception to their argument: the 1998 reduction in size of the Tasmanian parliament is a clear example of the Labor/Coalition cartel acting to thwart a new party challenger, the Greens (Smith and O’Mahony, 2006: 111–2). In their assessment of the growth of the Greens as a challenger to the two-party system,

19

Ariadne Vromen and Nick Turnbull (2006) identify obstacles faced by parties that operate outside the governing cartel. They argue that while the Greens may benefit from electoral reforms such as the introduction of public funding, the cartel’s reluctance to place limitations on corporate donations severely disadvantages smaller parties (Vromen and Turnbull, 2006: 177–8).

Arguing that there is not a strong cartel system in Australia, Murray Goot points out that most electoral reforms since the 1970s have been opposed by the major opposition party (2006: 190). Goot contends that these reforms have generally weakened the cartel parties’ firm control. In support of his ‘weak cartel’ argument, Goot notes the cartel parties’ reluctance to move away from compulsory voting, which he believes advantages smaller parties that attract the support of disaffected voters (2006: 191). Goot’s argument is that these disaffected voters would be likely to be non-voters in a voluntary system. The reluctance to introduce voluntary voting as a measure to limit minor parties may indicate a contest to gain ascendancy within the cartel. The shift to voluntary voting is generally considered to be more favourable to the Coalition and Labor has understandably been opposed to its introduction. However, despite having the numbers in the Senate from 2005 to 2008 to introduce voluntary voting, the Coalition was reluctant to initiate this reform. This was perhaps due to the uncertainty of impacts and the public opprobrium that such a reform may create. These differences in party positions can be seen in the Joint Standing Committee on Electoral Matters’ (JSCEM’s) 2004 election report, which extensively discussed the issue of compulsory versus voluntary voting, with Labor’s committee members tabling a minority report opposing the introduction of voluntary voting (JSCEM, 2005a: 372–3). 8

Fairness

The objective of the thesis is to determine whether electoral reforms have enhanced or diminished principles of fairness for election participants. Measurements of fairness in electoral systems take two distinct approaches. The first approach is based on philosophical concepts of democratic behaviour, with an emphasis on qualitatively assessing how fairness is viewed by participants in electoral contests –– that is, by

8 The JSCEM report includes a lengthy discussion on the arguments surrounding compulsory and voluntary voting, pp.183–204.

20 voters, candidates, and political parties. The second approach to studying fairness utilises quantitative measurements of electoral outcomes. While the two approaches overlap, as they both seek to measure the ability for citizens to meaningfully participate and express their choices in a democracy’s electoral processes, each maintains distinctive methodological techniques.

Defining Fairness

There is much debate over what actually constitutes fairness in relation to electoral institutions. In his definitions of the fairness of voting systems, Adrian Blau identifies five general concepts, which are not necessarily compatible: equality; populist; winner- takes-all; majority; and plurality (2004: 167). For example, Blau argues that while a strength of the populist , or majoritarian concept is that it supports the democratic principle of popular control of government, it can be in conflict with the equality concept, which calls for electors’ votes to be reflected in parliamentary representation (2004: 167, 177). While the principle of ‘one vote, one value’ is often used to describe the equality concept, such a measure can become problematic when considering the efficacy of a given vote. In a system of single-member electorates based on ‘one vote, one value’, it could be argued that voters’ choices in ‘marginal’ electorates have far more political weight and effect than votes in ‘safe’ seats.

With a focus on ballot access and design, British philosopher Michael Dummett defines fairness with reference to two imperatives: fairness to voters; and fairness to candidates. Dummett argues that when these two concepts are in conflict, it is the former that should prevail (1997: 109). For example, more stringent nomination rules may impede potential candidates from participating in an election, but also result in voters having a simplified ballot paper that provides a clearer choice and reduces accidental informal voting. If the choice of candidates becomes too limited, however, this lack of choice could also be unfair to voters. Another viewpoint is provided by Martin Harrop and William Miller, who describe fairness (in terms of strong government and fair representation) as a tension between ‘majority rule and minority reassurance’, and that the importance of one or the other will be determined by the specific circumstances of the democracy in question (1987: 42–3).

21

American political scientist Pippa Norris, who specialises in electoral democracy studies, extends the ‘fairness to voters’ concept to incorporate sub-groups of a democracy, as in measures to ensure fair representation for ethnic minorities (2004: 216). Such measures can be introduced through the design of the voting system, using either proportional systems (encouraging diversity of representation through reduced quotas and the use of list systems) or majoritarian systems. For example, New Zealand’s minority Maori community has had reserved seats in parliament since the 1860s, under both the previous first-past-the-post majoritarian system and the current Mixed Member Proportional system, with Maori voters having the choice to enrol on either the Maori or general rolls.

Canadian political scientist John Courtney (2004) cites an example in which ‘fairer’ representation for ethnic minorities is encouraged through electoral regulation, in this case by the use of boundary-drawing rules. Courtney cites Nova Scotia’s majoritarian system, where reserved race-conscious redistricting is implemented for Acadian and Black communities by creating four ethnic seats with lower populations than the provincial average. This is also known as ‘affirmative gerrymandering’ (Courtney, 2004: 61–4).

In their assessment of whether the Australian electoral system is ‘free and fair,’ Mercurio and Williams (2004) refine the two themes of freedom and fairness into four underlying principles: participation; free-conscience; election outcome; and knowledge (2004: 367–8). The participation principle requires all citizens (subject to reasonable limitations, such as age and mental capacity) to have equal opportunity to participate in elections as candidates, in the formation of political parties, and as voters. The free- conscience principle requires that voters are able to cast their vote without coercion or intimidation. The secret ballot goes a long way to ensuring this in Australia. The election outcome principle is based on the accurate counting and reporting of votes cast to ensure that people have confidence in the electoral process and outcomes. The knowledge principle requires that voters understand how the electoral process works, including enrolment and attendance requirements, and that the voter can make an informed choice when casting a vote, based on a knowledge of the candidates’ and parties’ respective policy positions and statements (Mercurio and Williams, 2004: 367– 9). In the case of above-the-line ticket voting, it is also important for voters to know how a party’s preferences will be distributed to other parties and candidates.

22

In studying New Zealand’s electoral system, Banducci and Karp (1999) base their assessment of fairness on the results of voter surveys, using indicators such as trust in government, perceptions of big business involvement, and connectedness with members of parliament. Their work uses data from the 1993 and 1996 New Zealand Election Studies. Similarly, data from the Australian Election Study, which conducts surveys after each Federal election (beginning with the 1987 election), and the Australian Survey of Social Attitudes provide information on confidence and trust in political parties and elected representatives. Drawing such national-level data together, the Comparative Study of Electoral Systems 9 provides extensive information on voter attitudes from 50 countries.

It can be seen that there are many definitions and interpretations of the fairness of electoral systems. For the purpose of this study, reference will be made to many of the concepts described above, as appropriate. However, emphasis will be placed on two of the principles discussed by Mercurio and Williams. The first is participation, that is, equality of opportunity to participate in elections as voters, candidates or as members of parties. The second is knowledge of the electoral process, so that voters can make an informed choice. While Mercurio and Williams’ other two principles –– free- conscience and electoral outcomes (in the qualitative sense) –– are equally important, they are currently considered less contentious or not under threat in the Australian setting.

Measuring Electoral Outcomes

Outlining the quantitative approach to determining the fairness of electoral outcomes by analysing election results, Goot (1985) identifies three primary approaches. The first, developed by Soper and Rydon (1958), seeks to measure a party’s ‘effective vote’. This is done by determining the vote required for a party to win its seats in the absence of bias, based on a ‘constructed median’ calculated by averaging a party’s support level in five seats either side of its median support base (Goot, 1985: 215). This concept emphasises the importance of marginal seats over safe or ‘unwinnable’ seats, but has subsequently been criticised by Jaensch (1977) and Maley (1980). Maley has argued

9 http://www.cses.org

23 that it is not valid to make assumptions about how electors would have voted if they had not been in a marginal or safe seat.

The second quantitative approach to measuring the fairness of electoral outcomes, developed by Hughes (1978), is to determine the notional two-party preferred (2PP) vote required to win a majority of seats. This measure has become the dominant method for measuring electoral fairness in Australian elections. Since World War II, the Federal elections in 1954, 1961, 1969, 1990 and 1998 (out of 25 elections) are notable for returning governments with less than 50 per cent of the 2PP vote (see AEC, 2007). Such results are argued to be unfair. South Australia is the only Australian State that has inserted a ‘fairness’ clause in its electoral legislation, specifying that electoral redistributions should endeavour to create results in which the party or coalition of parties that achieves the majority of the 2PP vote also achieves a majority of seats (Section 83 of the Constitution Act 1934 , inserted in 1991; also see Newton-Farrelly, 2008).

The third approach to fairness measurement identified by Goot comprises Rae’s Indices of Electoral Fractionalisation and Legislative Fractionalisation. These indices can be used to determine the variation between the support for parties contesting an election and their subsequent parliamentary strength. This is calculated by ‘adding the squares of the decimal shares of the votes or seats won by the parties at a given election and subtracting this from unity’ (Goot, 1985: 216). For example, in the 2004 Federal election, the Index of Electoral Fractionalisation was 0.68, while the Index of Legislative Fractionalisation was 0.59. It is argued that when using this method, a perfectly fair outcome will result in identical indices. However, such a result rarely, if ever occurs. In addition, this measure aggregates party results, and therefore obscures distorted outcomes for individual parties, both in a negative and positive sense. For example, parties might have broad appeal but insufficient support to win seats in single- member electorates; for example, the Greens received 7.2 per cent of the vote nationally in 2004, but did not win any lower house seats. Furthermore, distortion can occur for a party running a few candidates in areas with concentrated support, as regularly happens for the National Party. In 2007, the National Party contested only 24 of the 150 House of Representatives’ seats, but won 10 seats, or 6.7 per cent of the total seats, with 5.5 per cent of the nation-wide vote. The weakness inherent in these fractionalisation

24 measures is that they can produce a ‘fair’ result by balancing out over- and under- represented parties.

Subsequent to Goot’s survey, a more effective method of taking into account the disproportionality of individual party results was developed by Irish political scientist, Michael Gallagher. Gallagher’s Least Squares Index involves squaring and tallying the differences between each party’s proportion of votes and seats. The Index value is the square root of half of the total (Gallagher and Mitchell, 2005: 603; see Appendix C). While other indices, such as the Gini and Dauer-Kelsay indices, are useful in measuring the degree of malapportionment that exists in a voting system, the Gallagher Index measures the actual impact that such systems have on representation, and will therefore be used throughout the thesis. As noted by Western Australian voting system specialist Jeremy Buxton about measures of proportionality, malapportioned voting systems can still provide proportional results, depending on the spread of votes for the various parties (Buxton, 1985: 57). Consequently the Gallagher Index is used as an indicator of impact, rather than as a strict measure of the apportionment of a system.

Measures of proportionality will be used in this thesis as a way of determining the level of bias that is present in each system, and also as a longitudinal measure of the impacts of reforms. For example, shifts from single or dual-member constituencies to proportional representation in the upper houses of Western Australia and Victoria can be expected to improve proportionality. However, this study will also compare these regionally-based systems against a State-wide proportional system, to measure the impact of seat magnitude.

Studies of Electoral Institutions

Any comparative evaluation of Australian electoral reforms needs to draw on the frameworks used in the existing comparative literature on voting systems and electoral regulation. In addition, there is a burgeoning applied literature targeted at electoral administrators and seeking to promote electoral best practice. Such work is highly relevant in assessing how Australian reforms relate to internationally accepted norms for electoral management and fair elections.

25

Comparative Studies of Voting Systems and Electoral Regulation

Few comparative evaluations of national electoral systems exist in contemporary literature. One of the best comparative assessments of electoral laws is provided by Louis Massicotte, Andre Blais and Antoine Yoshinaka (2004), who examine regulatory factors in 63 countries, covering issues such as enrolment, voting and counting procedures. This study expands on the authors’ earlier work, which examined seven restrictions on voting in the same 63 countries: age; mental capacity; citizenship; residence in country; citizens abroad; and prisoners’ rights (Blais, Massicotte and Yoshinaka, 2001). Michael Gallagher and Paul Mitchell (eds, 2005) used a case-study approach to investigate the electoral systems of 22 countries, addressing their system origins, workings, reforms and impacts. Lijphart (1994) examined aspects of 27 democracies, including the impacts of electoral reform on disproportionality.

Vernon Bogdanor and David Butler (1983) provide a narrative approach for 13 (mainly European) countries. This includes useful information on the disproportionality of election results. Colomer (ed., 2004) expands on the work of Bogdanor and Butler, both in terms of detail and the number of countries examined. This work provides numerous examples from all populated continents, including established, transitional and emerging democracies and encompassing majoritarian and proportional systems. However, due to the breadth of this coverage, the comparative aspect of the assessment is limited, partially a result of the different scope adopted by the various chapter authors. Andrew Reynolds and Ben Reilly (2002) provide a tabulation of the voting systems of all of the world’s states, which is useful in explaining the systems in place and their related terminology. Yet this work lacks a critical comparative evaluation of the merits and detriments of these systems.

In cases where particular aspects of electoral regulation are addressed in a comparative assessment, more depth can be provided to the analysis, especially when the range of countries being assessed is limited. Such studies are often conducted with one specific country as a focus, and other countries used as comparators. A recent example of this approach is Graeme Orr’s (2008) critique of expatriate voting, with Australia being the focus of the study, compared against the United Kingdom, Canada and New Zealand.

26

In recent decades, some researchers have developed assessment frameworks for comparing electoral systems, and measures of democracy in general. These frameworks tend to be either ratings-based, comparison-based (as in Massicotte, Blais and Yoshinaka’s study, above), or discussion-based. An example of a ratings-based system is provided by Jørgen Elklit and Reynolds (2002; 2005), who use quantitative ratings of aspects of electoral systems in their development of an evaluation framework. A good example of a discussion-based model is the Democratic Audit framework, which examines electoral issues as part of a broader assessment of democracies. The main aim of the audit approach is to identify the strengths and weaknesses of a system and, through the publication of papers and reports, to stimulate public debate and action for improvement (for example, Orr, 2004; Young and Tham, 2006). Competing assessment frameworks are discussed more fully in the following chapter, in the course of developing a framework suitable for the purposes of this thesis.

International Norms for Electoral Management

The growth in the number of democracies worldwide has spawned a similar growth in activity to provide assistance in the development of electoral systems. Prominent in this work has been IDEA, a Stockholm-based intergovernmental organisation that regularly produces information on best practice for electoral administration and management. Typical of these works is the IDEA handbook on electoral management design (Wall et al., 2006), IDEA’s code of conduct for professional and ethical elections (1999) and its guidelines on the legal frameworks for elections (2002). In a similar way, the ACE Electoral Knowledge Network, which is partially funded through IDEA and the United Nations Development Programme (UNDP), regularly produces web-based information directed to electoral practitioners. Information includes comparative data on most aspects of electoral administration, which not only identifies best practice, but through comparative analysis, assesses how widespread such practices are.

The UNDP also independently produces reports on electoral governance issues, such as maintaining electoral management bodies within broader state structures (for example, López-Pintor, 2000). The International Foundation for Electoral Systems (IFES), which is largely funded by the United States Agency for International Development, is another organisation that seeks to improve electoral procedures and outcomes, with a particular focus on developing democracies. In doing this work, it also identifies preferred

27 procedures and trends in modern regulation (for example, see Blanc et al., 2007). Other quasi-governmental organisations also regularly fund publications in this area, as in the Inter-Parliamentary Union’s funding for the development of a code of conduct for elections (Goodwin-Gill, 1998).

Combined, the growth of the international governance industry has enabled relevant organisations to produce a relatively new body of work that contributes to the development of international norms for electoral management. The publications are primarily produced by electoral practitioners, and are designed for use by other practitioners. These organisations (particularly IDEA, ACE, and IFES) typically have web-based databases that allow for research on the existence and prevalence of current practices, including detail on major and relatively minor aspects of electoral design and administration.

Australian Electoral Institutions

The final area of literature to be reviewed is the valuable body of work focused on Australian electoral institutions. This literature addresses electoral system design, institutional structure, and electoral reforms. When considering electoral system design, it is important to note that Australian electoral systems were initially developed in the mid 19 th century in the absence of an effective party system. This allowed electoral management institutions to be designed as independent structures which were reasonably resistant to political interference (Sawer, 2001a). As a result of path dependency, this independence has been maintained to some extent, despite the development of a party system and the ability of governing parties to adapt electoral systems to their own advantage. The continuing influence of path dependency has meant a cultural resistance within electoral management bodies to change based on partisan self-interest, in the face of governing parties’ legislative ability to bring about such change.

Discussing institutional design in the Australian context, David Farrell and Ian McAllister (2005) provide evidence of major party domination of electoral systems. They argue that this party dominance has been brought about by the compulsory, fully- preferential nature of Australian systems, which instils in voters high levels of party

28 identification and, for parliamentary representatives, very high levels of party discipline (2005: 13). However, while there may be strong major party control over Australian voting systems, reforms that benefit minor parties do occur on occasions. Two such recent examples are the reforms of the Western Australian and Victorian upper houses, which moved from single-member elections (both States had dual-member electorates, with one member elected at each alternate election) to proportional representation (Farrell and McAllister, 2005: 50–51). As a result, minor parties have been elected to balance-of-power positions in each of these houses.

Goot (2006) argues that such reforms provide counter-evidence to the cartelisation thesis. Moreover, he points out that while the cartel parties are the major beneficiaries of increased incumbency entitlements (through having more members), minor parties also benefit, albeit to a lesser degree (2006: 205–6). In support of the cartelisation thesis one might argue that while minor parties may also benefit from incumbency entitlements, the resource gap between major and minor parties inevitably increases. The major parties remain adept at limiting the potential of minor parties through electoral mechanisms. Such mechanisms include the electorate magnitude for proportional representation, party registration requirements and public funding thresholds.

There exists a large disparity in the capacity to contest elections, which is determined by access to incumbency benefits. The disparity between incumbents and contestants without access to incumbency benefits has developed in recent decades. Incumbency benefits include parliamentarians’ travel, staffing, printing and communication entitlements, media coverage (such as involvement in televised debates), publicity resources (including access to public service resources), and voter databases (for example, see Young, 2003; Hughes and Costar, 2006; Young and Tham, 2006). There is a hierarchy of incumbency advantages (depicted in Figure 2.1), ranging from those enjoyed by the party in government, through those enjoyed by parliamentary parties, by parliamentarians without party status, and by parties without parliamentary representation, to the relatively limited benefits available to Independents. While this skewing of the electoral playing field through incumbency advantages is mostly separate from Australian electoral system design, it is important to acknowledge the impact it has on the fairness of Australian elections.

29

Figure 2.1 Hierarchy of Incumbency

Government Ministers

Government, parliamentary, parliamentary party President/Speaker and registered party Government Whips resources Government Chairs of Committees Government backbenchers

Opposition leaders Shadow ministers, whips Parliamentary, parliamentary party and registered party resources Opposition whips Opposition backbenchers

Parliamentary and registered party Minor party members of parliament resources

Parliamentary Independent members of parliament resources

Registered party Registered political parties without resources parliamentary representation

Non-parliamentary Independent candidates

30

The fact that the current two-party system has now been in place for more than 60 years suggests an entrenchment of the perception of the voting public that there are only two parties to choose from. This perception is reinforced by the two parties’ control of the political agenda, campaign funds and the electoral system itself, despite the occasional opportunities that arise for smaller parties, as previously mentioned. Minor parties can win more support at the expense of the two major parties where proportional representation exists. To illustrate this, on a national basis in the past ten Federal elections, Labor and the Coalition have always received a lower level of primary voting support in the Senate than in House of Representatives seats. By comparison, the Australian Democrats, successful at winning Senate seats from 1977 to 2001, have consistently received a higher vote when contesting more winnable positions in the Senate (see Table 2.1).

Table 2.1 Support at Federal Elections (%) –– House of Representatives v. Senate 10 Election Labor Coalition Democrats HoR Senate HoR Senate HoR Senate 1983 49.5 45.5 43.6 40.0 5.0 9.6 1984 47.6 42.2 45.0 39.5 5.5 7.6 1987 45.8 42.8 46.1 42.0 6.0 8.5 1990 39.4 38.4 43.5 41.9 11.3 12.6 1993 44.9 43.5 44.3 43.0 3.8 5.3 1996 38.8 36.2 47.3 44.0 6.8 10.8 1998 40.1 37.3 39.5 37.7 5.1 8.5 2001 37.8 34.3 43.0 41.8 5.4 7.3 2004 37.6 35.0 46.7 45.1 1.2 2.1 2007 43.4 40.3 42.1 39.9 0.7 1.3 Source: AEC. 11

The academic works that are particularly relevant to this thesis include Murray Goot’s (1985) comprehensive review of major studies of Australian voting systems and reform impacts. Goot’s review has a particular emphasis on electoral regulation and management issues including compulsory voting, postal voting, absentee voting, equality, fairness, preferential voting and proportional representation. Canadian

10 Percentages are national averages, therefore they under-represent Democrats support when the party did not stand candidates in all House of Representatives seats. An accurate example of increased voter support in the Senate can be seen in the 1998 figures, when the Democrats stood candidates in all House of Representatives seats. 11 This thesis sources its election data from official electoral commission publications and websites, as well as the Australian Government and Politics Database (AGPD, based at the University of Western Australia. Due to the number of sources used, only the organisations are listed. Website addresses are found after References at the end of the thesis.

31 political scientist David Elkins (1992) provides an international perspective in his comparative work on the Canadian and Australian electoral systems. Elkins presents arguments about why Australian electoral reforms are primarily driven by the governing parties to obtain a partisan advantage, and not for reasons of equality or fairness. He goes on to show that such reforms, while often at the expense of other players, are commonly justified by the governing party as being an improvement to democracy (Elkins, 1992: 68). Farrell and McAllister (2006) have undertaken a comprehensive assessment of Australian voting systems, including analysis of levels of disproportionality for selected jurisdictions. This work provides an historical and theoretical background on the development of Australia’s electoral systems, as well as providing the necessary contemporary context. There are numerous other works detailing Australia’s electoral systems, including those of Colin Hughes (for example, 1977; 1986; and with Graham, 1976) and Dean Jaensch (for example, 1974; 1995). However, these works tend to concentrate on election processes and outcomes, and do not investigate reforms, or the reasons for them.

A substantial body of publications and data relevant to this study is produced by parliamentary committees and electoral management bodies. The Australian Parliament’s Joint Standing Committee on Electoral Matters (JSCEM) regularly assesses aspects of the Commonwealth electoral system, particularly through holding an inquiry after each Federal election. Although this inquiry process provides arguably the greatest opportunity for public feedback and discussion, the Committee’s reports are invariably heavily influenced by the policy of the government of the day, which normally has majority control of the committee. However, the inquiries generate important data and discussion through their public hearings and submissions. For example, more than 200 submissions were received after the 2004 Federal election (JSCEM, 2005a). These inquiries provide useful insight into the institutional behaviour of political parties, electoral bodies, and other key stakeholders in the electoral process. In addition, the Electoral Council of Australia, which includes the electoral commissioners of all Australian jurisdictions, provides basic comparative information on Australia’s various electoral systems, while the Australian Electoral Commission (AEC) and equivalent State and Territory electoral offices have readily accessible data on election rules and results.

32

The electoral commissions themselves also make valuable contributions to academic and public discourse through the production of annual and election reports, and research papers and commissioned work on specific electoral issues. For example, in recent years, the AEC has published several research papers on compulsory voting, showing public support to be at 71–74 per cent (Evans, 2006: 9); identifying young people’s attitudes to participation and voting behaviour (through the Youth Electoral Study –– Print, Saha and Edwards, 2004 and 2005; Edwards, Saha and Print, 2006); and identifying reasons for levels of informal voting (Medew, 2003; Dario, 2005). Examples at the State level include the South Australian State Electoral Office’s paper on optional preferential voting (SEO, 2004), the Electoral Commission of Queensland’s ballot paper survey reports (ECQ, 2002, 2005 and 2007), and the Victorian Electoral Commission’s paper on lowering the voting age (Gribbin, 2004).

In a study on the impact of electoral reforms on fairness, Orr (2004) provides an assessment of political equality in Australian electoral systems. Orr’s study includes useful tables comparing regulation across the nine Australian jurisdictions for factors including prisoner franchise, political finance, candidate nomination fees, voter enrolment, and party registration. This snapshot approach highlights inconsistencies between jurisdictions and forms the basis for explanations and arguments on each of these issues. Young and Tham (2006) present a similar treatment of political finance issues, such as public funding, donation laws and disclosure regulation.

The literature on Australian electoral systems also includes academic work providing more detailed analysis of the impacts of specific reforms. For example, Bob Darcy, Ian McAllister, Malcolm Mackerras, and Jonathan Kelley consider the effect that ballot paper positioning can have in influencing candidates’ chances of success (Kelley and McAllister 1984; Darcy, 1986; Darcy and McAllister, 1990; Darcy and Mackerras, 1993). In part, these works respond to the Commonwealth’s change in 1983 from alphabetical listings on ballot papers to positions being drawn by lot. Another example is the work on compulsory voting by Lisa Hill and Jonathon Louth (Hill, 2004; Louth and Hill, 2005), and Scott Bennett (2008), who investigate the relationship with turnout levels, international practice, and public attitudes towards compulsion.

Finally, at the State level, Norm Kelly (2004), Glynn Evans (2005) and Paul Strangio (2004), among others, address recent electoral reforms. Kelly assesses the move to

33 reduce malapportionment in Western Australia, identifying the partisan interest behind the reform. In Evans’ work, the introduction and impact of South Australia’s ‘fairness’ clause is assessed, with consideration given to possible changes or repeal of the clause. Strangio provides an interesting history of reforms to the Victorian Legislative Council over the past half-century, culminating in the move to proportional representation in 2003, with an emphasis on the evolution of Labor’s policy on electoral reform.

Gaps in the Literature

The preceding review indicates that some substantial work has been conducted on comparative evaluations of voting systems between countries. However, there is little evidence of similar work at the sub-national level and, particularly for the purpose of this study, of comparative analysis across Australian States and Territories. While the most valuable recent contributions in this area are Farrell and McAllister (2005) and Orr (2004), neither of these works attempts to understand the forces behind electoral reform. In addition, there is only a small body of work comparing electoral regulation and management.

This study differs from those described above, as it assesses all Australian jurisdictions, with an emphasis on reforms in the past 25 years. It examines how and why these reforms have occurred and, using principles of fairness, measures the impact that these reforms have had in advancing or impeding electoral democracy. Similarly, while works have been published on particular reforms (or aspects of reforms), this study investigates the institutional framework in which such reforms occur, and assesses the interrelationship of systems and reforms at the Australian national and sub-national levels.

Considerable attention has already been given to the comparative, quantitative study of voting system outcomes, in terms of the methods of voting (for example, majoritarian versus proportional systems). Consequently, this thesis focuses primarily on electoral system regulation, management and administration. It assesses those aspects of electoral regulation and management that cumulatively may have major effects on equity and access, and considers these against international norms.

34

35

Chapter 3

Methodology

Research Design

This research has two primary objectives. The first objective is to present an evaluation, against internationally accepted norms for electoral management and fair elections, of significant electoral reforms that have occurred in Australia in the past 25 years. The second is to provide insight into the forces at play in such reform processes. In applying a normative framework to Australian electoral reforms the thesis adopts a comparative approach. As explained in Chapter 1, the comparison focuses on the electoral reforms introduced in Australia's nine jurisdictions over the 25-year period from 1983 to 2007.

The justification for including all nine Australian jurisdictions in this comparative study is as follows: First, the differences between jurisdictions warrant attention. Although there are a few features shared in common by the nine systems, such as compulsory voting and party registration, no two systems can be regarded as being the same. For almost every aspect of electoral administration and management, there is at least one jurisdiction that does things differently. Second, there are relatively strong relationships between the jurisdictions at the administrative level, in terms of the exchange of information, contractual agreements (for example, bilateral joint roll agreements between the Commonwealth and each of the States and Territories) and personnel crossover. There are also similarities at the legislative level, such as party registration and donation disclosure schemes, and at the political level, with Labor and the Coalition

36 parties being dominant throughout. The mix of diversity and commonality supports an inclusive and comprehensive study.

It is beyond the scope and resources of this research to give equal coverage to all nine jurisdictions on all possible issues. Therefore, the following chapters identify issues in electoral systems that provide evidence relevant to the hypotheses discussed in Chapter 1. Using the case study approach, jurisdictions are selected that best typify the electoral issue being considered. Through this process, insight is gained into the reform processes in play in all jurisdictions, and the impact such reforms have on fairness, equity and access, measured against internationally accepted standards for electoral management.

It was necessary to be selective in deciding which reforms to assess in detail, in order to pursue the comparative case study approach discussed above. Including all nine Australian jurisdictions imposes the need to limit the depth of analysis for each reform or aspect of the system. The focus is on significant reforms, particularly those that are distinctive to specific jurisdictions.

Assessment Frameworks

Three existing approaches to assessing democracy and fairness were considered in undertaking this study of Australian electoral systems: (1) a discussion-based model, generally used to inform and enrich debates about reform within the democracy being assessed; (2) comparative models, which identify differences using various indicators for assessment; and (3) ratings-based models, where a democracy 'score' is calculated, which is then often presented to international donors or financial institutions (Beetham et al., 2001). The present study draws on a combination of these three forms of assessment that have been devised for the comparative study of electoral systems and democratic governance. The following sections briefly describe the format, criteria and indicators that these frameworks typically use, before explaining how the frameworks will be applied to this research project.

37

Discussion-based model –– Democratic Audit Framework

The first form of assessment to be considered is the IDEA’s Democratic Audit framework, now also known as the ‘State of Democracy’ framework. The Democratic Audit framework enables the identification of strengths, weaknesses and sources of reform in individual dimensions of democracy and does not attempt to aggregate these factors. The audit framework seeks to inform research and debate in four ways: raising public consciousness about democratic values and how they relate to institutional arrangements; focusing on strengths and weaknesses of democratic practices; contributing to public debate on reform initiatives; and providing an assessment instrument to determine the impact of reforms. The assessment framework developed through IDEA is based on the two core democratic principles of popular control over public decision making and decision makers, and political equality of respect and voice between citizens in the exercise of that control (Beetham et al., 2001). It can be argued that when these two principles are in conflict — that is, where popular majorities threaten the equality of respect and voice for citizens –– it is the political equality value that must always take precedence over the value of popular control (Sawer, 2006a).

Table 3.1 Democratic Audit Assessment Framework Pillar of Section Overriding question democracy Citizenship, law Nationhood and Is there public agreement on a common citizenship and rights citizenship without discrimination? Rule of law and access Are state and society consistently subject to the to justice law? Civil and political rights Are civil and political rights equally guaranteed for all? Economic and social Are economic and social rights equally guaranteed rights for all? Representative Free and fair elections Do elections give the people control over and accountable governments and their policies? government The democratic role of Does the party system assist the working of political parties democracy? Effective and Is government effective in serving the public and responsive government responsive to its concerns? The democratic Does the parliament or legislature contribute effectiveness of effectively to the democratic process? parliament Civilian control of the Are the military and police forces under civilian military and police control? Integrity in public life Is the integrity of conduct in public life assured? Civil society and The media in a Do the media operate in a way that sustains popular democratic society democratic values? participation Political participation Is there full citizen participation in public life? Decentralization Are decisions taken at the level of government that is most appropriate for the people affected?

38

Democracy External influences on Is the impact of external influences broadly beyond the state the country’s supportive of the country’s democracy? democracy The country’s Do the country’s international policies contribute to democratic impact strengthening global democracy? abroad Source: Beetham et al., 2008: 73–8.

Of relevance to this study are the audit framework’s sections on free and fair elections (questions 1–6) and the democratic role of political parties (question 1), as listed in Table 3.2. IDEA’s guide to making audit assessments (Beetham et al., 2008) goes on to provide positive and negative indicators for each of these questions. It can be seen from these questions (and IDEA’s guidelines) that there is an emphasis on accessibility, equality and fairness. This emphasis is consistent with the qualitative and quantitative concepts of fairness discussed in the previous chapter.

Table 3.2 Democratic Audit Assessment Questions Free and fair elections 2.1.1 How far is appointment to governmental and legislative office determined by popular competitive election, and how frequently do elections lead to change in the governing parties or personnel? 2.1.2 How inclusive and accessible for all citizens are the registration and voting procedures, how independent are they of government and party control, and how free from intimidation and abuse? 2.1.3 How fair are the procedures for the registration of candidates and parties, and how far is there fair access for them to the media and other means of communication with the voters? 2.1.4 How effective a range of choice does the electoral and party system allow the voters, how equally do their votes count, and how closely does the composition of the legislature and the selection of the executive reflect the choices they make? 2.1.5 How far does the legislature reflect the social composition of the electorate? 2.1.6 What proportion of the electorate votes, and how far are the election results accepted by all political forces in the country and outside? The democratic role of political parties 2.2.1 How freely are parties able to form and recruit members, engage with the public and campaign for office? Source: Beetham et al., 2008: 74–5.

Comparative models

The second form of assessment uses traditional comparative methodologies. Two models that focus on electoral systems are those of Mozaffar and Schedler (2002), and Massicotte, Blais and Yoshinaka (2004). Mozaffar and Schedler use a three-level approach to the comparative study of electoral governance. The levels are: rule making (electoral competition and electoral governance); rule application; and rule adjudication

39

(see Table 3.3). Their design is mainly intended, as is much of the democracy assessment literature, for use in the assessment of developing democracies. Their model has an emphasis on the explanation of democratic progress, rather than the evaluation of electoral systems at particular points in time (Mozaffar and Schedler, 2002: 12).

Table 3.3 Three Levels of Electoral Governance Level Elements 1a. Rule making –– Formula –– District magnitude –– Boundaries –– Assembly size –– electoral competition Timetable –– Franchise 1b. Rule making –– Voter registration –– Party and candidate registration –– campaign finance electoral governance –– Observation –– Ballot design –– Polling stations –– Voting and counting –– Election management bodies –– Dispute resolution 2. Rule application Registration of voters, candidates and parties –– Election observers –– Education –– Organisation –– Voting and counting 3. Rule adjudication Admission of complaints –– Processing –– Publication and implementation Source: Mozaffar and Schedler, 2002: 8.

In their study of 63 countries, Massicotte, Blais and Yoshinaka (2004) also focus their comparative assessment on a number of themes: voting restrictions; candidacy; voter registration; electoral administration; and polling day (see Table 3.4). However, the emphasis of this comparative model is on the tabulated assessment of all democracies (not just developing democracies), which is then analysed by distinguishing between developed and developing democracies. The model also emphasises the influence of colonial heritage, for example by grouping British Commonwealth democracies as a subset of the whole.

Table 3.4 Establishing the Rules of the Game Theme Elements Voting restrictions Age –– Mental capacity –– Citizenship –– Period of residence –– Expatriates –– Prisoners Candidacy Citizenship and residence –– Deposit –– Public service ineligibility –– Compatibility with other elective mandates Voter registration Initiated by state –– First registration –– Compulsion –– Inquiries –– Close of register –– Party involvement Electoral Nature of authority –– Method of appointment –– Term of office administration Polling day Number of days –– weekday or rest day –– Employees given time off –– Opening hours –– days between election rounds –– polling officials –– Nomination procedures –– Scrutineering –– Identification requirements Source: Massicotte, Blais and Yoshinaka, 2004.

Several of the elements in the Massicotte, Blais and Yoshinaka model are generally not seen as contentious in the Australian context. As an example, the voting age in all

40

Australian jurisdictions is set at 18 years, and although there is some debate about lowering the voting age, the current age of eligibility is not viewed as a constraint on democracy for most Australians. Likewise, the rules in place for polling day opening hours and the use of scrutineering are also generally uniform and seen as non- controversial. Other elements, however, such as restrictions on prisoner voting and rules for voter registration (enrolment) have been highly contentious in Australia in recent years, with a variety of reforms taking place in several jurisdictions. The thesis therefore takes a selective approach to applying the elements listed in these comparative models.

Ratings-based model –– the Elklit and Reynolds Framework

The third form of assessment considered for use in the thesis is based on rankings systems, which provide aggregate 'scores' derived from indicators. Probably the best- known is the annual Freedom House ratings, which have been produced since the 1970s. The Freedom House ratings are based on seven-point scales for political and civil liberties, ranging from best practice to worst practice. In recent years, Jørgen Elklit and Andrew Reynolds (2002, 2005) have developed a methodology that is specifically designed for elections and electoral systems. By providing ‘scores’ for freeness, fairness, and the administrative efficacy of elections, Elklit and Reynolds seek to avoid the possible manipulation or ‘spin’ that is sometimes applied to election outcomes by election observers for the sake of political expediency (2005: 147).

Elklit and Reynolds (2005) attempt to simplify their task with an overarching question – – ‘[are] the rules, as written, applied fairly and without partisan bias?’ (2005: 151). They have developed 54 questions categorised into 11 themes, or steps, for their framework assessment. Table 3.5 outlines the themes. The framework has been designed to include issues that are highly important for developing democracies. Consequently, not all of the questions are pertinent for the Australian situation. In many cases, the issue has already been dealt with in a way that is acceptable to the vast majority of stakeholders. For example, the legal framework for elections in Australia is clearly established and widely accepted (questions 1.1–5); violence and direct vote buying are generally absent from Australian elections (questions 8.3 and 8.9); and vote counting procedures tend not to be controversial (questions 9.1–5), except on some occasions when the results are particularly close. Other issues, however, such as the

41 impartiality of electoral management, informal voting, fairness, and voter turnout, are highly relevant to the Australian situation, and therefore are incorporated in this study.

Table 3.5 Election Assessment Framework Theme Issues Considered Form of Measurement Legal Framework Legal foundation –– Electoral timetable –– Expert panel assessments Timeliness –– Implementation – Legitimacy

Electoral Legitimacy –– Impartiality –– Quality –– Polling evidence management Transparency Expert panel Survey of stakeholders Constituency and Accepted structure –– Access to information –– Expert panel polling district Fair limitation and allocation systems Stakeholder surveys demarcation Voter education Access to education –– Informal voting –– Needs Surveys of voter of ‘at-risk’ groups –– Turnout of first-time voters education outreach Data from register, polling, and election results ‘At-risk’ defined as historically marginalised groups Voter registration Proportion registered –– Bias –– Convenience –– Data from register Accuracy –– Confidence –– Fairness Expert panel analysis Ballot paper Party registration –– Independents –– Expert panel assessments design Discrimination Campaign Access to media –– Public funding –– Media bias Expert panel assessments regulation –– Use of state resources Polling Turnout –– Violence –– Conforming to rules –– Data from election results Control of fraud –– Accessibility –– Observation – and observer reports – Vote buying –– Intimidation Expert panel assessments based on data Counting Integrity and Accuracy –– Transparency –– Expert panel assessments Access –– Timeliness –– Observation based on data from observer reports Result Complaint resolution –– Timeliness –– Expert panel assessments verification Legitimacy –– Access to justice Reports Legislation Post-election Access to statistics –– Audit of electoral Expert panel assessments management –– Election review Source: Elklit and Reynolds, 2005: 152–4.

The three assessment frameworks –– Democratic Audit, comparative, and ratings –– all seek to inform debate on the strengths and weaknesses of electoral systems. However, as discussed, each of the frameworks includes several aspects that are not relevant for this study. Consequently, a combination of the three frameworks is used to identify and explore features that support or undermine the freedom and fairness of Australian elections.

42

Scope of the Study

There is a degree of overlap between the comparative frameworks reviewed above, and together they are useful in providing general guidance for the comparative case study assessment of electoral management. For the purpose of the current study, however, the aim is to evaluate the impact of recent electoral reforms, and to provide a greater understanding of how and why reforms occur. Importantly, Elklit and Reynolds argue that the design of any assessment framework needs to be cognisant of the particular conditions of the jurisdictions involved (2005: 151). Consequently, the present study developed a framework drawing on aspects of those reviewed, first, to better reflect the Australian electoral environment, and second, to improve on existing international models by generating a framework that is more applicable to developed democracies.

The assessment framework used in this thesis has been designed to reflect two principal themes: the administrative structure of Australian electoral management bodies; and the electoral reform process, including the impact that reforms have on election participants –– that is, voters, candidates and political parties. The first theme focuses on the independence of Australia’s electoral commissioners. In evaluating the independence of commissioners, all three of the assessment approaches outlined above are drawn upon. Question 2.1.2 of the Democratic Audit discussion-based framework (Table 3.2) is particularly relevant to this theme –– How inclusive and accessible for all citizens are the registration and voting procedures, how independent are they of government and party control, and how free from intimidation and abuse? In addition, the analysis utilises Mozaffar and Schedler’s (2002) and Massicotte, Blais and Yoshinaka’s (2004) methodologies for comparative assessment of electoral administration (Tables 3.3 and 3.4). Finally, a modified version of Elklit and Reynolds’ (2005) ratings-based framework was developed to differentiate the levels of independence enjoyed by the various commissions. A comprehensive comparative assessment was supported by interviews with the commissioners for all nine jurisdictions.

The second theme –– the electoral reform process –– largely uses a discussion-based approach, with comparative assessments also made where relevant. The modified ratings-based methodology is not a suitable methodology for this component of the

43 study, for two main reasons. First, because a case study approach is adopted, not all aspects of electoral law were considered across all jurisdictions. Second, because some of the key stakeholders, such as ministers and shadow ministers, were not available for interview, it was difficult to make a comprehensive assessment across all jurisdictions.

There are several electoral matters that had to be excluded from this study, given the demands of covering nine jurisdictions. The issues to be addressed were largely identified through the research process. Some issues, however, were found to be beyond the scope of the study at an early stage in the design of the research framework. These included appeals mechanisms for disputed returns and the processes for conducting boundary redistributions. While these areas are clearly relevant to the election process, they are not included in this study. Appeals mechanisms are regarded as primarily a legal issue, and are well addressed in the academic literature, for example, by Graeme Orr and George Williams (2001), whose study includes a cross- jurisdictional comparative component.

The boundary redistribution process was significantly changed at the Commonwealth level in 1983, when the Hawke Labor Government removed parliament’s veto power over the determination of boundaries for House of Representatives’ electorates. The process since 1983 is overseen by an independent committee appointed by the AEC, and includes three rounds of public consultation. Final decisions can only be challenged on the basis of an error of law, and not on any political considerations. Similar processes are in place at the State and Territory levels. As a result, the Australian system is considered best practice by many, with American political scientist Richard Engstrom stating that it is ‘essentially a quasi-judicial, or even bureaucratic, task that is performed without partisan referents’ (2005: 2). By comparison, the United States system is highly politicised and heavily litigated. The Hawke Government’s 1983 reform defused redistributions as a political issue, and as a result, it is now seen as a largely non- contentious area, and therefore not covered in this study. Recent academic works on redistribution processes include Colin Hughes’ (2007) analysis of redistributions in New South Wales and Queensland, and Glynn Evans’ (2005) assessment of the ‘fairness’ clause in South Australia.12

12 In addition to political science approaches such as Evans (2005) and Hughes (2007), the study of redistributions also benefits from interdisciplinary research incorporating geographic and demographic modeling methods.

44

Another area not explored in great depth in this research is the resourcing and effectiveness of voter education. All Australian electoral management bodies regard voter education as an important issue, but neither administrators nor political stakeholders consider it to be a controversial area. Nevertheless, two relatively contentious aspects of voter education are addressed in this thesis: education for Indigenous voters; and education to explain differences between the Federal and State systems, especially in regard to optional preferential or full preferential voting.

The indicators that are central to this study are derived from two principles of electoral system fairness, as discussed in Chapter 2. These are participation, or equality of opportunity to participate in elections, and knowledge, both of the electoral process, and in terms of being able to make an informed choice. The equality principle demands that electoral administration is independent and non-partisan, so that elections are conducted without bias to particular candidates or parties. The equality principle also requires that registration and voting procedures are easily accessible to all citizens, that individuals’ votes have equal weight and importance, and that these votes are accurately reflected in parliamentary representation.

Table 3.6 Indicators for Comparative Analysis of Australian Electoral Systems Themes Indicators Parliamentary Representativeness of committee membership oversight committees Input from extra-parliamentary parties Engagement with citizens and civil society Electoral Level of independence administration Capacity to amend electoral regulations Control over budget Accountability Powers Appointment procedures Franchise Inclusiveness Levels of participation Voters Ease of enrolment Accuracy of the roll Ease of voting Accessibility Levels of informality Party registration Barriers to entry Membership threshold Candidates Ability to communicate with voters Funding Regulation of private funding Access to public funding Election outcomes Proportionality of results

45

The knowledge principle emphasises the need for election participants to be aware of and understand electoral laws, including how to make a formal vote, and, in cases of ticket voting, knowing what happens to a vote if the preferred choice is eliminated. This principle also requires voters to be able to make an informed choice by being aware of who the candidates and parties are, and what policies they stand for. Table 3.6 outlines the themes and indicators that form the basis for this research. These themes and indicators, drawn from the assessment methodologies discussed above, underpin the questions asked of the interviewees. The interviews also identified specific issues that required further case study and comparative analysis.

The first theme, parliamentary oversight committees, is dealt with in the following chapter, while electoral administration is the subject of Chapter 5. The remaining themes –– the franchise, voters, party registration, candidates, funding, and election outcomes –– are explored using a comparative case study approach in Chapters 6 to 8.

Data Collection

Data was drawn from four main sources for this thesis. First, an assessment was made of the existing secondary literature, as reviewed in Chapter 2. A more detailed analysis of the relevant literature is provided in the chapters that follow. Second, a comparative review was conducted of the existing electoral legislation and electoral commission reports in each jurisdiction, including legislative amendments that have occurred during the 25-year study period. From this, ‘best practice’ among the various jurisdictions was identified. The results of the comparative review were then combined with analysis of the third main data source –– parliamentary inquiries and debates (including Hansards and committee inquiry submissions and reports). Together, these three data sources were used to generate an understanding of how legislative reforms occur. The process of comparative assessment includes an initial desk-top analysis of electoral reforms, from which a number of major electoral issues and reforms were selected for intensive analysis. These issues and reforms were further explored through quantitative assessment and face-to-face interviews (interviews being the fourth main data source).

46

Recent reforms were preferred for the in-depth analyses, because the participants involved in negotiating these reforms were more readily available for interviews.

The data collected from the personal interviews were considered to be critical to the overall research questions for this study. In particular, interview data were critical in answering the question: What are the drivers for reform –– that is, how and why do electoral reforms occur? Issues identified in the general assessment of reforms were used to construct unique semi-structured guides for the interviews. The interviews were conducted with the key stakeholders in each jurisdiction: electoral commissioners and senior officials; ministers; shadow ministers; and parliamentary committee members. The interviews provided insight into legislative reform processes, the operation of electoral administrations, and the relationships between the electoral institutions of the parliament, executive and the administration. The interviews were important not only in illuminating the reasons for the reforms, but also in exposing the processes involved and the role that personal relationships and interactions play in facilitating or thwarting reforms.

Figure 3.1 Research and Reporting Model

Research Framework

Data Collection

Exploration of Issues

Description

Analysis

Interpretation and Findings

Sources: Creswell, 1998; Stake, 1995.

47

The research followed a seven-stage case study reporting model, as outlined by Creswell (1998) and Stake (1995), and depicted in Figure 3.1. As the model emphasises, data collection (particularly interviews, for this study) and data analysis provide important ways to identify issues for more selective and detailed analysis. This iterative process of data collection and analysis ensured that the critical aspects of the research focus were identified and then given due analytical treatment.

Analysis of Data

Data are analysed in the following chapters, with the normative assessment being based on quantitative, institutional and other indicators. The three assessment models are applied to varying degrees depending on the subject matter. A scoring scale is used in Chapter 5 to rate the nine jurisdictions, and values are assigned using a modified version of Elklit and Reynolds’ framework for assessing fairness (2005: 154–6). This framework was selected as being relevant for the current study because it represents the best attempt in contemporary literature to assign values to the various aspects of electoral administration. Other comparative studies, such as that of Massicotte, Blais and Yoshinaka (2004), Bogdanor and Butler (1983), and Reynolds and Reilly (2002), are useful in identifying and explaining differences in voting systems and electoral regulation, but do not attempt to quantify the importance of such differences. For the present study, Elklit and Reynold’s framework was modified to suit research on electoral administration in established democracies. While the factors assessed here differ from that of the Elklit and Reynolds framework, because the latter was designed to assess both established and developing democracies, the Elklit and Reynolds method of scale and standardisation is an appropriate research tool for this study.

Figure 3.2 shows the rating scale used by Elklit and Reynolds in their work. As can be seen, they provide a four-point scoring scale based on the broad range of proficiencies expected in established and developing democracies, from very good to poor (2005: 154). In addition, they provide a weighting based on the relative importance of a category depending on whether a developing or developed democracy is being assessed.

48

Figure 3.2 Elklit and Reynolds Framework –– Scoring Methodology (2005)

54 indicators under 11 categories are each given a score 3 – very good 2 – good 1 – not satisfactory 0 – poor

The sum of the indicator scores for each category is tallied and standardised relative to the value of 10

Each category is given a weighting based on its importance to developed or developing democracies (for example, voter education is important in developed democracies, but essential in developing democracies) x3 – essential x2 – important x1 - desirable

The sum of the standardised and weighted category scores are tallied and standardised relative to the value of 100

The Elklit and Reynolds framework is modified for this study to reflect the narrower range of proficiencies generally expected of developed democracies, and particularly in Australia. To allow for differentiation between jurisdictions, a five-point scale is used to provide a sufficient number of gradations between best practice and worst practice. In addition, as the subjects of this study are all developed democracies, the weightings based on importance were not considered to be relevant. It is anticipated that this modified framework will be applicable to developed democracy comparisons world- wide. Figure 3.3 illustrates these changes.

49

Figure 3.3 Kelly Framework –– Modified Elklit and Reynolds

Indicators under eight themes are each given a score 5 – best practice 4 – 3 – mid-point 2 – 1 – worst practice

The sum of the indicator scores for each theme are tallied and standardised relative to the value of 10

The sum of the standardised theme scores are tallied and standardised relative to the value of 100

Similar to Elklit and Reynolds’ framework, indicators in the modified framework are grouped into themes (categories) and the scores are standardised. There is always a subjective element in such normative assessment and it is also true that some indicators may have greater importance than others within the same theme when assessing the fairness of electoral systems. For the purpose of the analysis in this study, however, weightings are not given to individual indicators. As Elklit and Reynolds point out, it would be erroneous to view the application of such scales as a perfect method for assessing the conduct of elections (2005: 149). By using the scale of best practice to worst practice, such subjective evaluations provide gradings rather than outright values. However, the gradings produce consistent final scores that can be used for comparative purposes. For this study, the modified framework is tested against the electoral administration theme.

Personal Interviews

The comparative data is enriched by personal interviews conducted with the commissioners from all jurisdictions, and many of the relevant ministers and shadow ministers. The interviews provide snapshots of the views and opinions of key actors in

50 electoral governance, giving an insight into reform processes and providing data which will be useful for future research. The absence of any similar comparative study, and the paucity of comparative research at the sub-national level, means this interview data is covering new ground, and will be a valuable addition to political science literature.

This study uses a semi-structured interview format, which is a combination of the general interview guide and the informal conversational interview techniques. Topics are pre-determined, but the interviewer is not required to follow a particular order of questioning (Selltiz, Wrightsman and Cook, 1976: 318). This allows the interviewer to be responsive to the answers being provided. In the informal conversational, or recursive interview, questions are not pre-determined. Rather, questions and topics emerge through the course of the interview (Minichiello, et l., 1992: 112). The semi- structured interview format therefore allows specific questions and core themes to be covered, but also provides the freedom to pursue additional topics as raised by the interview participants.

There are a number of reasons for adopting the semi-structured interview approach. The interviews were in part designed to provide insights into the relationships between the various participants involved in developing and administering electoral systems. These participants were either political representatives, who may have partisan positions to promote, or electoral commission officials, who generally took a more apolitical approach in line with their position as independent statutory officers. A list of interview questions was prepared prior to each interview and structured specifically for each individual interviewee. This approach was adopted because it was the purpose of each interview to seek a unique mix of data based on the particular history and electoral system in each participant’s respective jurisdiction. In addition, the semi-structured format enabled lines of investigation to be pursued on issues that were raised by the participants in the course of the interview. In some cases, this approach was beneficial in raising new themes and topics not originally anticipated in the research design.

The data sought in the interviews included matters that were politically sensitive and, in the case of electoral officials, were often also personally sensitive as they touched on the manner of their appointments. It was decided that conducting the interviews using an informal conversational approach would increase the probability that the participants would divulge information. In the majority of cases, interviews were conducted in the

51 participants’ offices, as, according to Smith (1996), interviewing participants in familiar surroundings enhances the reliability of data collected (1996: 7). To encourage more open and frank responses, the interview subjects were assured of confidentiality in the way their transcripts were to be used. In addition to obtaining consent for the use of the data prior to each interview, it was explained to each participant that when the use of a direct quote could identify them, they would be contacted for permission to use that quote. While some participants waived this requirement for follow-up permission, this assurance meant participants were more open in their responses than might otherwise have been the case. It has also been possible to draw extensively on the interview transcripts to illustrate particular issues without needing to identify the source of the quotes.

The interview participants were selected according to their positions in the respective jurisdictions. For each jurisdiction, the minister and shadow minister responsible for electoral matters were identified. In addition, in those jurisdictions where a parliamentary oversight committee existed (the Commonwealth, NSW, and Queensland), members of these committees were interviewed. A number of other people were also identified as key sources, to broaden and diversify the range of knowledge and views sought on how electoral reforms develop and are implemented. Such people included representatives of minor parties with portfolio responsibility for electoral matters and electoral officials below the Commissioner level who are involved in legislation and parliamentary relations.

In total, 31 interviews were conducted, including all full-time electoral commissioners (that is, the chief executive officers of their organisations). Some ministers and shadow ministers were not available or willing to be interviewed; it was possible, however, to interview the minister and corresponding shadow minister in five of the nine jurisdictions. To compensate for the relative lack of interviews in the remaining four jurisdictions, extra effort was made to research parliamentary debates and media statements. Interview participants are listed at Appendix A.

52

The Role of the Researcher

It is widely acknowledged that data collection and interpretation occurs through the filter of the researcher’s experiences and background (Rossman and Rallis, 1998: 68). Therefore, researchers need to be rigorous in their approach to ensure that personal perspectives, including bias, opinions and prejudices, do not corrupt the quality of data obtained. Applied research in the social sciences often deals with complex and conflict- ridden situations, in which there are competing values, perspectives and ideologies (Denzin, 1970: 5). This is especially true of research into matters of public policy.

Research is particularly sensitive when it has the potential to threaten or put at risk those being studied. Research that impinges on the vested political interests of people and institutions can be included in this category (Lee, 1993: 4). Such sensitivities are a matter of concern for the current research. The interview participants in this study were generally experienced in being interviewed, both by researchers and the media, and were therefore aware of the implications in revealing sensitive information. It was my aim to gather sensitive data, so that a thorough analysis of policy processes and decision-making could be undertaken. Thus, there was a need to treat the information gathered in a confidential manner.

My previous background assisted in understanding sensitive contextual issues. I brought to the study the experience of having been a member of the Western Australian Legislative Council for four years, representing the Australian Democrats from 1997 to 2001. I resigned from the party in 2003 and have not been politically active since that time. In addition, I conducted a series of interviews with political representatives as part of my honours research (Kelly, 2004). While this combination of experiences allowed me to understand some of the contextual issues for this doctoral research, the interviewees did not consider me to be an active political protagonist who would seek political advantage through gaining information on sensitive issues. The combination of my personal experience and role as a researcher assisted in establishing a comfortable interview environment and rapport with the interview participants. My own political experience also assisted in understanding the procedures, strategies and nuances of the various electoral reform debates.

53

Summary

This thesis aims to advance the normative study of electoral systems, through the comparative assessment of electoral systems management across Australian jurisdictions, and of the various aspects of these electoral systems that impact on equity and fairness for election participants. The thesis develops a framework, incorporating discussion-based, comparative, and ratings-based methodologies, for assessing electoral systems in developed democracies, in relation to international norms for electoral management and fair elections. This framework is used to provide a comprehensive comparative analysis of Australian electoral systems at the Federal, State, and Territory levels. The modified Elklit and Reynolds assessment framework is used for comparative analysis in Chapter 5, and this modified framework is ideally suited to further comparative studies in both the Australian context, and in other developed democracies.

This research also assesses the impacts of reforms over the past 25 years in terms of access and equity, across a range of indicators drawn from internationally accepted standards for the conduct of free and fair elections. By providing insight into reform processes and the drivers behind reform, this study also contributes to an understanding of the dynamics of legislative change and the sometimes competing objectives of governments, oppositions, and electoral administrators.

54

55

Chapter 4

Institutional Structure

The strength of the committee, and with most committee structures, is not what’s in the report, but in the process. You normally get a situation where you get a good spread of witnesses. You get a good spread of opportunities for scrutiny and debate around the circumstances of what’s occurred in an election.

People should have no doubt that it is the most political of committees, given the nature of what it’s looking at, and so its recommendations are invariably, often political…invariably you end up with dissenting reports, practically more than any other committee that I’m aware of. – an Australian politician

The institutional structure of an electoral system provides the environment in which electoral laws are administered and amended. An understanding of this environment is necessary to evaluate the fairness of a system, in terms of both democratic behaviour and electoral outcomes. Australia’s electoral management bodies have an international reputation for their professional, non-partisan and independent performance. Yet how strongly entrenched is that integrity and independence? Is the reputation deserved? Australian electoral administrations are relatively similar in terms of structure and responsibility; however, their powers vary depending on their enacting legislation. In most cases the legislation also limits the ability of commissions to act independently in the interests of best electoral practice and in providing a level playing field for election participants. In addition to their regular interactions with the government of the day, as in negotiating budgets and liaising with the responsible minister, electoral commissions

56 are subject to different levels of parliamentary oversight, in terms of inquiries into the conduct of elections and other electoral matters.

This chapter presents an assessment of the institutional framework of Australian electoral administration, beginning with a description of the electoral administration institutions and parliamentary oversight committees. The chapter also considers the relationships that exist between commissions and the parliamentary committees, as well as between parliamentary committees and citizens and civil society organisations. On this basis, an assessment is made of the fairness of the institutional structure, particularly in terms of the representativeness of parliamentary committees and engagement with citizens and civil society. Chapter 5 continues the theme of examining electoral administration by focusing on the performance of electoral commissions, with an emphasis on the degrees of independence they exhibit.

Australian Electoral Management Bodies

The origins of Australia’s professionalised and independent electoral administration lie in 19 th century colonial society, with a lack of strong local authorities or precedent, which enabled professional, efficient and non-partisan electoral systems to develop (Sawer, 2001a: 15). As Sawer describes, administration was initially the responsibility of public servants working in a government department, with greater independence later being achieved through the use of statutory bodies, or ‘offices’, and finally with the removal of ministerial direction and the establishment of commissions.

At the Federal level, the Australian Electoral Commission (AEC) was established in 1984 by section 4 of the Commonwealth Electoral Legislation Amendment Act 1983 . The AEC replaced the Australian Electoral Office, a statutory authority founded in 1973. As Hughes (2001) notes, although the 1983 legislation effectively removed the Minister from the principal Act for all matters except the tabling of reports in parliament, the government retained control over budgetary and legislative matters (2001: 156). At the State and Territory levels, all electoral administrations, apart from South Australia, are commissions. Although the South Australian administration currently remains an ‘Office’, it is planned for it to become a Commission in the near

57 future, and has had an Electoral Commissioner as the Chief Executive Officer of the Office since 1973 (Macilwain, 2007: 20). In this thesis, references to all of the Australian electoral management bodies will use the term ‘commission’ in a generic sense.

There are strong relationships between electoral administrations at the Federal, State and Territory levels, and innovations in the management of electoral systems are regularly disseminated between the jurisdictions. The commissions regularly assist each other in the conduct of elections by, for example, providing pre-polling facilities, specialist staff, and exchanging information on technological advances, such as electronic voting. The names of the nine Australian administrations, and the years the administrations were established as Commissions, are provided in Table 4.1

Table 4.1 Australia’s Electoral Management Bodies Jurisdiction Electoral management body Commission established Commonwealth Australian Electoral Commission (AEC) 1984 NSW New South Wales Electoral Commission (NSWEC) 2006 Victoria Victorian Electoral Commission (VEC) 1995 Queensland Electoral Commission of Queensland (ECQ) 1992 Western Australia Western Australian Electoral Commission (WAEC) 1987 South Australia State Electoral Office (SEO) – Tasmania Tasmanian Electoral Commission (TEC) 2005 ACT Australian Capital Territory Electoral Commission 1992 (ACTEC) Northern Territory Northern Territory Electoral Commission (NTEC) 2004

NSW Election Funding Authority

Typically, the commission is the sole institution with responsibility for the conduct and administration of elections in their jurisdiction (except for the drawing of electoral boundaries, which is determined by a separate statutory body administered by the commission –– see Economou, 2007). However, whereas other jurisdictions entrust their commissions with the task of administering party registration, public funding, and other political finance matters such as financial disclosure, in 1981 New South Wales established a separate authority under the Election Funding Act 1981 for these purposes. This is the three-member NSW Election Funding Authority (the Authority). While the

58 establishment of a separate agency for political finance purposes is not in itself a problem in terms of fairness, the membership of the Authority is problematic.

Briefly, one of the Authority’s responsibilities is the administration of public funding to parties and candidates, including the provision of public funds for ‘electoral education’. NSW is unique in Australia in providing funds to the parties for the production and dissemination of information about policies and party-related work. This expenditure can be used exclusively for providing services to party members, if the party chooses. Funding is calculated on the results at the previous general election, with six of the 16 parties that contested the 2007 election receiving funding based on their 2003 election results (EFA, 2007: 23).

The Electoral Commissioner sits as the Chair of the Authority, with the other two members nominated by the Premier and the Leader of the Opposition. The current members are Steven Lewis, nominated by the previous Labor Premier Morris Iemma, and Ted Pickering, a former Liberal parliamentarian and Minister, nominated by the Opposition Leader, Barry O’Farrell. The ability of the two major parties to nominate members of the Authority makes New South Wales the only Australian jurisdiction in which appointments to an electoral management body are made on a partisan basis. This appointment process impacts on perceptions of the independence of the NSW Electoral Commissioner, who is required to preside over this authority and yet is outnumbered by political appointees.

The Authority has the power to initiate legal proceedings against candidates who do not comply with the requirements of the Act. This places the Authority’s two nominated members in a position of potentially starting action against their own party colleagues – – a clear conflict of interest. It can be argued that having an independent Chair holding the casting vote may prevent blatantly partisan decisions that favour one party over the other. However, there is also the capacity for the two party appointees to act in collusion, as a cartel operating against fairness principles and against the interests of other parties and election candidates. For example, the two party appointees could support each other to avoid actions against their party colleagues. The Authority’s 2006–07 annual report lists (mostly local government) candidates against whom proceedings were initiated. Of 47 separate actions, 43 were against independent

59 candidates, with the remaining four candidates representing minor parties (Unity Party and the Christian Democratic Party) (EFA, 2007: 13–15).

The fact that no candidates from the appointees’ parties were subject to proceedings against them may simply indicate the report’s focus on the local government level, at which party campaigning is less overt, or that the parties’ candidates had all complied with the law. However, the perception of a lack of independence remains, due to the presence of political appointees. In 2008, the NSW Select Committee on Electoral and Political Party Funding (SCEPPF), which included Labor and Liberal party representatives, recommended that partisan appointments should cease, to remove any perception of bias (SCEPPF, 2008a: 213).

It would appear to be a simple exercise to incorporate the functions of the Authority with those of the Electoral Commission, as in other jurisdictions. At an administrative level this would be a seamless transition, as the Authority is essentially subsumed by the Commission in any case –– the Authority does not employ any staff, with the Commission providing all staffing and administrative resources. The fact that the Corporate Plan for the Commission and the Authority is a joint document is another indicator of the symbiotic nature of the relationship. The closeness between the two institutions is highlighted by a comment from the Commissioner: it was very unclear to me, when I came into the position of Electoral Commissioner, who was running and administering the Election Funding Authority provisions (SCEPPF, 2008b: 21). However, the Commissioner has expressed opposition to the Commission taking on the Authority’s responsibilities: …whilst the NSW Electoral Commission is a body corporate it has one member only and that is the Electoral Commissioner. I do not think it would be in the public interest to have one person effectively dealing with all of this. (SCEPPF, 2008b: 10)

However, one official stated in an interview that it would not be a major problem to integrate the Authority’s functions with those of the Commission, at the same time overcoming the Commissioner’s concern about having one person responsible for funding and disclosure issues: One of the things that in time might eventually happen is the Election Funding Authority would get subsumed into the actual Commission, with probably some sort of similar model to what Queensland has, where you have an augmented

60

Commission when you’re dealing with the funding stuff. So you could have a nominee of the Auditor General and a nominee from somewhere else.

As the SCEPPF found, there is merit and support for removing partisan appointments, based on internationally accepted norms for electoral management and concepts of independence (for example, see Wall et al., 2006: 9, discussed more fully in Chapter 5). At present, neither the Labor nor Liberal leadership has expressed a willingness to give up their right to nominate members of the Authority. The possible integration of the Authority with the Commission is less contentious, because in its current form the Authority is essentially administered by the Commission. Despite the partisan appointments, no allegations of corruption or partisan bias have been made, but the concern remains that the principle of fairness is diminished by the perception of partisan behaviour.

Parliamentary Oversight of Electoral Matters

Parliaments are made up of election winners. As the beneficiaries of the electoral system, parliamentarians have a clear vested interest in maintaining (or changing to their own advantage) a system from which they have benefited. Therefore, their parliamentary activity requires close scrutiny to determine whether their actions are based on principles of fairness, or on their own self-interest and partisan considerations.

Table 4.2 Parliamentary Oversight Committees Jurisdiction Committee Committee name established Commonwealth 1983 Joint Standing Committee on Electoral Matters (JSCEM, formerly the Joint Select Committee on Electoral Reform, 1983–87) NSW 2004 Joint Standing Committee on Electoral Matters (NSW JSCEM) Victoria 2007 Electoral Matters Committee Queensland 1990 Legal, Constitutional and Administrative Review Committee (LCARC, formerly Parliamentary Committee for Electoral and Administrative Review)

Four Australian jurisdictions have parliamentary oversight committees with a distinct role in overseeing and inquiring into electoral systems and the conduct of elections (see

61

Table 4.2). As the larger parliaments have more members to draw on for issue-specific committees, it is not surprising that the three largest jurisdictions (in terms of parliamentary size) have committees dedicated solely to electoral matters. These three jurisdictions are the Commonwealth, with 226 parliamentarians (150 MPs, 76 Senators); NSW, with 135 (93 MLAs, 42 MLCs); and Victoria, with 128 (88 MLAs, 40 MLCs). In Queensland (with a parliament of 89 members), the Legal, Constitutional and Administrative Review Committee (LCARC) examines electoral matters as part of its remit to inquire into broader issues of a legal and constitutional nature, not all of which are related to the electoral system.

The remaining jurisdictions, with parliaments ranging from 17 (ACT) to 95 (Western Australia) members, tend to use their legislative or other committees to conduct inquiries into electoral matters. The issue in question often determines to which committee such matters are referred to in these jurisdictions. For example, the ACT’s Standing Committee on Legal Affairs (SCLA) inquired into possible changes to the overall size of the Assembly and electorate magnitude in 2002 (SCLA, 2002), while in 2006–07, the Standing Committee on Education, Training and Young People (SCETYP) inquired into a proposal to lower the voting age (SCETYP, 2007). In Western Australia, the Standing Committee on Legislation inquired into the Gallop Labor Government’s ‘one vote, one value’ legislation in 2001 (Kelly, 2004: 56). The larger jurisdictions also send certain electoral matters to other standing committees. The following sections explain the inquiry activity of these committees, with a focus on partisan influences at work. Four standing committees with specific references for electoral matters are examined, as well as two examples of other committees that have received electoral matter referrals (a Commonwealth Senate standing committee and a NSW upper house select committee).

Commonwealth

The oldest Australian electoral committee is the Commonwealth’s Joint Standing Committee on Electoral Matters (JSCEM), formerly the Joint Select Committee on Electoral Reform (JSCER). The Commonwealth committee has developed a substantial history of conducting inquiries, with a total of 31 reports published in the past 26 years (a list of the committee’s reports is provided at Appendix B). The committee’s main focus is inquiring into the conduct of each general election. Additionally, the

62 committee conducts inquiries into specific electoral issues, such as levels of representation (1988 and 2003), the integrity of the electoral roll (2001 and 2002), and political finance issues (2006).

The depth of the committee’s election inquiries has grown in the 25 years since its inception. The growth in the number of submissions it receives after each election is evidence of this increasing depth, with 129 submissions for the 1987 election inquiry, increasing to 221 submissions following the 2004 election. As Hughes notes, the JSCEM election inquiries, which include extensive public hearings after initial submissions have been received, is possibly the best existing form of scrutiny into the electoral system, adding transparency that is lacking elsewhere (2001: 153–4). Being a joint committee, that is, drawing on members from both houses of parliament, means that there is some representation available for minor parties, with the Australian Democrats had a member on the committee from the initial formation of JSCER in 1983 through to 2008. From July 2008, Greens Senator Bob Brown is the minor party/independent representative.

In interviews, members of the committee expressed the view that the regular election inquiries are very important to the overall administration and conduct of elections. One member noted: You get a good spread of witnesses. You get a good spread of opportunities for scrutiny and debate around the circumstances of what’s occurred in an election. I’d also say that generally in respect of reports, you get issues pretty reasonably canvassed in the body of reports. The body of a report, for example, normally gives you a pretty good account of what’s occurred, the points of differentiation.

Concerns were raised, however, that the committee operates on an extremely partisan basis that is largely unavoidable, as in this observation from a committee member: People should have no doubt that it is the most political of committees, given the nature of what it’s looking at, and so its recommendations are invariably, often political. Invariably you end up with dissenting reports, practically more than any other committee that I’m aware of. The government of the day has a significant influence on the outcome of the reports, and you’d expect that.

An electoral administrator commented on the partisan nature of the committee, noting the difference in approach when the Howard Government had a Senate majority from 2005 to 2007:

63

Now it’s in an environment where the government has a majority in the Senate and that changes the dynamics greatly because there is less need in the joint standing committee for the politicians to come to a compromise. Because the government members feel they’ve got their senior colleagues behind them. They can go for broke.

It is understandable that there would be a close relationship between an electoral matters committee and the elections administrator. The JSCEM holds regular (twice yearly) private hearings with the AEC. Although these hearings are recorded by Hansard, the transcripts are not released to the public; they are regarded as briefings and an opportunity to discuss the mechanics of electoral administration away from public scrutiny. The AEC also provides substantial submissions to JSCEM inquiries, for example, more than 1,000 pages of information and comment were submitted following the 2004 and 2007 elections.13 Typically, the AEC’s submissions present data on the conduct of the election, as well as providing information and comment on specific electoral issues that have been raised through the media, or may have been foreshadowed or requested by the committee.

2004 Federal Election Inquiry –– the Richmond Result One example of the partisan nature of the JSCEM is contained in the inquiry into the 2004 Federal election. The issue was the outcome in the Richmond electorate, where the sitting member, National Minister Larry Anthony, was defeated by the Labor candidate by 301 votes, a margin of 0.19 per cent. Significantly in such a close contest, the Liberals for Forests candidate received 1,417 votes (1.8 per cent). The 10-member committee, with five government members, including the Liberal Party chair (who resolves any tied vote), decided to investigate this outcome in Richmond. The investigation was based on complaints from the National and Liberal parties on election day that the design of the Liberals for Forests how-to-vote cards, which gave preferences to Labor head of the Nationals, misled voters into thinking that Liberals for Forests were connected to the Coalition parties.14 The AEC provided advice to both

13 AEC submissions to JSCEM are publicly available and can be found on the AEC and JSCEM websites, at http://www.aec.gov.au/Elections/australian_electoral_system/jscem/index.htm and http://www.aph.gov.au/house/committee/em/reports.htm. For the 2004 election, see submissions 74, 165, 168, 172, 182, 205, and 221. For the 2007 election, see submission 169. 14 Section 329(1) of the Commonwealth Electoral Act 1918 states that ‘A person shall not, during the relevant period in relation to an election under this Act, print, publish or distribute, or cause, permit or authorize to be printed, published or distributed, any matter or thing that is likely to mislead or deceive an elector in relation to the casting of a vote’.

64

Coalition parties on election day, based on legal advice from the Government Solicitor, that a breach had not occurred.

The committee held a public hearing in the electorate, and used some of the evidence given to this hearing in its final report. There appeared to be a clear intention by Liberal members of the committee to gather evidence that the Liberals for Forests how-to-vote cards were deceptive. The leading nature of the questioning, and the subsequent report, is typified by the following exchanges at the hearing:

Senator BRANDIS (Liberal): —In any event, there is no doubt in your mind that you are not alone in being misled—that there were a substantial number of other people misled in the same way as well. Bronwyn Smith —That is right. (JSCEM, 2005b: 29)

In reference to this exchange, the Committee report stated that ‘Ms Smith characterised the number of people who were misled as “substantial”’ (JSCEM, 2005a: 119). Although Ms Smith agreed with the comment, she did not use the word ‘substantial’; this was suggested by the Liberal Senator.

Senator BRANDIS (Liberal) —What was it? Tell us. Mrs Flower —It was a bogus party, set up to steer votes away from the National Party. CHAIR (Liberal) —To deceive people. Mrs Flower —To deceive them. Although someone in the room has another theory. (JSCEM, 2005b: 38)

The report printed the majority of this exchange, except for the final sentence referring to another theory (JSCEM, 2005a: 118). The Committee’s report found that ‘Ms Elliot [the Labor candidate] was elected as a result of preferences on the basis of deceptions by Liberals for Forests’ (JSCEM, 2005a: 119). However, the Labor members of the Committee included a minority report, attacking the majority report for its ‘inflammatory allegations’, concluding that ‘the allegations made in the Majority Report are nothing more than a political stunt on behalf of the Coalition’ (JSCEM, 2005a: 368– 9). The regular insertion of minority reports in the JSCEM reports indicates the political nature of the committee. The report was subsequently referred to in parliamentary debates on the Howard Government’s 2006 electoral reforms, which included the deregistration of parties such as the Liberals for Forests. The investigation

65 into the Richmond result demonstrates inherent partisan biases in the inquiry process and committee reporting.

New South Wales Joint Standing Committee on Electoral Matters

The New South Wales Joint Standing Committee is a seven-member committee which was established in 2004. Its terms of reference require it to have a majority of government members. Currently the three non-government members represent the Liberal, National, and Greens parties. Like its Federal counterpart, the committee is developing a practice of inquiring into general elections, having conducted inquiries into the 2003 and 2007 elections. It has also inquired into voter enrolment issues (partly in response to reforms at the Federal level). Unlike its Federal counterpart, however, the NSW committee has not developed a partisan culture of majority and minority reports. Nor does it appear to have a comparable level of public interest in its inquiries, with only 14 submissions received for its 2003 election inquiry, and 19 in 2007. Two days of hearings were held for each inquiry.

Victorian Electoral Matters Committee

Victoria’s Electoral Matters Committee was only formed in mid-2007, but has already completed an inquiry into the State’s November 2006 general election. The inquiry received 28 submissions and held two days of hearings. It is currently conducting two inquiries, on political donations and disclosure and on voter participation and informal voting, with reports due in 2009. The seven-member committee has a government majority, currently with four Labor and three Liberal party members. The other parties in parliament, the Nationals, Greens, and Democratic Labor Party, do not have representation.

Although it is difficult at this early stage of the committee’s work to determine a particular culture of reporting, its first report is encouraging. The 269-page report provides an extensive analysis of various aspects of the 2006 election, including enrolment procedures, party registration, electronic voting, and ballot paper design. It is interesting to note the absence of dissenting (minority) reports, and the fact that, of the report’s 72 recommendations, 56 are directed to the Victorian Electoral Commission, with only 11 recommending legislative action (Electoral Matters Committee, 2008: xxi– xxvi). This may indicate an emphasis on administrative detail rather than electoral

66 system reform, but also the greater powers that the Victorian Electoral Commissioner holds in comparison with his Commonwealth counterpart.

Queensland Legal, Constitutional and Administrative Review Committee

Queensland’s LCARC is a seven-member committee, with a Labor government majority and representatives of the other parties in parliament (National, Liberal, and One Nation). The intention to hold regular election inquiries is clearly evident in the case of the three largest jurisdictions. However, the situation for Queensland is not as clear. LCARC’s remit is centred on reform, with four areas of responsibility: administrative review reform; constitutional reform; electoral reform; and legal reform. Of the 66 reports published since 1996, only nine have specifically dealt with electoral reform issues: truth in political advertising; electoral legislation; how-to-vote cards and appeal processes; electoral fraud; young people engaging with democracy; and Indigenous participation. Because LCARC is concerned with issues beyond electoral matters, it has been difficult to establish regular election inquiries since its inception, with only one report so far on the 1998 election (receiving 25 submissions). In August 2008, however, the committee announced an inquiry into ‘Certain Contemporary Electoral Matters,’ with a focus on the 2004 and 2006 general elections.

Among the previous and current Queensland committee members interviewed, there were different views on the merit in having an automatic reference to initiate an inquiry immediately following each general election. One position was that a regular inquiry was necessary, and this was best done as soon as possible after an election while issues remained fresh. It was also stated that the committee should not necessarily wait for a request from the Attorney-General before an inquiry commenced –– an inquiry immediately after an election should be a standing order . An alternative view was that, as the committee dealt with other issues as well as electoral matters, a regular election inquiry would take up too much time and resources. This committee member also supported the incorporation of electoral matters into constitutional and legal issues based on the broader concepts of democracy and representation: I don’t think the community or the public is that desperately concerned. The media does its job to scrutinise the winners and losers for the rest of the next term, but the community just gets on with it. I don’t think the committee would get to do anything else, if that was dictated to them.

67

Senate Finance and Public Administration Standing Committees 15

The committees discussed above have an ongoing responsibility to inquire into electoral matters. Other committees receive inquiry referrals on a ‘case-by-case’ basis. For example, the Commonwealth Electoral and Referendum Legislation Amendment Bill 2006 was referred to the Senate Standing Committee on Finance and Public Administration (SSCFPA) in December 2006. The Bill was consequential to the major electoral reforms of the Howard Government, which had been passed earlier in 2006, and many of the amendments were considered to be ‘machinery’ in nature. The inquiry took just over two months, received only three submissions (from the AEC, the Human Rights and Equal Opportunity Commission, and the Department of Defence), and did not hold any public hearings. The eight-member committee delivered a unanimous 14- page report recommending that the Bill be supported in its entirety (SSCFPA, 2007).

In contrast, earlier in 2006 the Senate Finance and Public Administration Legislation Committee (SFPALC) considered the highly-contentious major Howard Government reforms (discussed in Chapter 6). The inquiry received 53 submissions, held a day of hearings, and was completed within an intensive seven-week period. Despite numerous concerns being raised in submissions and hearings, the report essentially rubber- stamped the government legislation, making only one recommendation –– that the legislation be passed without amendment (SFPALC, 2006: 33). Almost half of the 92- page report consisted of dissenting reports written by the Labor and Democrats members of the committee (SFPALC, 2006: 35–74). These two examples of committee work demonstrate the diversity that can exist in the inquiry process. The latter example indicates that where partisan interests are at stake, an inquiry will more readily split on party lines instead of seeking a common view consistent with fairness principles and accepted norms for electoral management.

NSW Select Committee on Electoral and Political Party Funding

Political imperatives obviously play a role in how parliaments choose to inquire into electoral matters. A good example of this occurred in the New South Wales Parliament in mid-2007. Despite New South Wales having a dedicated electoral matters committee, the Legislative Council established a select committee, the Select

15 Due to a restructure of Senate committees in 2006 – combining legislation and reference committees into one committee – the two committees referred to in this section have different names but are essentially the same committee.

68

Committee on Electoral and Political Party Funding (SCEPPF), to inquire into political finance and public funding issues. The political manoeuvring involved in establishing the committee, and later in determining the committee membership, is obvious from the Hansard transcripts.

The motion to establish the committee was debated on 27 June 2007, with the mover, Liberal Don Harwin, arguing that a select committee was a better option than the already-established NSW JSCEM, as he envisaged ‘looking over the horizon and looking at policy issues…not about getting bogged down in the administrative minutiae of particular elections, particular donations and particular items of expenditure’ (Harwin, 2007: 1807). Although the intention may have been genuine, the result was that by using an upper house committee the Labor Government would not have a majority, as was the case on the NSW JSCEM. By winning support from other right-of- centre parties, the Liberal party was also able to pass their motion to set up a committee ahead of another motion from the Greens, who sought to establish a similar (but joint house) inquiry.

The committee membership was determined three months later, with agreement that there would be two Labor Government members, two Coalition opposition members, and two cross-bench members. However, in what was obviously further manoeuvring by the parties, the Council had to vote to appoint two of the three cross-bench members who nominated from the Shooters Party, the Christian Democratic Party (CDP), and the Greens –– with the Greens. The Greens candidate, Lee Rhiannon, lost out (NSWLC, 2007: 2264). As the Shooters Party and CDP are considered right-of-centre parties, this outcome could be perceived as being part of an agreement made to establish the committee. A positive aspect of the outcome, however, was that five parties were represented on the six-member committee, providing broader representation of minor parties compared with a joint house committee.

The committee carried out a nine-month inquiry, receiving 189 submissions and conducting five days of public hearings. When compared with the NSW JSCEM’s 2003 and 2007 election inquiries, the SCEPPF attracted far greater public interest and media attention. It tabled an extensive 276-page report on 19 June 2008, with 47 recommendations, including a $1,000 limit on the size of donations, a cap on campaign expenditure, and increasing the frequency of donation disclosures to every six months

69

(currently every four years). There was general consensus on the recommendations, with only one brief dissenting report from the two Labor members, opposing one recommendation and suggesting a minor change to one other (SCEPPF, 2008a: 258).

The inquiry was, however, overtaken by revelations from the Independent Commission Against Corruption that Labor party members had been involved in corruption at the local government level, with political donations from developers a critical factor (the ‘Wollongong scandal’). This created a major crisis for the Government. In February 2008, Premier Morris Iemma announced the Government’s intention to reform political finance legislation. The Government introduced a Bill the day before the SCEPPF tabled its report. Among other changes, the legislation introduced six-monthly donation disclosures, with a disclosure threshold of $1,000. The Bill was passed the following week, progressing through both houses in less than 24 hours.

In terms of being an impetus for reform, the committee report, while substantial in its considerations and findings, was secondary to the media exposure on political finance issues, and the willingness of the incumbent governing party to initiate reforms. The committee did not have a Labor majority, and the Iemma Labor Government chose to pursue a separate process of reforming electoral laws. There are however, many issues raised by the SCEPPF inquiry that were not addressed by the Government’s reform Bill, and the report remains useful as a potential driver for future reforms.

Engagement with Citizens and Civil Society Organisations

As well as producing inquiry reports and recommending changes, the engagement that inquiries generate is an indicator of democratic behaviour. Such engagement is directly related to the principle of having equal opportunity to participate in the electoral process. It can be seen that the parliamentary committee activity discussed in this chapter provides opportunities for individual citizens, civil society organisations, and other groups to engage in the reform process. While it is to be expected that political parties are regular and substantial contributors to inquiries, the make-up of the committees highlight their lack of impartiality; the vast majority of committee members

70 are members of Labor or the Coalition. The strong party discipline that exists in the major parties creates an environment where it is difficult for committee members to put forward recommendations that do not reflect a party’s policy on electoral matters.

Importantly, however, the committee inquiry process allows the general public, academics, civil society organisations and interest groups to raise concerns and put forward suggestions reflecting the democratic principles of fairness and equity that do not necessarily have a partisan bias or advantage. Canberra-based political scientist Malcolm Mackerras is one academic who has been a regular contributor over recent decades to committee inquiries and discussions, playing an important role in the JSCER’s deliberations in 1983 on increasing the size of the parliament (Green, 1986: 97). Similarly, former Australian Electoral Commissioner Colin Hughes, now an Emeritus Professor at the University of Queensland, is a regular contributor, and is called upon for his combination of administrative and academic expertise. The Democratic Audit of Australia, a research unit based at the Australian National University, has made regular submissions to committee inquiries in recent years, drawing upon the expertise of numerous academics around Australia who are knowledgeable about the relevant subject matter of each inquiry. 16 Another influential contributor is elections analyst Antony Green, who works for both the Australian Broadcasting Corporation and the NSW Parliamentary Library.

There are a number of organisations that have been established in Australia with the express purpose of advocating on electoral reform issues, and they use committee inquiries to both influence legislation and to promote their causes to the broader public. Two examples are the Proportional Representation Society of Australia (PRSA),17 and the H S Chapman Society. 18 PRSA (including its predecessors) has existed for more than 100 years and, as its name suggests, is primarily concerned with reforming voting systems from majoritarian to proportional systems. In addition, PRSA provides expertise on the intricacies of vote counting within proportional systems, and its member organisation, the Electoral Reform Society of South Australia, has made submissions to (at least) the past five JSCEM Federal election inquiries. The H S Chapman Society was founded in 1996 and has consistently made submissions to

16 See Democratic Audit of Australia website at http://arts.anu.edu.au/democraticaudit 17 The PRSA includes the Electoral Reform Society of South Australia and Electoral Reform Society of Western Australia as member organisations. Website – http://www.prsa.org.au 18 Website – http://www.hschapman.org

71 parliamentary inquiries and campaigned on issues of removing opportunities for electoral fraud, primarily in terms of enrolment procedures. Despite being debunked by some (for example, see Sawer, 2001b), a former president of the Society, Amy McGrath, has regularly published on these issues (see McGrath, 1996; and Copeman and McGrath (1997).

The conduct of regular parliamentary inquiries also provides opportunities and forums for other organisations to make submissions based on their constituent concerns. For example, there were at least four submissions 19 to JSCEM’s 2004 Federal election inquiry from organisations representing the interests of blind and vision-impaired citizens. All of these submissions advocated the use of electronic voting, an important advance when considering the secrecy of the ballot.20 The JSCEM’s report recommended an electronic voting trial at the 2007 election (JSCEM, 2005a: 258), which was then conducted by the AEC. The trial had limited success, with only 850 vision-impaired and blind voters choosing to vote this way, at an administrative cost of $2,597 per vote (AEC, 2008: 61). However, as the AEC has noted, consideration must be given to the value of all voters having a secret and independent vote, and it has recommended that the trial be extended at the next election (AEC, 2008: 62).

Religious groups also use the inquiry process to re-affirm their positions. A good example is contained in the submission from the Association of Australian Christadelphian Ecclesia to the 2004 Federal election inquiry, 21 which re-asserts the group’s conscientious objection to voting. At a brief hearing in July 2005, the Association expressed satisfaction that its members were able to formally express their conscientious objection to voting, in response to letters from the AEC asking to explain their failure to vote. Interestingly, one Liberal Senator cited the situation of the group’s members as an argument for the reintroduction of voluntary voting (JSCEM, 2005c: 34). Collectively, these contributions from civil society organisations are valuable instances of democratic engagement based on the principle of equal opportunity to participate. As previously discussed, however, the value of inquiries in this respect is tempered by the influence of partisan interests on inquiry outcomes.

19 Submissions 54, 101, 135, and 138 at http://www.aph.gov.au/house/committee/em/elect04/subs.htm 20 Without electronic voting, blind and vision-impaired citizens must rely on other people to assist them to vote, and therefore cannot exercise a secret ballot. 21 Submission No. 27 at http://www.aph.gov.au/house/committee/em/elect04/subs/sub027.pdf

72

Summary

It can be seen from this chapter that Australia’s electoral institutional structure is strongly focused on parliamentary activity, both in legislating for the administrative and operational regimes in which the electoral commissions operate, and in creating forums for public debate on electoral matters. The legislative focus is not surprising, given that Australian parliaments and governments use legislation to maintain a tight control on the form of electoral systems and administration of elections. In this environment, the parliamentary committee inquiry process is a critical element of the institutional structure of Australian electoral systems. In the absence of independent agencies with the powers to implement reforms, committee inquiries have become the primary forum for public debate on electoral matters. Inquiries provide a conduit through which electoral commissions and citizens can reach the electoral lawmakers. Because the submissions are made publicly available on committee websites, a wealth of information is disseminated to interested groups, media attention is raised, and the wider public becomes aware of electoral issues.

The governing party has a membership majority on all of the electoral matters committees, followed by representation from the opposition, and typically one member representing minor party interests. It is a concern, however, that the Victorian committee has only Labor and Liberal party members, with no representation from the other three parliamentary parties. While the membership of these committees is generally representative of the composition of the respective parliaments, it does raise questions about the usefulness of the committee process, and the outputs of the committees. By reflecting the make-up of the parliaments, the committees simply reinforce any disproportionality or bias that is created by election outcomes. In doing so, they create a significant incumbency advantage, primarily for the incumbent government over the opposition and minor parties, but also for parties with parliamentary representation over non-parliamentary parties (reinforcing the hierarchy of incumbency structure depicted in Chapter 2).

Governments’ dominance over these committees runs counter to the participation principle of fairness, which stresses equality of opportunity. While it is not feasible or

73 necessarily ideal for all parties to have equal representation, an improvement on current structures consistent with the participation principle would be to have an equal number of government and non-government members on these committees. In regard to issues- specific committees, the establishment of the non-government dominated SCEPPF in New South Wales is a good example of partisan interests at work. Although the SCEPPF membership was reflective of party numbers in that State’s upper house (albeit with minor parties over-represented), the debate over membership between parties is indicative of the importance that parties place on being involved in electoral matters that directly impact on their own future prospects.

Several committee members commented in the interviews that electoral matters committees are extremely partisan and are primarily a means for furthering parties’ interests rather than assessing the need for reform on the basis of fairness and equity, consistent with international standards for electoral management and fair elections. It can be argued that much of the committees’ work is of limited value, as members are simply pursuing a pre-determined policy agenda. This is particularly evident with the Commonwealth JSCEM, given its long history of reporting and with dissenting reports a common occurrence. While there are some encouraging signs of bipartisan co- operation occurring within State-based committees, it must be remembered that all of the parliamentary oversight committees are dominated by Labor and the Coalition parties. Co-operation can be a sign of party cartelisation rather than working towards democratic values of fairness and equity.

The creation of electoral committees over the past 25 years indicates the importance of this policy area, and the number of submissions and hearings for these committees is indicative of the level of interest and knowledge that Australian society has about the electoral process. Citizens and civil society organisations that contribute to committee inquiries are often guided by principles of democracy and fairness. Benefits that flow from the advocacy of representative organisations, such as the trial of electronic voting, are indicative of the value of these consultative processes. However, there is also clear evidence that submissions (and not only those from political parties) are used to seek partisan advantage. It is disappointing that all of these contributions, irrespective of motives, are often reduced to arguments for committee members to pursue their own partisan objectives in committee reports and dissenting statements.

74

The experiences detailed in this chapter suggest that, on principles of fairness, a more appropriate process for inquiry into electoral issues could be the creation of a forum outside parliament that involves a broader range of political parties, electoral administrators, civil society organisations and other electoral experts, with processes and reporting not under the control of government or other parliamentary parties. Such a forum could be influential as a driver for electoral reform. However, given that parliaments have the power to regulate almost every aspect of elections and electoral administration, the recommendations of such a forum would ultimately be subject to the wishes of governing parties and party cartels.

With its partisan appointments, the NSW Election Funding Authority is an anomaly in the institutional structure of Australian electoral management. There appears to be no evidence of corruption, or serious practical problems created by these appointments in the administration of funding, or in initiating proceedings against non-compliant candidates. With these political appointments in place however, a perception of bias in the Authority is created, and this could impact negatively in the future on the perceived independence of the NSW Electoral Commission and Commissioner. A simple solution would be to replace the partisan appointees with suitably qualified experts who have no party connections.

Despite the strong governing party control of electoral legislation, commissions have been established in all Australian jurisdictions to administer the electoral system ‘independently’ of government. In the following chapter, the thesis moves to an examination of how independently Australia’s electoral commissions really are.

75

Chapter 5

The Independence of Electoral Management Bodies

There would be a perceived political advantage for the government if I was doing a particularly good job getting people on the roll and getting them voting.

I know one electoral commissioner who, when the results came in, said ‘that’s my job gone’. –– two Australian electoral commissioners

The previous chapter presented an overview and assessment of the Australian institutional framework for electoral administration, concentrating on the establishment of electoral commissions, the structure of the New South Wales Election Funding Authority, and the work of parliamentary oversight committees. It is generally accepted that for the conduct of free and fair elections, electoral management bodies should be allowed to act independently of any partisan influence.

This chapter provides further analysis of the institutional framework by assessing the role and performance of Australian electoral commissions, with a particular emphasis on determining to what degree they are independent of government influence and control. Matters considered include the definition of independence, commissioner appointment processes and security of tenure, and funding and budgetary processes. In addition to an analysis of the relevant legislation for each jurisdiction, material is drawn from the interview data to further identify how these factors impact on the independence of electoral administrations in practice. The chapter concludes with an analysis using

76 the modified Elklit and Reynolds framework to determine best practice among the nine jurisdictions.

Models of Independence

In the literature on electoral system management it is widely accepted that, to ensure free and fair elections, electoral management bodies should be independent, both of the government of the day and of any political partisan connections (Dacey, 2005: 1). The International Institute for Democracy and Electoral Assistance (IDEA) argues that legitimacy is enhanced if electoral authorities are perceived to be impartial and not subject to political interference or control (IDEA, 2006: 71), the argument being that election results are more likely to be accepted by the electorate if there is a strong perception of independence, irrespective of any basic measures of independence. Orr, Mercurio and Williams identify electoral authority independence as being the single most important factor in ensuring free elections (2003: 399).

The ACE Electoral Knowledge Network (ACE) distinguishes two different dimensions of independence. First, ACE identifies ‘structural’ independence, in which the electoral management body is formally separated from the executive branch of government, through constitutional or legislative mechanisms. Second, there is ‘fearless’ or ‘behavioural’ independence, which is a ‘normative independence of decision and action’ that does not allow government, political or partisan influences to alter behaviour or actions (ACE, 2007). ACE goes on to point out that while the two concepts of independence are linked, structural independence does not provide any assurance that an electoral management body will act as a fearlessly independent organisation.

Different strategies can achieve such independence. In some countries, partisan balance is sought by allowing various parties to make appointments to electoral bodies. In other countries responsibility for electoral management is delegated to a non-partisan body. Both of these strategies are used by governing and opposition parties as ‘structures of mutual constraint’, to achieve a neutral bureaucracy (Mozaffar and Schedler, 2002: 16).

77

In Australia, delegation to a non-partisan body has been the typical method. The level of bipartisanship achieved in delegating control to a non-partisan public service bureaucracy has, however, been influenced by the degree of control the ruling party wishes to maintain over the electoral legislation, especially when the government has a parliamentary majority. One example of the control that governing parties can exert over electoral management is in section 206 of the Commonwealth Electoral Act 1918 , which stipulates that pencils, and not pens, are to be used in marking ballot papers. To change to the use of pens would require an Act of parliament.

López-Pintor (2000) draws on the work of Garber (1994) and Harris (1997) to describe four approaches to organising electoral management bodies: a governmental approach, with elections conducted by civil (public) servants; a judicial approach, in which judges are appointed to administer the election; a multi-party approach, where the electoral body is composed of party representatives; and an expert approach, in which political parties, by consensus, delegate responsibility to a group of experienced individuals with a reputation for independence (López-Pintor, 2000: 20).

In another typology, Massicotte, Blais and Yoshinaka (2004) identify three approaches to establishing electoral authorities. Their analysis focuses on who is appointed as the person responsible for making decisions about election administration (2004: 83). In their model, Massicotte, Blais and Yoshinaka differentiate between: the appointment of multiple commissioners to represent a diversity of (usually political) views; the practice of allowing a government minister to be in charge of the electoral process; and the appointment of a single commissioner (Massicotte, Blais and Yoshinaka, 2004: 83). It is the last of these that is favoured in Australia.

As previously mentioned, Australia has a long history of professional electoral management bodies that administer elections in a relatively fair and non-partisan manner. All Australian jurisdictions, except South Australia, currently have electoral commissions. South Australia has an electoral ‘Office’, but is expected to move to the commission model in 2008. In addition, the senior electoral official in South Australia is the Electoral Commissioner, and electoral officers and politicians share the view that the Office already acts as a commission in most aspects of its work. While each electoral commission is a statutory body under the relevant legislation in its jurisdiction,

78 the same legislation can also limit its ability to act independently and its capacity to provide a level playing field for participants in elections.

In research conducted for IDEA, Wall et al. (2006) identify three models of electoral authority. These models are sorted according to a range of criteria: institutional arrangements; implementation; accountability; powers; composition; security of tenure; and budget control (Wall et al., 2006: 9, see Table 5.1). Briefly, the three models are: independent (being institutionally independent from the executive); government (within or under the direction of a minister and department); and mixed (a combination of the first two models, with a degree of institutional independence, but still within the direction and control of the government of the day) (Wall et al., 2006: 9).

While IDEA identifies the Australian system as an example of the independent model (Wall et al., 2006: 304), there is one important criterion that the Australian Electoral Commission, and other Australian commissions, do not meet –– namely, having the power to independently develop the electoral regulatory framework (2006: 9). In Australia, this power resides with the parliament and government of the day, and is critical in limiting the commissions’ ability to operate independently. In this respect, Australian electoral authorities conform more to IDEA’s government or mixed models. As discussed later in the chapter, it is also questionable whether Australian commissions meet the independence criterion of having the ability to manage their budgets without day-to-day government control.

Table 5.1 IDEA’s Model of Independent Electoral Management Bodies

Aspect Independent electoral management body Institutional arrangement Is institutionally independent from the executive branch of government Implementation Exercises full responsibility for implementation Formal accountability Does not report to executive branch of government but with very few exceptions is formally accountable to the legislature, judiciary or head of state Powers Has powers to develop the electoral regulatory framework independently under the law Composition Is composed of members who are outside the executive branch while in office Term of office Offers security of tenure, but not necessarily fixed term of office Budget Has and manages its own budget independently of day-to-day governmental control Source: Wall et al., 2006: 9.

79

The term ‘independence’ is often used interchangeably with ‘neutrality’, ‘non- partisanship’ and ‘impartiality’; however there are differences between the meanings of these terms. An electoral authority may be established as an independent body, but exhibit bias through partisan actions. Conversely, an authority that is not independent, for example one that is an office entirely within a government department, may operate in a non-partisan and impartial manner due to a lack of partisan direction from the responsible minister and the neutrality of its public service bureaucracy. Paul Dacey (2005) argues that the Australian Electoral Commission was created with the intention that it would be both independent (that is, not influenced by others and thinking for itself) and impartial (without allegiance or obligation to any political parties, candidates, or other political players) (2005: 2–3). It cannot however, be argued that the commission functions in a totally non-partisan manner, as it is constrained by the legislative environment in which it is required to operate.

Furthermore, it is generally agreed that electoral authorities should not only be independent and impartial, but also that they should not allow for any perception of dependence or partiality to occur (Dacey, 2005: 4; Orr, Mercurio and Williams, 2003: 400). To this end, IDEA identifies five essential criteria for ethical electoral administration: respect for the law; non-partisanship and neutrality; transparency; accuracy; and service to voters (IDEA, 1999: 9–15). Under the ethical principle of non- partisanship and neutrality, it is the perception of neutrality that is seen as the critical factor in successful elections. IDEA argues that one way to maintain neutrality is for electoral administrators to refrain from expressing any view that could become a political issue in an election (1999: 12). However, when the electoral system, or a component of it, becomes an election issue, administrators can be caught between the two options of remaining silent on a matter, in a way that may advantage a particular party or parties, or expressing a view that is based on the principle of electoral fairness for voters, but which may be against government policy. This dilemma was expressed by one commissioner in these terms:

If something’s come up that’s already partisan, you’ve got to be very careful particularly, as I said, my own personal view was, if I had to put a personal view I would’ve been supporting the [change in legislation], which is a bit odd I know, but I didn’t think it was in anybody’s interests on either side for me to say that.

80

A different view, at least in the context of the Australian Commonwealth system, is given by Dacey, who argues that section 7 (1) of the Commonwealth Electoral Act 1918 places the commission at the centre of political debates on electoral matters. Through this section, the commission is required to promote public awareness, provide advice to parliament and conduct and publish research. To maintain independence, Dacey argues that the commission should comment from the viewpoint of improving the quality of the electoral process (Dacey, 2005: 6). Dacey is a long-time senior administrator in the AEC, 22 and his view is from the perspective of electoral professionalism, in which best electoral practice is seen as taking precedence over any partisan impacts that such practices may have. For example, independent commissions may aim to maximise both the comprehensiveness and the accuracy of the electoral roll. There may however, be considerable tension between these dual goals. Improving the accuracy of the roll may occur at the expense of its comprehensiveness and, importantly, this may have partisan impacts whereby some groups of voters, who are more inclined to support certain parties, can become disenfranchised.

In terms of independence, another aspect of electoral administration that has been debated in recent years is whether electoral commissions should be involved in policing ‘truth in political advertising’ legislation. Currently South Australia is the only jurisdiction with such legislation (discussed further in Chapter 7), which requires the State Electoral Office to initiate proceedings in cases of false and misleading advertising. One electoral administrator explained the impact on independence this way: If the Commissioner then takes prosecutorial activity against MPs and candidates –– does he prosecute if his job comes up in two years time, and the party has just won government, does he prosecute or not. It’s the ‘without fear and favour’ stuff. How can you act without fear and favour if you’re worried about your tenure.

An Australian Democrats proposal to replicate this power in the AEC for Federal elections was opposed by the AEC on the basis that such powers would impact on the Commission’s independence and ‘reputation for political neutrality’ (AEC, 2001: 2). In its submission to a parliamentary inquiry, the AEC argued that it if it had responsibility for monitoring advertising would be accused of partisanship. As an alternative, it proposed that an ‘Electoral Complaints Authority’, resourced by staff from other

22 Paul Dacey was appointed Acting Electoral Commissioner in September 2008, replacing Ian Campbell.

81 agencies including the AEC, be established for election campaign periods (SFPALC, 2002: 90). In the same inquiry, however, the South Australian Commissioner stated that he ‘welcomed measures that reinforce the probity of the electoral and political systems’ (SFPALC, 2002: 89). The reluctance of the AEC to take on this role is a typical example of the tension that may exist in electoral administration –– in this case, between voters being able to make an informed decision based on truthful information, and having confidence that the election is being conducted by an independent and non- partisan authority.

Based on the models of independence described above and an examination of legislation, eight factors have been identified as being pertinent to the independence of Australian commissions. These are: commissioner experience; the size of commissions; appointment processes; political affiliations; length of tenure; security of tenure; reporting mechanisms; and budget processes. The following sections analyse these factors for their impact on the fairness and independence of electoral administration.

Commissioner Experience

Commissioner experience has not typically been included as a factor in models of electoral authority independence. It is clear from the Australian experience, however, that a group of electoral professionals exists, for whom electoral administration is a lifetime career. In this environment, working to professional standards rather than to political or management direction is an important component of a culture of independence. A commissioner’s previous experience in electoral administration prior to appointment can impact on his or her ability to act in an independent manner. Generally Australian commissioners have each established a lengthy professional career based on the application of non-partisan principles of fair elections (see Table 5.2). Of the nine commissioners, only one, Ian Campbell, came into the position with no previous electoral experience. Campbell’s professional background lies in senior public service administration, including time as a departmental deputy secretary and deputy president of the Repatriation Commission. This lack of electoral experience was a concern for one observer: I think he’s misunderstood his position to a degree. That is, he’s emphasised more the public servant role and the implementation of government policy due to

82

his background, rather than his independence and guardianship of the system role which is his as a statutory office holder.

While commissioners’ previous electoral experience can and does have an impact on the independence exhibited by commissions, it is not included in the framework analysis of the independence of electoral administration. This is due to the transient nature of this factor. It is suggested that future comparative work on electoral management bodies should take into consideration the variable of professional versus generalist appointments to determine its relevance to independence.

Table 5.2 Electoral Experience of Current Australian Electoral Commissioners (as at 1 July 2008) Jurisdiction Commissioner Year first Previous Previous electoral appointed electoral appointments experience Commonwealth Ian Campbell 1 2005 Nil Nil NSW Colin Barry 2004 16 years Electoral Commissioner, Victoria Victoria Steve Tully 2005 17 years Electoral Commissioner, South Australia (1997–– 2005) Queensland David Kerslake 2006 4 years Assistant Commissioner, Industrial Elections, and Funding and Disclosure AEC (4 years) Western Warwick Gately 2006 3 years Deputy Commissioner (8 Australia months) Acting Commissioner (2 years) South Australia Kay Mousley 2006 + 20 years Various –– finally Director of Operations, AEC (South Australia) Tasmania Bruce Taylor 2005 21 years Director, Industrial Elections (AEC, Tasmanian Division, 1984–93) Deputy and Chief Electoral Officer, Tasmania (1994–2005) ACT Phillip Green 1994 13 years Australian Electoral Office (then AEC), from 1982 Northern Bill Shepheard 2005 26 years Western Australian Territory Electoral Commission (7 years), Australian Electoral Office (then AEC) (c. 17 years) Northern Territory Electoral Office (2 years) 1 Campbell was replaced as Commissioner in September 2008, three years into his five-year term. Paul Dacey served as acting Electoral Commissioner, prior to Edward Killesteyn being appointed as Commissioner in December 2008. Like Campbell, Killesteyn has a public service background.

83

The Size of Commissions

The number of commissioners to be appointed to commissions has not been a contentious issue in Australian electoral management. All jurisdictions operate with a single commissioner appointed in a full-time capacity as the chief executive officer of the organisation. In three jurisdictions –– the Commonwealth, Tasmania, and the ACT –– two other commissioners are appointed on a permanent part-time basis, and tend to be drawn from the legal profession (a senior judge) and the senior public service (usually with expertise in statistics and demographics). In the remaining jurisdictions, additional commissioners are appointed to a separate body when there is a need to conduct a redistribution of electoral boundaries. For the purposes of this study, unless otherwise specified, a reference to a ‘commissioner’ refers to the commissioner appointed as the full-time chief executive officer of the organisation.

Current commissioners have differing views on the merits of having a single commissioner, as against a multi-member board commission. Some prefer autonomous decision-making but there is also support for collaborative and consultative management. A preference for the latter was expressed in the following terms by one commissioner: A commission can stand as three people and as a Body we’ve considered this and we’ve done it, and it’s not, like, that you’ve personally had any biased personal opinion, that you can be labelled yourself. It’s a body. It’s not just me thinking this way because of any biases. This body of three people sat down and deliberated and decided this. What my personal views are, who knows. It gives it that air of independence that the body’s doing it.

Generally, there was agreement among commissioners that the size of the commission did not have an impact on the independent conduct of commission work, and that an ideal model for the size of a commission depended on each commissioner’s personal preference.

84

Commissioner Selection and Appointment Processes

Clearly, the independence of an electoral management body is closely related to the method of appointment of its chief executive officer. Where appointment processes are controlled by government it may be difficult to sustain confidence in the impartiality of the process. Four general methods of appointment exist in Australia (see Table 5.3). In order of increasing accountability, they may be listed as follows. The first is appointment by the Governor or Governor-General, on the recommendation of the government of the day. This process is used in the three largest jurisdictions, the Commonwealth, NSW and Victoria, where the entire selection and appointment process remains within the government’s control and oversight. A second type of process is used in the remaining six jurisdictions, which require consultation with other political parties represented in parliament. The third method, used in Queensland and South Australia, requires consultation with a parliamentary committee prior to the ratification of an appointment. Finally, the South Australian legislation also requires a resolution from both houses of parliament before the Governor can appoint the electoral commissioner.

Table 5.3 Appointment Processes for Australian Electoral Commissioners 23 Jurisdiction Appointment process (section of principal legislation) Commonwealth Governor-General appoints (s.21) NSW Governor appoints (s.21AA) Victoria Governor appoints (s.12) Queensland Governor in Council appoints. The position is advertised nationally and the process includes consultation with all leaders of parliamentary parties and with parliamentary committee (s.23) Western Australia Governor appoints, on recommendation of the Premier, who is required to consult with parliamentary party leaders (s.5B) South Australia Governor appoints, on recommendation from both houses of parliament (s.5) Tasmania Governor appoints, on recommendation of the Premier, who is required to consult with parliamentary party leaders and the President of the Legislative Council (s.8) ACT Executive appoints. Consultation with leaders of all parties and independents required. Appointment is disallowable (s.22) Northern Territory Administrator (equivalent to a Governor) appoints, on recommendation of the Chief Minister, who is required to consult with parliamentary party leaders and independents (s.314)

23 References to section numbers in tables in this chapter relate to each jurisdiction’s primary legislation, listed at Table 1.1.

85

In the three jurisdictions not requiring consultation, there is a danger that appointments may be perceived as partisan. One example of this was the Howard Government’s appointment of Ian Campbell as commissioner in 2005. One of Campbell’s initial tasks was to advise on the significant reforms being proposed by the Government at the time, and following this, to manage their implementation once they were passed in 2006. In February 2008, the Rudd Labor Government announced a transparent and merit-based selection process for senior public servants (Faulkner, 2008a), which is to involve public advertising and involvement by the Public Service Commissioner. However, there remains no requirement for the government to consult with other parties prior to appointing an electoral commissioner.

Appointment by the government of the day, without reference to other parties, leaves such appointees open to claims of bias and partisanship. Typical was this observation from one commissioner: I’ve read all the articles about [the Commissioner’s] appointment and how that allegedly had the Prime Minister’s tentacles all over it.

Such allegations have accompanied both Labor and Coalition appointments. For example, in an interview Tasmanian Liberal, Michael Hodgman, was critical of the treatment he received from Labor-appointed commissioner, Colin Hughes: I went along to that review expecting a fair hearing, and I walked out and they said ‘how do you think you’ve gone?’ I said,…My appeal’s going to be dismissed, absolutely’…I said outside the hearing, ‘I’ve not had a fair hearing’. I know I was causing [Prime Minister] Hawke a lot of pain and angst, and I reckon he said to [Labor Special Minister of State, Mick] Young, ‘get rid of Hodgman’ and Young said, ‘the way to do that is change the boundaries’.

Two Labor perspectives on Coalition-appointed commissioners are similarly sceptical: We were very concerned with [the commissioner] when he was appointed, and as to how he would perform.

He was seen to be partisan, being a particularly good mate of [the Minister].

A clear majority of comments from politicians however emphasised the importance of maintaining perceived and actual non-partisanship. Commissioners were seen, almost universally, as honest and incorruptible, but concerns were frequently raised that if appointments were made without consultation, the subsequent decisions of those commissioners would be open to arguments that the appointment was partisan in nature.

86

This was particularly the case where commissioners’ decisions resulted in measurable partisan advantages or detriments.

The Requirement to Consult

The requirement of consultation with other parties means that it is less likely that a person with perceived partisan views will be appointed. However, while there is a requirement in six jurisdictions for some form of consultation to take place, there is no requirement, except in South Australia (and to a lesser degree in the ACT), for the government to genuinely take the views of the other parties into account. As one minister stated: It was something that [the Premier] is alleged to have done. He said to the other parties, ‘we’re going to appoint so-and-so, you have now been consulted’.

In the case of the ACT, however, commissioner appointments are disallowable instruments. This means that a government without a majority in the Legislative Assembly could be exposed to the possibility of its appointment being rejected in a very public manner. It also means that an Opposition that does have concerns about an appointment needs to decide whether such concerns warrant moving a disallowance motion. Although the ability to disallow an appointment is theoretically a safeguard mechanism that allows non-government parties a voice in the appointment process, it is a blunt instrument that is not likely to be used in practice. It would only be effective when Opposition parties have the numbers in parliament and, irrespective of the outcome, moving such a motion in itself would call into question the office of commissioner. If a motion was successful, it would result in the commissioner being removed from office, damaging the public image of the office and labelling that person as partisan, without that person having a right to defend himself or herself. If a disallowance motion was unsuccessful, however, it would place the commissioner in a very difficult, or even untenable, situation. As one commissioner argued: The other problem with that is if you have a disallowance motion is it can be lost if the government has the numbers in the house, and you get a poor lame duck electoral commissioner who says the other side don’t want me.

How can I maintain any credibility if they don’t want me? So to try to engage them all at some stage in the process and bring the other sides of politics on board in the first place is a much better way than people jumping up and down afterwards.

87

Instead of taking the negative and post-appointment approach that a disallowance motion entails, the South Australian process requires the positive and pre-appointment method of an affirmation resolution by both houses of parliament, following a multi- party committee process outlined below. This is the most thorough and publicly- accountable consultation process of all the Australian jurisdictions. It requires the Governor to appoint a commissioner based on a recommendation made by both houses of parliament. When this requirement was imposed in 2005, for the first and so far the only time, the matter was considered by the Statutory Officers Committee (SOC), a joint house parliamentary committee that was established for the purpose of appointing electoral commissioners, auditors general and ombudsmen. The SOC is made up of six members, three from each house, with only two parties (Labor and Liberal) and an independent represented. Although not all parties are represented on the SOC, the subsequent requirement for parliamentary ratification by way of an affirmation resolution should ensure that other parties are consulted prior to such a vote. Such consultation would not be expected to change the outcome (given the major parties’ representation on the SOC); however it could assist in alleviating any concerns that other parliamentarians may have.

After advertising nationally for the position, an interview panel, made up of an electoral commissioner from another jurisdiction and senior public servants, interviewed a short- list of applicants that had been drawn up by a personnel agency. This panel provided its recommendations to the SOC, which then interviewed the recommended applicants. Based on this second round of interviews, the SOC reported its recommendation to parliament, which was agreed by both houses without debate (Hansard, 6 July 2005). In 2006, the Victorian Public Accounts and Estimates Committee recommended that the South Australian model be followed in Victoria, whereby future appointments would be made by a resolution of both houses of parliament, following a recommendation from an appropriate parliamentary committee (PAEC, 2006: 69).

Queensland is the other jurisdiction that involves the use of a parliamentary committee in the selection process. In Queensland’s process, the requirement is to consult with the Legal, Constitutional and Administrative Review Committee (LCARC), which also has a remit to inquire into broader electoral and other matters. When this requirement was activated in 2005 for the appointment of the current commissioner, the Chair (Lesley Clark, Labor) and Deputy Chair (Fiona Simpson, Nationals) of LCARC both sat on the

88 selection panel (as did the Labor Attorney-General, Linda Lavarch). Although the Act does not require committee representation on the actual selection panel, the involvement of government and opposition parliamentarians also assisted in fulfilling the additional requirement to consult with other parties. As one parliamentarian stated: My experience of [previously] interviewing for the information commissioner where it was just me, really reinforced –– where you have got the external accountability-type office holders –– that you have bipartisan support in that early stage. Not just in the parliament, that’s really after the event.

It was acknowledged by some interviewees that there could be a benefit in the Commonwealth’s Joint Standing Committee on Electoral Matters having a role in the appointment of commissioners, either prior to or after the appointment. However, the dominant view was that the pronounced partisan nature of the committee’s membership could politicise the process and jeopardise the independence of the commissioner’s position. By contrast, Queensland’s use of LCARC in appointing the commissioner appears to have been successful, although it has only been used once so far. One theme that came through strongly from committee members in response to questioning on the suitability of using committees in the selection process was the importance of personal relationships and goodwill between committee members. Such goodwill appears to be evident in the Queensland situation but clearly lacking in the Federal committee.

Political affiliations

Another mechanism in Australian legislation that guards against partisan appointments is the inclusion of rules preventing members or previous members of political parties or parliaments from being eligible for appointment (see Table 5.4). Four jurisdictions (NSW, Victoria, Tasmania, and the ACT) prevent people who have been members of a political party within the previous five years from being eligible for appointment to a commissioner’s position. Queensland limits the prohibition to existing members of political parties. In addition, four jurisdictions (Western Australia, Tasmania, ACT, and the Northern Territory) place restrictions on current or previous members of parliament. Such requirements have generally been introduced as a result of one jurisdiction’s reforms being adopted by other jurisdictions. For example, the Northern Territory’s Electoral Act 2004 was heavily influenced by the ACT’s Electoral Act 1992 . While such restrictions have merit in ensuring there is a check on overt partisan influence, if a

89 partisan appointment were to be made, it is more likely to be of someone with less well- known or obvious political leanings. It is partly to counter this potentiality, and to encourage cross-party support for commissioner appointments, that recent electoral legislation often states a requirement for the government to consult with other political parties before deciding on a preferred candidate, as discussed above.

Table 5.4 Prohibitions on Appointment of Party Members or Parliamentarians Jurisdiction Party members Parliamentarians Commonwealth No provisions in the Act No provisions in the Act NSW Members of political parties, and any No provisions in the Act person who has been a member of a political party in previous five years, are ineligible. (s.21AB(4)) Victoria Members of registered political parties, and No provisions in the Act any person who has been a member of a political party in previous five years, are ineligible (s.12(3)) Queensland Members of political parties are ineligible No provisions in the Act (s.23(4)) Western Australia Act is silent Any person who is or has been a member of a parliament or legislature anywhere in Australia is ineligible (s.5B(10)) South Australia No provisions in the Act No provisions in the Act Tasmania Members of political parties, and any Members of parliament, and any person who has been a member of a party in person who has been a member the previous five years anywhere in of parliament in the previous Australia, are ineligible (s.8(3)) five years anywhere in Australia, are ineligible (s.8(3)) ACT Any person who is or has been a member of Any person who is, or has been a political party in the previous 5 years is a member of a parliament or ineligible (s.12A) legislature anywhere in Australia in the previous 10 years is ineligible (s.12A) Northern No provisions in the Act MLAs ineligible for Territory appointment (s.327)

Length of tenure

The length of a commissioner’s tenure can have significant impacts on his or her independence and ability to act without fear or favour. If a commissioner lacks long- term security, then his or her actions may be, in a real or perceived sense, related to a desire for re-appointment. The timing of a potential re-appointment, irrespective of the length of appointment, can also have an impact, especially if this coincides with an

90 election. As shown in Table 5.5, Australia’s electoral commissioners generally have reasonable lengths of tenure, with eight jurisdictions providing for appointments from five to 10 years. In seven of these jurisdictions, however, this is the maximum length, with the relevant Acts allowing tenure ‘up to’ that term. The remaining jurisdiction, South Australia, has potentially longer security, with the appointments of commissioner (and deputy commissioner) lasting to the age of 65.

Table 5.5 Length of tenure for Australian Electoral Commissioners Jurisdiction Length of tenure Commonwealth Up to 7 years. Eligible for re-appointment (s.8)

NSW Up to 10 years. Eligible for re-appointment, for no more than one further term of up to 10 years (s.21AB(1)) Victoria 10 years. Eligible for re-appointment up to a further 10 years (s.12) Queensland Up to 7 years (s.23(5)) Western Australia Up to 9 years. Eligible for re-appointment (s5B(4)) South Australia To age of 65 (s.7) Tasmania Up to 7 years. Eligible for re-appointment (s.17) ACT Up to 5 years. Eligible for re-appointment (s.25) Northern Territory Up to 5 years. Eligible for re-appointment (s.320)

The ‘up to’ provision provides governments with flexibility in determining the length of an appointment, and it has been argued by ministers that appointing for a lesser term fits with standard practice for senior public service appointments. However, shorter-term appointments may be associated with weak independence, if the commissioner is seeking re-appointment. At the Commonwealth level the first commissioner, Colin Hughes, was appointed for the maximum seven years (and served just less than six before resigning). All subsequent commissioners have only been appointed for five- year terms. Similarly, in Western Australia, where terms can be up to nine years, recent appointments have been for five years. The current commissioner, Warwick Gately, who had already served two years as acting commissioner at the time of his appointment, was appointed as commissioner for only three years, to make a total of a five-year term.

In the case of the current South Australian commissioner, her appointment to the age of 65 is effectively a 14-year term. Her deputy commissioner was appointed at a younger age, and therefore has an effective appointment in that position of 26 years.

91

Internationally, the only jurisdictions with a similar length of security are Ghana (to 70 years), Canada and Malaysia (to 65 years), India (six years or to 65 years, whichever is earlier), and Poland (to 70 years) (Wall et al., 2006: 93; Massicotte, Blais and Yoshinaka, 2004: 85, 87).

In the interviews, Commissioners expressed a range of views on what they considered to be an ideal length of appointment. One argument in favour of five-year terms was expressed as follows: Five years, if you’re going to shake a show up, then if you haven’t got everything accomplished that you want in five years, you’re not going to get it, and if you go on about it, people are going to say ‘he’s still on about that’. So five years is your effective working life. Secondly, if you were a dud, five years of coasting along, they will have a chance of replacing you.

Generally though, commissioners expressed support for a longer term. Typical of such views were these commissioners’ comments: Having tenure in the appointment of more than five years is desirable from a perspective of long-term planning in electoral matters. I think you need to plan over two election cycles, so eight to ten years is about a good time in my view. Otherwise you just get short-term bites at the planning and no look above the horizon.

Tenure’s good for being absolutely seen and perceived as free and fair and impartial and having confidence that you can do your job properly without any perception of interference. If you’re in a job around eight years, people are getting sick of you, you’re getting sick of them, and it’s good for the organisation and for the Commissioner to do something else.

The Australian Electoral Commission tends to appoint people as their last job, as Commissioner, and they tend to serve five years, and my sense is that one of the things that misses out there is the longer term vision for electoral administration.

If you’ve gone through two elections, you’ve probably established more independence in the role, and perhaps a fairer decision can be made.

There was also support for appointments to the age of 65: With the right person, the ideal is to the age of 65. I’d probably have preferred ten [years]. I think ten gives it a greater longevity, and there’s more security for people coming into the job. There have been in some other jurisdictions questions over the commissioner being swapped every time a government swaps. I don’t know how true that is, but it does raise the question in some cases.

92

A dominant view from the interviews was that commissioners should be in place for more than one election, and that five-year appointments, combined with three- or four- year electoral cycles, often meant that a commissioner was only in the position for the one election. This could result in a reduced ability to oversee the implementation of administrative reforms based on previous experience in that jurisdiction. The timing of appointments, and whether the end of the term was close to an election, was also an issue of concern for some commissioners. Interviewees argued that if a possible re- appointment coincided with the conduct of an election, this could impact on the actions of the commissioner during the election period.

Security of tenure

Once appointed, security of tenure can be enhanced or diminished by the conditions under which an appointment can be terminated. All jurisdictions provide for dismissal from office under specified circumstances, such as physical or mental incapacity, bankruptcy, and misconduct. However, for such a dismissal to take permanent effect, a resolution passed by each house of parliament is required in seven of the nine jurisdictions, with the Commonwealth and Queensland being the only exceptions to this requirement (see Table 5.6). In these two exceptions a government still needs to substantiate its reasons for dismissal (under separate employment legislation). However, the need for a parliamentary resolution provides an obviously greater safeguard against governments acting vexatiously. Some jurisdictions specify a number of sitting days within which a resolution for dismissal needs to be passed for it to take effect. Depending on the parliamentary sitting schedule, and the timing of parliament being notified, this could equate to a period of several months. Some Acts do not specify a time period, but a deadline for action may be covered in other legislation, such as constitution Acts or in parliamentary standing orders.

93

Table 5.6 Dismissal Processes for Australian Electoral Commissioners

Jurisdiction Dismissal – reporting to parliament Commonwealth Governor may terminate, for specified reasons (e.g. misbehaviour) (s.25) NSW Governor may suspend the Commissioner. A statement from the Minister explaining the suspension is to be provided to parliament within 7 sitting days. Commissioner can then be removed by the Governor if each house of parliament passes a resolution within 21 days of statement being tabled (s.21AB(3)) Victoria Governor may suspend the Commissioner. The Minister is then to notify the Speaker, President and leaders of (parliamentary) political parties within 2 hours (s.140). Commissioner can then be removed by resolution of both houses of parliament (s.12(4)(e)) Queensland Governor in Council may terminate (s.26) Western Australia Governor may suspend the Commissioner. A statement explaining the suspension to be provided to parliament within 7 sitting days. Commissioner may be removed on resolution of both houses of parliament within 30 sitting days (s.5C) South Australia Governor may suspend the Commissioner under specified circumstances . A statement explaining the suspension is to be provided to parliament within 3 sitting days. Commissioner may then be removed by resolution of both houses of parliament (s.7) Tasmania Governor may suspend the Commissioner under specified circumstances . A statement explaining the suspension is to be provided to parliament within 7 sitting days. Commissioner can then be removed by resolution of both houses of parliament (s.21) ACT Executive may suspend. Minister is to present a statement to the Assembly on the next sitting day. Assembly resolution within 7 sitting days required for appointment to be ended. (s.29) Northern Territory Suspension by the Administrator. Minister must present statement to Assembly within 3 sitting days. Assembly must pass a resolution for dismissal to take effect (s.323)

Reporting Mechanisms

The independence of electoral administrations is also influenced by the method with which they report to government and parliament. Reporting directly to parliament provides a transparent process that is accessible to all political stakeholders at the same time. Reporting only to the government can provide the political parties in government with the advantage of having earlier access to information on sensitive issues, thereby allowing the government to prepare its response prior to the public release of information. In Australia, there is a mix of reporting mechanisms for the various electoral administrations, as shown in Table 5.7. Although possibly not a major issue, delays between a government receiving a report and tabling it in parliament can lead to accusations of a closer relationship between the electoral administration and the government than is desirable. For example, the 15 sitting day period allowed in section

94

17 of the Commonwealth Electoral Act 1918 could stretch out for several months, depending on parliament’s sitting schedule.

Table 5.7 Reporting Mechanisms for Australian Electoral Administrations Jurisdiction Reporting mechanism Commonwealth Annual, election, and financial disclosure reports to the Minister, who then must table the reports in Parliament within the next 15 sitting days (s.17) NSW Not specified in Act. Election report sent to Premier, with a request that it be tabled in parliament (from interview data) Victoria To both houses of parliament, on elections and polls (s.8). Annually to both Houses of Parliament in relation to the provision of enrolment information (s.35) Queensland Not specified in Act. Report to parliament (from interview data) Western Australia Annual report to the responsible Minister by 31 August, who then presents the report to the parliament within 21 days of the Auditor General's report. (s.62 and 64 of the Financial Administration and Audit Act 1985 ) South Australia Not specified in the Act; however, s.8 specifies that the Commissioner is responsible to the Minister Tasmania Annual and other reports directly to both houses of parliament (s.13) ACT Annual report to the Minister responsible (Attorney-General), who then presents the report to the Assembly (s.10) Northern Territory To the Speaker, then tabled in the Assembly within 3 sitting days (s.313)

Budget Processes

Dundas (1994) and Wall et al. (2006) identify the budgetary independence of electoral management bodies as one of the primary guarantors of electoral commission independence, pointing out that the manner in which a commission is funded can affect its independent status. Dundas argues that the need for an electoral authority to negotiate its budget can undermine its primary role as an independent agency (1994:40). Wall et al. conclude that the electoral administration should be provided with its own budget, and should be free of day-to-day government interference in administering that budget (2006:9).

All Australian administrations (except South Australia) have moved to the independent statutory commission model in the past 25 years, but remain reliant on governmental budget processes for their budgetary allocations. Consequently, it is possible for governments to maintain a significant degree of influence and control over the ‘independent’ commissions. It is important to understand the processes through which the commissions receive their financial allocations. The standard method for Australian

95 commissions is to negotiate funds through their parent departments or with the relevant finance or treasury department, as part of the whole-of-government budgetary process. This process can diminish the commission’s independence, especially when compared to the practices of some other countries, such as Canada, where the electoral body has a portion of its budget guaranteed by law (Rose, 2000:7).

Budgeting processes for electoral administrations are often contained within finance and public administration legislation, rather than in electoral legislation, and personal relationships between electoral officials and treasury officials can influence these processes. In order to understand budget processes, all commissioners and ministers interviewed were asked to describe how the budget process works in their respective jurisdiction. A summary of this data is provided in the following paragraphs.

Commonwealth : The AEC undergoes the same process as all government agencies, and is therefore subject to government-wide funding cuts. The AEC negotiates its budget with the Department of Finance and Administration (DoFA). Requests for funding of special programs go to the Special Minister of State. DoFA provides budget and policy advice (beyond just monetary matters) to the Minister responsible. In 2005, the Minister determined the AEC’s resources following a joint finance agencies review.

New South Wales : The NSW commission undergoes the same process as all government agencies. It negotiates its budget with Treasury and, if necessary, the Cabinet budget committee.

Victoria : The Victorian commission negotiates its budget with the Department of Justice; its budget is an output within the Department’s budget. There appears to be no restrictions on accessing funds to conduct general elections, as the following administrator’s comment attests: The Victorian Electoral Commissioner has a bottomless pit to draw on to run a State election, but it’s accountable as to how it’s spent, after the election, and that’s rightly so. But there is no real impediment. The Commissioner can’t say ‘there wasn’t enough money available’. Because if the Commissioner believes the money has to be spent, then they’ll just spend it and be accountable for it later.

Queensland : The Queensland commission negotiates its budget with Treasury, for both recurring and election budgets. Appropriations for special projects require ministerial

96

(Attorney-General) support. There are no restrictions on accessing funds to conduct general elections.

Western Australia : The Western Australian commission negotiates its budget with Treasury and then seeks ministerial (Attorney-General) support.

South Australia : The South Australian Electoral Office experiences the same budget discipline as other agencies. It must negotiate its budget with the Attorney-General’s Department and with Treasury, and then seek ministerial (Attorney-General) support. Special projects will require Cabinet approval.

Tasmania : The Tasmanian commission has a recurrent budget provided through the Department of Justice and is subject to government-wide funding cuts. It must negotiate its budget with the Department and seek the Attorney-General’s support, if necessary. Budgets for general elections are ‘reserved by law’, and are therefore exempt from government restrictions. A commercial trust account is used for the conduct of local government and other elections.

Australian Capital Territory : The ACT commission negotiates its budget with the Department of Justice and Community Safety, and then seeks the Attorney-General’s support, before going to Cabinet. Funding goes directly to the Department, which then passes funds onto the Commission, minus departmental expenses.

Northern Territory : The Northern Territory commission must seek Treasury approval for its budget, based on previous recurrent and election budget costs. It is required to seek Cabinet approval for additional funds.

Budgetary Independence

Around Australia, commissioners expressed a range of views about the level of ministerial or departmental control over their ongoing and election budgets. Concerns included: being part of a departmental appropriation, rather than having a separate agency budget line; being susceptible to government-wide budget cuts; the potential for a government to interfere with an upcoming election budget; and having to argue with departmental officials, rather than the minister, for funding to be maintained for specific

97 programs within the electoral commission budget. Typical of these comments were the following commissioners’ views: So there is a line item in the budget for electoral services, but the amount of money that’s in electoral services is given to the department. They cream an amount that they decide from the top of that to use to fund their corporate things that they say are devoted to electoral services

Because we fit for portfolio purposes under that department, which is a very disparate group of individual organisations, they want to hold the purse strings and when the government says there’s a seven per cent cut in the budget, well they say they want your seven per cent as well.

You can have your budget fiddled with. And I always thought that was dangerous particularly where the CEO’s responsible to the minister, and therefore there’s a direct line with interfering with the election budget.

A further issue concerning whole-of-government budget cuts relates to the application of efficiency dividends, where a commission has to make savings in line with a government-wide standard. As two administrators stated: Each year the Department applies a one per cent, or one and a half per cent [reduction] of your last year’s appropriation. They say, we’re going to take that off you, and you’ll have to be more efficient, because you’re getting less money. They’re taking the dividend now, and forcing you to be more efficient.

The trouble is, this has now been in operation for maybe 15 years, maybe 18 years, so one per cent, one per cent, one per cent, one per cent. In round figure terms, we might be 20 per cent less of a budget than it would have been had it not applied. We’re the same as every other agency, though.

One minister provided a government’s perspective of working through a department as part of the budget process, acknowledging the potential for departmental interference in a commission’s budget: [The Commissioner] puts forward all of his own proposals for funding. Now they are, for administrative reasons, brought in through the portfolio and then brought to Cabinet for the budget process. I want to see all the bids from the Commissioner and statutory officers, so that they’re not filtered.

And that’s a lot of the arguments about, can a statutory officer like the Commissioner communicate fully to the Minister without it being filtered through the bureaucracy, and I think the answer to that is, ‘yes, he can’. If a Minister wasn’t on the ball about that, and if a department wasn’t playing by the rules, things could get filtered I guess, potentially. If the Minister lets it happen and the department was of a mind to do that, but I think an appropriate understanding of the role of the department would mean that that wouldn’t occur.

98

The majority of commissioners expressed a reasonable level of satisfaction with the budgets they receive, but personal relationships with departmental and treasury officials were often mentioned as being very important in ensuring a smooth budgetary process. Typical of this view was the following comment from a commissioner: As long as we’re not looking like we’ve got the hand in the till, and it costs them three times as much as every other State for a similar event, Treasury will just make sure. It’s really independent.

Other commissioners expressed concern that they could not necessarily implement the programs they felt were important. Instead of having ‘a bottomless pit to draw on’, the situation is quite different. As one commissioner said, nobody has a blank cheque. No western country gives any organisation a blank cheque.

In 2006, the Victorian Public Accounts and Estimates Committee recommended that in the future, the Electoral Matters Committee should review the electoral commission’s budget, and report to parliament ahead of the appropriation being passed (PAEC 2006:82). Such a process would open the commission’s budget to the scrutiny of non- government parties, and therefore may assist in maintaining the independence of the commission. However, this proposal, which models the process already in place for the Victorian Auditor-General, is yet to be adopted or implemented.

Determining Best Practice for Independence

The discussion above presents an analysis of the factors that influence the level of independence exercised by Australian electoral administrations. Using the modified framework (based on the Elklit and Reynolds model) described in Chapter 3, a normative assessment can be made to determine which jurisdictions display best practice for each factor, and overall. Of the eight factors discussed in this chapter, six are included as indicators in the best practice analysis shown in Table 5.8. Two factors are not included as indicators: commissioner experience, which relates to individual incumbents and is not comparable over time; and the size of a commission, which has been shown to have negligible, if any, impact on independence. In their model for assessing the fairness of electoral systems in developed and developing democracies, Elklit and Reynolds found that some factors have greater importance than others, and

99 weighted them accordingly. It was beyond the scope of this thesis, however, to determine which, if any, of the indicators are more important in maintaining the independence of the commissions. Consequently, weightings have not been applied in this analysis.

Table 5.8 Best Practice Analysis –– Independence of Electoral Administration Performance Common- NSW Vic Qld WA SA Tas ACT NT indicators wealth Appointment process 1 1 1 4 2 5 2 3 2 Political affiliations 1 3 3 2 2 1 4 5 2 Length of tenure 3 4 5 4 2 4 3 2 1 Security of tenure 1 4 5 1 2 3 2 5 3 Reporting mechanisms 2 1 5 4 2 3 5 2 4 Budget process 3 3 5 4 3 1 4 2 4 Total Score (max. 30) 11 16 24 19 13 17 20 19 16 Intermediary Step 3.7 5.3 8.0 6.3 4.3 5.7 6.7 6.3 5.3 Score Note: 5 = best practice
1 = worst practice Intermediary Step Score is the sum of scores standardised relative to the value of 10

The determinants of best practice for each of the indicators are explained as follows. For the first indicator, Appointment process , multi-party committee involvement followed by parliamentary ratification was determined to be best practice, as this is a transparent process that eliminates any perception of bias. The requirement to consult with other parties is an improvement on direct government appointment with no external consultation, which is considered worst practice. South Australia ranked more highly than Queensland for this indicator due to its requirement for parliamentary ratification.

The presence of prohibitions on Political affiliations reduces the risk of partisan appointments, and there are two components to this indicator: affiliations based on party membership and being a member of parliament. Ratings were applied in terms of the length of prohibition, and whether one or both components are proscribed. The ACT has the most stringent restrictions of five years (for party members) and 10 years (for parliamentarians). Exemplifying worst practice, the Commonwealth and South Australia have no restrictions for either component.

100

Best practice for Length of tenure was largely determined from interview data and international standards (for example, as detailed in Massicotte, Blais and Yoshinaka, 2004), with a fixed term of about two and a half electoral cycles (7.5 –– 10 years) considered to be preferable. Shorter or longer terms were therefore considered less than best practice, as were terms with no fixed minimum period (that is, specifying ‘up to’). The actual length of appointments was also considered. Victoria, with a 10-year fixed term, is considered best practice. It can be argued that a longer term, such as the current South Australian commissioner’s appointment for 14 years, may give a commissioner greater independence. However, appointments to the age of 65, as is the case in South Australia, do not take into account the age of the appointee (for example a 60-year-old would only have a five-year term). Consequently, length of tenure in South Australia is considered less than best practice.

Security of tenure relates to dismissal processes and, although dismissals may be rare or unlikely, the process provides a good indicator of a commissioner’s ability to express independent views without fear of dismissal. Factors include the time required to notify parliament of a dismissal and then the time required to act on a resolution to dismiss. Victoria and the ACT are top-ranked for this indicator, the former due to the requirement to notify parliament within two hours and the latter for requiring the Assembly to resolve the matter within seven sitting days.

For Reporting mechanisms , it is considered best practice for the electoral commission to report directly to parliament, so that the commission is seen as being independent of the government. Another indicator is whether the minister is required to table reports in a specified number of days, as reporting to the government can result in delays in the public release of information. On this score, Victoria and Tasmania ranks highest, while New South Wales, where there is no legislative requirement to table election reports in parliament, ranks lowest.

To determine rankings for the final indicator, Budget process , it was necessary to evaluate commissioners’ experiences in obtaining an acceptable level of funding, and the steps they are required to go through in achieving that. Scores for this indicator were largely based on interview data, with Victoria designated as having best practice, due to the apparent ease of accessing money for the conduct of elections. South

101

Australia ranked lowest, based on concerns of budget cuts, and indirect negotiations for funding.

Based on the combination of these indicators, the Victorian Electoral Commission (VEC) is rated as the most independent of the Australian jurisdictions. The VEC is rated as having the best practice for four of the six indicators, with its commissioner appointment process being the only area of serious concern. The Australian Electoral Commission is rated as the poorest in terms of independence, having worst practice for three indicators and best practice in none.

Summary

This chapter has provided an insight into various aspects of the performance of Australia’s electoral management bodies. As previously discussed, the results of the best practice analysis are best used as a comparative guide between the jurisdictions assessed. It is clear that additional indicators could be used and that weightings could be applied to each indicator in an attempt to quantify their relative importance in the overall independence of an electoral administration. It is necessary, however, to first reach this evaluation point before adding complexity to the calculation of values.

In its current form, the framework analysis is a valuable guide, and can be used as a tool for jurisdictions to improve their ranking in terms of independence. For example, if South Australia was to include in its legislation a prohibition on appointments of people with political affiliations, the State’s overall ranking could jump from fifth to second. Similarly, if New South Wales required its commissioner to report directly to parliament, its ranking would move from equal sixth to equal second.

The electoral commissions provide the organisational basis for the conduct of elections, and this chapter has made a comparative assessment of the fairness and independence of these institutions. The next three chapters take a case study approach to assessing aspects of electoral systems in the various jurisdictions. These chapters will analyse the impact and fairness of aspects of electoral systems and reforms that have been implemented during the 25-year study period.

102

103

Chapter 6

Reform Processes I –– The Commonwealth

The Hawke Government endeavoured to secure as much bi-partisan support for its electoral proposals as possible by reaching a consensus via the Joint Select Committee on Electoral Reform. As a consequence the electoral legislation enacted in 1983 and 1984 largely lacked partisan bias. However, the compromises necessary to reaching this consensus have resulted in an electoral system that still cries out for some improvement. Phillip Green (1986: 129)

Our democracy is diminished. Late last Friday, the Senate passed a new law that is a major step backwards for our electoral system…Electoral reform is needed, but it should encourage democratic participation and increase transparency, not frustrate these aims…The new law has an Orwellian title that asserts that it is an “Electoral Integrity” measure. In fact, it has undermined the integrity of our electoral system. George Williams (2006)

Win Government, Implement Electoral Reform

There have been two significant waves of electoral reform in the past 25 years at the Commonwealth level. In both cases a change of government has been the trigger for reform –– the election of the Hawke Labor Government in 1983, and the Howard Coalition Government winning office in 1996. Both of these governments sought to change electoral law upon coming to power. The Hawke Government was successful in

104 passing substantial reforms within a year of coming to office, thanks to minor party support in the Senate. In contrast, the Howard Government had to wait nine years until 2005, when the Coalition gained a Senate majority, enabling it to make its preferred changes the following year.

This chapter outlines some of the major aspects of each set of reforms, investigates the forces driving the reforms, and assesses the impacts they have had in terms of fairness and equity. I argue through this chapter that the propositions of cartelisation theory were to some extent supported by the evidence from the 1983–84 reforms, with benefits primarily shared between Labor and the Coalition. Significant, but lesser benefits accrued to the Australian Democrats, whose support was necessary for the reforms to pass the Senate. However, because the Howard Government’s control of the Senate obviated the need to act as part of a cartel, its 2006 changes promoted a distinct partisan advantage for the Coalition parties.

Labor’s 1983–84 Reforms

The election of the Hawke Labor Government in 1983 was the catalyst for the most significant changes to Commonwealth electoral legislation since 1948 (when the size of Parliament was substantially increased and a proportional representation system introduced for the Senate). The Joint Select Committee on Electoral Reform (JSCER) was established, and in September 1983 released its first report. The Committee’s Labor majority recommended important changes to the electoral system which were subsequently introduced and passed by the end of the year. Further reforms were implemented in 1984. Combined, these reforms created a new structural regime for administrators, legislators and political parties. Uhr (2000) summarises the six major reforms of this legislation as follows: • The establishment of the Australian Electoral Commission to administer the electoral system, and removing ministerial direction; • An increase in the size of parliament, from 123 to 148 House of Representatives’ members and from 64 to 76 Senators; • The provision of public funding for political parties and Independent candidates, when candidates receive at least four per cent of the formal vote;

105

• The introduction of ‘above-the-line’, or ticket, voting for the Senate, with the use of group voting tickets to determine parties’ preferences • The formal recognition of political parties, through a system of registration and identification on ballot papers; and • The removal of parliament’s veto power over the determination of House of Representatives’ electoral boundaries.

The previous chapter has dealt with the first issue –– that of independent electoral administration. The next three issues (the size of parliament, public funding and Senate ticket voting) are addressed in the following sections, while party registration matters are dealt with in Chapter 8. Hughes (2007) provides a thorough evaluation of the sixth issue –– independent boundary determinations –– with a particular emphasis on the history of redistributions in New South Wales and Queensland. As explained earlier, redistribution matters are not addressed in this thesis.

Two other reforms that occurred in 1983, not mentioned by Uhr, were an alteration in the counting rules for transferring vote surpluses in the Senate count, and the relaxation of the requirement for full preferential voting in the Senate. As these changes can be considered minor, especially compared to the six major reforms, they are not covered in detail here. However, the change in counting rules can have critical implications in a very close count, or where there is an unconventional pattern of below-the-line voting. Farrell and McAllister (2003; 2006) provide excellent accounts of the change in counting rules to a modified version of the Gregory method, and its implications.

In addition to the amendments mentioned above, the 1983 reforms also made changes to: • The close of polls (from 8pm to 6pm); • Enrolment and voting for Indigenous Australians (made compulsory); • The franchise (qualification was changed to Australian citizenship, with British citizens on the roll prior to 1984 retaining enrolment rights); • The regulation of political finance (a disclosure system was introduced for political donations and expenditure); • Closure of the roll (a statutory period of seven days was introduced before rolls were closed following the calling of an election);

106

• Postal voting (introduction of ‘general postal voter’ registration); • Ballot order (determined by lot rather than alphabetical listing); • Extension of the prisoner voting franchise (previously, prisoners serving sentences for offences punishable by imprisonment for one year or longer were ineligible to vote –– this was changed to offences punishable by imprisonment for five years or longer).

Postal voting processes are discussed later in this chapter, as are financial disclosure issues and the closing of the roll, in relation to the 2006 reforms. Issues relating to Indigenous voters and prisoner voting are discussed in Chapter 8.

The Size of Parliament –– The House of Representatives

As discussed in Chapter 2, one of the concepts of the fairness of electoral systems is equality of representation, sometimes defined as ‘one vote, one value’. Another aspect of fairness is that citizens have reasonable access to their elected representatives. Equality of representation and accessibility were put forward by the JSCER in 1983 as reasons why it recommended an increase in the size of parliament. On the first issue of representativeness, it was argued that due to the constitutional guarantee of a minimum of five House of Representatives’ seats for each State, Tasmania’s seats were malapportioned in their favour, and an increase in the overall number of seats would dilute this disparity (JSCER, 1983: 140). On the second issue, accessibility of members, the committee noted that the number of voters had grown considerably since the last significant increase in seats, which had occurred in 1949. As a result, a decrease in voters per electorate was required so that members could more effectively serve their electorates (JSCER, 1983:137).

An interesting question in relation to the 1983 increase in the size of the House of Representatives, from 125 to 148 seats, is whether it had an impact on the representativeness of the House in proportional terms. While acknowledging that results are influenced by voting behaviour, Table 6.1 indicates that there has been little, if any, impact on proportionality based on the increase in seats. Using Gallagher’s Least Squares Index (LSI), there was an average Index score of 10.98 for the six elections prior to the increase (1972–83), and a similar average score of 10.26 in the nine elections since the change (1984–2007). This is understandable as the reform was

107 primarily a correction for the growth in the number of voters enrolled in each electorate, rather than a change in the majoritarian, single-member voting system. Therefore, the single-member majority system, combined with the strong two-party system that exists across Australia, makes it difficult for the proportionality of the House to be significantly improved by providing opportunities for lesser parties that gain small but significant levels of support.

Table 6.1 Proportionality in the House of Representatives Election Major party Gallagher’s Least support % Squares Index 1972 92.1 6.96 1974 94.2 6.08 1975 95.9 14.22 1977 87.8 15.16 1980 91.6 12.94 1983 93.1 10.54 1984 92.6 7.87 1987 91.9 10.57 1990 82.9 12.94 1993 89.2 9.08 1996 86.1 11.23 1998 79.6 11.34 2001 80.8 9.80 2004 84.3 9.00 2007 85.5 10.50 Note: The smaller the Index figure, the more proportional the result. The method of calculating the Index, and the primary data used, is shown at Appendices C and D.

What is more significant in terms of the representative nature of the House is the increase in support for geographically-dispersed minor parties, for whom winning seats is extremely difficult under the single-member electorate system. Table 6.1 shows the levels of support for the major parties (Labor and the Coalition) prior to and throughout the study period. Although major party support has recovered in recent elections from its low of 79.6 per cent in 1998, it remains significantly lower than the levels experienced in the 1970s and 80s. This is largely a reflection of voting behaviour rather than the impact of reforms. As Western Australian political scientist David Charnock notes, the impact of the increase in the number of seats has diminished over time as enrolment numbers have steadily increased. By comparison, Charnock argues that the redistribution methods introduced in 1983 are possibly a more significant factor in terms of any partisan bias from the reforms (1994: 498–9).

108

The Size of Parliament –– The Senate

Section 24 of the Australian Constitution requires that for every seat in the House of Representatives there ‘shall be, as nearly as practicable, twice the number of senators’. An increase in the size of the House of Representatives requires an increase in the size of the Senate to retain the approximate 2:1 ratio. The 1983/84 reforms therefore included an increase in the size of the Senate from 64 to 76 seats. As the Senate is based on a proportional representation voting system, an increase in seat magnitude has the potential to impact on the ability of minor parties to win seats, and therefore on the proportionality of representation. The increase in seats resulted in a reduction in the quota required to win a State Senate seat in an ordinary half-Senate election from 16.7 per cent (for five seats) to 14.3 per cent (six seats). For the less-common full Senate elections, the quota reduced from 9.1 per cent (10 seats) to 7.7 per cent (12 seats). The impacts of this change on proportionality are shown in Table 6.2.

Table 6.2 Proportionality in the Senate Election Major party Gallagher’s Least support % Squares Index 1970 80.4 3.18 1974 1 91.2 4.23 1975 1 92.6 3.34 1977 82.3 8.49 1980 85.7 2.05 1983 1 85.5 3.94 1984 2 81.7 5.71 1987 3 84.8 2.93 1990 85.4 5.01 1993 87.3 3.54 1996 80.2 5.53 1998 75.0 7.24 2001 76.1 8.45 2004 80.1 8.33 2007 80.2 6.05 1 Full Senate election –– 10 seats per State. 2 Seven Senators elected from each State to increase the overall size of the Senate to 76. 3 Full Senate election –– 12 seats per State. Note: Primary data is shown at Appendix D.

It can be seen from these figures that proportionality has actually decreased since 1983, although there was greater proportionality in the only double-dissolution election since the increase in seats –– 2.93 in 1987 –– than the average of 3.84 in the three double- dissolutions in the decade prior to the change (in 1974, 1975, and 1983). The most obvious impact has been the decreased proportionality in the four most recent elections

109

(1998–2007), a period that has also seen the highest non-major party vote in history. The voting figures (at Appendix D) suggest that the increased disproportionality is a combination of the major parties becoming more over-represented, and some minor parties winning significant support without winning seats. In fairness terms, the reduced quotas required to win Senate seats improves the ability for the Senate to be more representative of voters’ choices. However, while the reform is an improvement on the fairness of election outcomes, voter behaviour (that is, who voters actually choose) has resulted in less proportional results.

As for the reasons behind the increase in the size of the parliament, current ACT Electoral Commissioner Phillip Green noted in 1986 that a major concern of the JSCER was how an increase in the Senate size would impact on the possibility of a major party having control of the Senate. Green suggests that the committee may have preferred to increase the Senate to 88 (14 Senators per State, plus four from the Territories) to maintain the pattern of each State electing an odd number of Senators at half-Senate elections (seen by some as advantageous for the major parties). However, such an increase was considered to be too extreme, especially with the 2:1 requirement for the House of Representatives (1986: 96). Green goes on to argue that the real impetus for reform came from the National Party, which had suggested the increase to the JSCER as a way to consolidate and possibly expand their House of Representatives seats in rural Australia. Meanwhile, the Labor Party thought such an increase might be a way to reduce the impact of the Democrats (1986: 98), who had gained the balance of power in the Senate in 1981. The reforms were passed by Labor with Nationals support, with the Liberals and Democrats opposing the measure.

Public Funding

The importance of public funding for the fairness of funding regimes is suggested by the Democratic Audit assessment question, …is there fair access for [candidates and parties] to the media and other means of communication with the voters (question 2.1.3, see Table 3.2). With a few exceptions, the public funding of parties’ and candidates’ participation and campaigning in elections is a relatively recent democratic initiative, occurring in Australia since the early 1980s.

110

The arguments for public funding put forward by American political scientist Michael Johnston is that some level of funding encourages electoral competition, and strengthens parties both organisationally and for campaigning purposes (2005: 9). Sally Young and Joo-Cheong Tham contend that public funding can provide a way to introduce accountability measures, such as financial reporting and disclosure (2006: 37). Some parties, particularly those in government, are able to attract substantial private funding. Public funding provides a mechanism to cover basic administration and campaign costs for all parties, based on criteria such as registration, meeting reporting deadlines, voting support, and party membership. Johnston goes on to argue that while partial public funding is beneficial for the health of a democracy, full public funding can be counter- productive, introducing disincentives for citizen mobilisation and the possibility of corruption (Johnston, 2005: 15).

The introduction of public funding at the Commonwealth level in 1983 changed the nature of electoral competition by providing a guaranteed source of income for parties and candidates who achieve at least four per cent of the formal vote. When introduced ahead of the 1984 election, funding was based on the reimbursement of election campaign expenditure, with the rate of funding (per vote) based on the cost of a postage stamp for each of the three years between elections (JSCEM, 2005a: 313), and indexed to the rate of inflation. This amount (originally 90 cents) was then divided, with two- thirds for House of Representatives votes and one-third for Senate votes. Parties and candidates had to submit expenditure receipts for reimbursement from their public funding entitlement. The application of different rates for the House and the Senate was explained by arguing that MPs experienced greater constituency demands than Senators. However, it potentially disadvantaged those parties that targeted Senate elections and thereby gained greater numbers of Senate votes than House votes.

In its first report in 1983, the JSCER stated that the Labor party and the Australian Democrats were supportive of a public funding system. Labor’s support was based on the argument that elections should be decided on the quality of policies rather than the ability to raise campaign funds. Labor went on to argue that public funding would ‘narrow the differential in the financial resources available to the various competing parties’ (JSCER, 1983: 148). The Liberal and National parties were initially opposed to public funding, citing the lack of public support, and that public funding would entrench incumbent parties and disadvantage potential new entrants (JSCER, 1983: 150).

111

In 1995, two major changes occurred to the public funding regime, following recommendations by the JSCEM. The Commonwealth Electoral Amendment Act 1995 removed the need to prove expenditure, thereby changing the nature of funding from a reimbursement to an entitlement. This change broke the direct connection between funding and election campaign costs. Secondly, the funding rate was substantially increased. There was a 50 per cent increase for House of Representatives votes from $1.01 to $1.50 and, in a move to equalise funding between House and Senate votes, public funding for Senate votes was trebled from 50.4 cents to $1.50. The rationale put forward during the parliamentary debate for the Senate increase was based on the JSCEM recommendation –– that House and Senate votes should receive the same level of funding, because ‘as much effort is required to gain a Senate vote as a House of Representatives vote’ (JSCEM, 1994: 9). However, while the JSCEM had recommended equal funding, it left unanswered the question of whether funding levels should be increased or simply redistributed equally between the two houses.

The Hansard records show that both the Labor and Liberal parties strongly supported the changes. In his second reading speech, the Labor Minister for Administrative Services, Frank Walker, referred to JSCEM’s rationale for equalising House and Senate payments, but made no reference to the fact that the rate was being substantially increased, and did not provide a supporting argument for the increase (Walker, 1995). In a similar vein, Liberal MP Peter Slipper simply stated, ‘The rate will be increased to $1.50 per vote and will be indexed. The coalition will not be opposing this’ (Slipper, 1995). Apart from Labor and the Coalition, the Australian Democrats were the most significant beneficiaries of the increase. Their spokesperson, Senator Meg Lees, made no reference to the increase in her second reading speech.

In contrast, the Greens received far less public funding during this period, and were more vocal in their opposition to the increase. Greens spokesperson, Senator Christabel Chamarette, stated that: The amendments to the bill referring to public funding which the government moved in March were not recommended by the joint committee…they were part of a private deal which appeared to be engineered by the Federal Secretary of the ALP and agreed to in letters to the Minister for Administrative Services from the Liberals, the Nationals and the Democrats…Everybody is going to have at least a doubling of their funding. (Chamarette, 1995).

112

As a result, the cost of public funding has substantially increased since the 1996 election, as Table 6.3 shows. Due to the funding rate being linked to the Consumer Price Index, the rate has increased from $1.50 per vote in 1995, to its current (July to December 2008) rate of $2.19.24

Table 6.3 Public Funding payments ($million) 1984 1987 1990 1993 1996 1998 2001 2004 2007 7.81 10.30 12.88 14.90 32.15 33.92 38.56 41.93 49.00 Sources: JSCEM, 2005a: 316–7; AEC.

By its very nature, the public funding scheme ensures that the major vote-winners at an election are the major funding recipients. Currently, the major parties –– Labor, Liberal and the Nationals –– can be relatively assured of a certain amount of public funding, as can the Greens (and previously the Democrats), to a lesser extent. For parties, particularly new parties, it is not easy to predict whether the four per cent threshold in votes, which is required to receive any public funding, will be achieved. It is difficult therefore for such parties to budget for an election campaign. Table 6.4 shows the level of funding received by selected parties and candidates over the past three elections.

Table 6.4 Public funding payments –– selected parties and candidates ($million) Party / Candidate 2001 2004 2007 Labor 14.917 16.710 22.030 Liberal 14.492 17.956 18.134 Nationals 2.845 2.967 3.240 Democrats 2.412 0.008 0.000 Greens 1.594 3.317 4.371 Family First –– 0.158 0.141 Country Liberal Party 0.139 0.159 0.169 Pauline Hanson’s One Nation 1.170 0.056 –– Pauline’s United Australia –– –– 0.213 Peter Andren MP 0.073 0.079 –– Bob Katter MP 0.064 0.064 0.068 Tony Windsor MP 0.064 0.090 0.111 Nick Xenophon –– –– 0.312 Other 0.789 0.418 0.435 Total 38.559 41.926 49.003 Source: AEC.

24 Current rates of funding are provided on the AEC website at .

113

It can be seen from the table that for most parties public funding provides a relatively reliable income stream. For parties that experience greater fluctuations from election to election, such as the Democrats, a drop in voter support can have serious financial implications in planning an election campaign. In the Democrats’ case, their vote dropped for the House of Representatives from 5.41 per cent (2001), to 1.24 per cent (2004) and then 0.72 per cent (2007). For the Senate, the Democrats’ vote decreased from 7.25 per cent (2001), to 2.09 per cent (2004), and then 1.29 per cent (2007). The amount received by parties or candidates without parliamentary representation is quite small, as can be expected. This may indicate that the public funding scheme is simply supporting existing parties, rather than new entrants. However, the growth of the Greens, One Nation, and Family First since 1983 is a counter-argument to this proposition. The numbers of parties and candidates receiving public funding since 1993, shown in Table 6.5, suggests that there may be a trend to fewer parties receiving support, but further analysis would be required to determine the reasons for this decline.

Table 6.5 Number of Parties and Independents Receiving Public Funding 1993–2007 1993 1996 1998 2001 2004 2007 Parties 11 8 12 13 10 7 Independents 17 12 16 21 15 15

Although the public funding rate was originally calculated on a three-year funding cycle, by-elections held between general elections also attract funding. The cost of this additional funding is currently about $150,000 per by-election (depending on the size of the electorate and results). The by-elections held in the seats of Mayo and Lyne in September 2008 had a combined cost of almost $299,000 in public funding. The substantial cost raises the question of whether public funding should be provided for by- elections, especially where by-elections are caused by the voluntary resignation of an MP, as was the case in these two examples.

In sum, public funding has produced mixed results in terms of fair access to communicate with voters. While the major parties are the significant beneficiaries of the regime, smaller parties have been able to compete and grow, with public funding an important component of their financial support. However, the primary driver behind public funding, both in its implementation and in subsequent amendments, is partisan self-interest. As one Labor politician commented in an interview, both the Liberal party

114 and the Labor party could see that they were going to pick up an enormous amount of taxpayers’ money for nothing .

Senate Ticket Voting

A significant reform flowing from the JSCER’s first report was the adoption of ticket voting for the Senate. JSCER recommended this change on the basis of the high Senate informal voting rate, which was primarily due to voters having to number all boxes on the ballot paper. The alternative proposed by JSCER was to give voters a simpler option of numbering one box for a party, and allowing the party to determine where the voter’s preferences would flow, based on a ‘ticket’ previously lodged with the AEC by the party (JSCER, 1983: 64).

Another alternative way to reduce the informality rate was to introduce optional preferential voting; however this was rejected due to partisan considerations. The Liberal party had a philosophical objection to optional preferential voting and the Democrats, whose support in the Senate was needed to pass the reform, was also opposed (Green, 1986: 90–3). It was likely that the Democrats opposition was largely based on the benefits the party could receive in negotiating preferences with other parties –– a significant electoral advantage for a party that rarely gained sufficient primary votes to win Senate seats.

Table 6.6 Senate Informal Voting Rate (%) 1961 1964 1967 1970 1974 1975 1977 1980 1983 10.6 7.0 6.1 9.4 10.8 9.1 9.0 9.7 9.9 1984 1987 1990 1993 1996 1998 2001 2004 2007 4.7 4.0 3.4 2.6 3.5 3.2 3.9 3.8 2.6 Sources: AEC; AGPD.

Introduced for the 1984 election, group ticket voting (or above-the-line voting), had an immediate and ongoing impact in reducing levels of informal voting. The average rate of informality has dropped from 9.1 per cent for the nine elections prior to the change, to only 3.5 per cent in the nine elections since the change (see Table 6.6). Levels of informal voting across both Federal houses and other jurisdictions are discussed in Chapter 8.

115

The use of above-the-line voting has steadily increased since the 1984 election, from 85.7 per cent in that year, to 96.8 per cent in 2007 (see Table 6.7). This is understandable as voters become more familiar with the system and, as time goes on, a greater proportion of voters are those who reached voting age after 1983 and have therefore known no other Senate voting system. In addition, the political parties encourage voters to vote above the line. Particularly for minor parties, a high above- the-line vote is useful in negotiating preferences with other parties, as the party is able to dictate the flow of preferences. However, it is the major parties, Labor and Liberal, who tend to tell people to ignore the possibility of voting below the line. This is illustrated by a Liberal how-to-vote card (Figure 6.1), that gives the specific instructions ‘Do not mark any other box’ and ‘No need to fill in this section’.

Table 6.7 Senate Group Voting Ticket Usage (%) and Number of Candidates (in bold italics), 1984–2007 1984 1987 1990 1993 1996 1998 2001 2004 2007 NSW 81.2 83.8 94.3 96.2 95.8 96.2 97.1 97.6 98.2 40 50 62 66 63 69 65 78 79 Victoria 90.8 90.4 90.2 95.1 95.4 96.8 96.8 97.7 98.0 36 52 38 51 44 63 52 65 68 Queensland 86.3 88.0 91.0 93.6 95.6 95.8 95.5 94.8 97.3 28 36 33 42 48 57 40 50 65 Western 85.1 88.1 91.8 95.0 94.2 94.2 94.7 95.8 97.3 Australia 28 45 30 33 29 45 46 40 54 South 92.6 92.9 93.1 94.7 93.9 92.9 92.2 94.1 93.2 Australia 37 31 26 36 31 35 26 47 46 Tasmania 70.0 68.3 77.1 78.9 72.6 78.2 80.4 81.2 84.2 16 21 16 20 19 32 29 26 28 ACT 64.3 69.2 76.6 84.4 83.5 78.2 77.9 79.1 82.8 10 11 12 14 14 17 14 13 16 Northern 74.7 77.4 77.3 79.0 81.4 85.6 90.3 89.7 92.1 Territory 7 9 6 4 7 11 13 11 11 Australia 85.7 86.7 91.4 94.4 94.4 94.9 95.2 95.8 96.8 202 255 223 266 255 329 285 330 367 Sources: Newman, 1999; AEC.

116

Figure 6.1 Liberal Party How-to-Vote Card, 1998

117

Parties such as the Greens and Democrats have typically encouraged voters to decide how to vote, by explaining how to formally vote below the line (see Figure 6.2) or showing a full list of preferences (as per their group ticket, see Figure 6.3). Some parties, who may have negotiated a favourable preference exchange with non- likeminded parties, typically will be reluctant to publicise where their party preferences are being directed.

Figure 6.2 Australian Greens How-to-Vote Card, 1996

118

Figure 6.3 Greens (WA) How-to-Vote Card, 1998

119

Table 6.7 details the use of ticket voting across the nine jurisdictions since its introduction in 1984, as well as the number of candidates contesting each election. Noticeable in the table is the fact that Tasmania, the Australian Capital Territory, and (to a lesser degree) the Northern Territory have significantly lower rates of above-the- line voting in comparison to the other jurisdictions. Newman (1999) suggests three factors that may influence these differing rates. First, the differences could be related to the number of candidates contesting an election, the argument being that if fewer candidates contest, it is less onerous to fill out boxes below the line. Second, the use of the Hare-Clark system at the State or Territory level (in Tasmania and the Australian Capital Territory), where voters are required to preference all candidates, might also encourage below-the-line voting at Senate elections. Finally, the use of ticket voting at the State level (currently in place in New South Wales, Victoria, Western Australia and South Australia), may encourage above-the-line voting at the Federal level.

When figures for the number of candidates contesting Senate elections are considered, it can be seen that the use of the Hare-Clark system of proportional representation in Tasmania and the ACT may have a more significant effect on ticket voting levels than, as in the Northern Territory, having a small number of candidates,. The ACT’s combination of the Hare-Clark voting system and low candidate numbers appears to be behind the fact that the ACT has the lowest rate of ticket use in the country. It can be argued that the current requirement for voters to number at least 90 per cent of all boxes below the line for a vote to be formal disadvantages electors in jurisdictions with many candidates. It therefore becomes relatively more feasible to vote above the line in those jurisdictions, which empowers parties to direct preference flows according to group voting tickets –– this disempowers electors, and runs counter to the principle of equal opportunity to participate.

National Voters Voting Liberal A good example of voter disempowerment can be observed in the results of the 2004 Victorian Senate election, in which 97.7 per cent of electors voted above the line. The Liberal-National Coalition ran a joint ticket, on which the only National candidate was placed in the number two position. More than 99 per cent of Coalition supporters in Victoria voted above the line. This meant that by voting above the line, many National party supporters actually voted for a Liberal candidate ahead of a National candidate. The results can be seen in the seat of McMillan, which is the only Federal House of

120

Representatives electorate in Victoria where there were both Liberal and National candidates standing for a seat.

Table 6.8 shows the voting results for both the Liberal and National candidates in the House of Representatives seat of McMillan, and the results for the Senate votes cast in the electorate. Assuming that each Coalition party lost a proportion of voters in the Senate election equivalent to the drop from their combined House of Representatives vote, then it could be expected that there would be 6,472 National supporters in the Senate figures.25 However, only 66 National voters took the trouble to number at least 59 boxes below the line to formally vote for the National party candidate, Julian McGauran, who was placed second on the Coalition ticket. Although McGauran was considered to be in a safe winnable position on the ticket, a fact that National supporters may have considered, this result typifies the ability for the governing parties to devise a system that strengthens the parties’ influence –– by choosing where they place candidates on a list, as well as directing preferences to other groupings –– ahead of empowering voters.

Table 6.8 2004 Federal Election Results for the Coalition –– McMillan, Victoria HoR Senate ATL Senate BTL Party Votes % Votes % Votes % Liberal 32,379 42.93 218 0.29 37,908 49.83 National 6,676 8.85 66 0.09 Source: AEC.

Reform Attempts –– 1996–2004

When the Howard Liberal/National Coalition Government was elected in March 1996, it moved quickly to make its mark on electoral administration. One of the Howard Government’s first actions was to abolish the Aboriginal and Torres Strait Islander Election Education and Information Service (ATSIEEIS). However, for nine years the Government was thwarted in its efforts to make substantial changes to electoral legislation because it did not have a majority in the Senate. During this period, the Government made clear its intention to ‘tighten’ enrolment procedures and to relax the

25 The combined Coalition vote dropped from 51.78% for the House of Representatives to 50.20% for the Senate. This is a decrease of 3.05%. The National House of Representatives candidate received 6,676 votes, which equates to an expected 6,472 votes for the Senate (6,676 less 3.05%).

121 disclosure requirements for party donors. Until 2006, the Government had only been successful in making minor modifications to legislation, such as reducing the prisoner franchise, and changing a redistribution formula to ensure the retention of the two Northern Territory House of Representatives seats. Both of these changes occurred in 2004 and both required the support of the Labor Party.

The ATSIEEIS Case

The abolition of the ATSIEEIS in 1996 is a good example of a government impacting directly on the ostensibly independent operations of an electoral management body, and potentially providing partisan advantages to the governing parties. The ATSIEEIS was a unit of the AEC which had been established in 1979 to target remote and urban Indigenous communities, with an ongoing focus on encouraging enrolment (Alport and Hill, 2008: 5). In the nearly two decades for which it operated, the unit was seen as instrumental in improving the rates of Indigenous enrolment and voting, especially in remote areas. The value of ATSIEEIS was explained by one commissioner: They used to rip these poor Indigenous people off the roll, send out an objection letter, wouldn’t matter if you couldn’t read the bloody thing, you’re off the roll. In the early 80s they were ripping them off the roll. It was a shocker. Then ATSIEEIS came along and started rectifying a lot of the stuff to quite a meaningful extent, but then when ATSIEEIS stopped, the capacity to send people out with non-voters notices also stopped…Without ATSIEEIS there, we’ve had this build-up…on the roll, because no-one can actually say where that person’s gone –– they could have gone across the border.

It is argued that since the abolition of the unit, enrolment rates for Indigenous people have dropped markedly (Snowdon, 2004). This is seen as a distinct advantage for the Coalition, due to the high level of support for Labor in remote Indigenous communities. As one administrator stated: The issues are –– we’ve got a high Indigenous rate, we’ve got a lot of people living in remote areas, it’s a very young population, we’ve got a highly mobile population –– so the more people you get on the roll, the more people you educate to participate in a meaningful way in elections. [It] would tend to advantage the ALP. That’s not the reason we do it.

The Howard Government’s move to abolish this unit was the first time that the AEC had been given an explicit instruction about how it may spend its budget. This instruction therefore raises the issue of whether such an intervention is a challenge to the AEC’s independence. The impact of removing the ATSIEEIS program is summarised by two electoral commissioners as follows:

122

Taking away ATSIEEIS’s program in ’97 had a significant impact on the delivery of education services and enrolment potential…What ATSIEEIS was doing was streamlined towards the end and you go out there and people remember it, so it must be having some impact. In ten years you can actually see how it’s affecting the roll.

The best example of political interference in funding of electoral authorities is the fate of the Commonwealth’s Aboriginal electoral education scheme. It was a separate line item in the budget and they said ‘no, we’re not giving you that any more, and you’re not allowed to spend any other part of your budget on it. If you do, we’ll take it off you’. I think that’s entirely inappropriate and it should never have happened.

It was confirmed in interviews that if the AEC wished to reinstate a program such as the ATSIEEIS , it would not do so independently, but instead would make a request to the government to fund what is considered by the AEC as a politically sensitive program. As one administrator said: If there’s sensitivity about it, then you go through the Minister’s office…given that it was a program that was abolished by government, then yeah [it was politically sensitive]. No matter how independent you are, everybody is dependent upon somebody for money .

The AEC’s dependence on government approval to undertake a program that could improve the accuracy and comprehensiveness of electoral rolls diminishes the independence of the electoral administration. That such a program is considered politically sensitive indicates that decisions may not be made in the interests of best electoral practice, but in response to the political leanings of government. The correlation between voter turnout and the proportion of Indigenous people in an electorate, and the reasons for Indigenous people’s apparently low turnout, are discussed in Chapter 8.

The Coalition’s 2006 Reforms

At the 2004 election, the Coalition was able to win control of the Senate, taking effect from 1 July 2005. With a Senate majority, the Coalition was able to implement its long- standing electoral policy agenda. The Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Act 2006 contains many measures that the Coalition had first proposed in 1996 and consistently pursued since then, both through legislation and

123 through the JSCEM. Following two committee inquiries and 30 hours of parliamentary debate, the legislation was passed, virtually unamended, in late June 2006.

The following sections assess some of the more notable aspects of the 2006 legislation: the earlier closure of the electoral roll; more stringent proof-of-identity provisions (both for enrolment and provisional voting); an increase in the donation disclosure threshold; and an increase in the level of tax-deductible political donations. Two other important aspects of the legislation, full prisoner disenfranchisement and the deregistration of parties, are analysed in Chapter 8. Combined, these reforms have serious implications for the principles of fairness promoted by Mercurio and Williams (2004) and others –– the opportunity for voters and candidates to participate equally; and the ability for voters to make an informed choice, through knowledge of candidates and parties.

Earlier Closure of the Electoral Roll

Ease of enrolment is an important element of access and equity for voters in the electoral process. Early closure of electoral rolls has the potential to impact in a negative way on such access, particularly for people who are in the process of moving location or are living in a remote area, and for young people wanting to enrol for the first time. Australia does not have fixed-date elections at the Federal level. The Prime Minister is able to call an election, within a specified time frame, without notice. The writs for the election are issued shortly thereafter. In February 1983, Prime Minister Malcolm Fraser called an election for which the rolls closed ‘almost immediately’ after the announcement (JSCER, 1983: 110). The combination of an election being called with almost no warning and the sudden closure of the rolls meant that many citizens who would normally be eligible to vote were disenfranchised. As a result of this, the JSCER recommended that the rolls remain open for at least seven days following the formal announcement of an election. This seven-day period was duly brought into legislation in late 1983. 26 During the seven days, citizens were able to newly enrol, or update their enrolment if they had moved address or changed their name.

26 See section 45 of the Commonwealth Electoral Legislation Amendment Act 1983 , which created a new section 61A in the Commonwealth Electoral Act 1918 .

124

As a result of the 2006 legislation, however, the electoral rolls now close for new enrolments at 8pm on the day that writs are issued,27 and three business days after the issuing of the writs for enrolment changes. The Government’s stated justification for this change was that it would enhance the integrity of the roll by preventing people from making fraudulent enrolments after an election is called. The Government argued that by closing the roll earlier, the AEC would have additional time to verify new and updated enrolments. The Government majority on the JSCEM recommended this change, with one of its arguments being that significant numbers of voters delayed updating their enrolment when legally required, waiting for the calling of an election to be prompted into action (JSCEM, 2005a: 35). Labor members of the Committee argued otherwise, stating that the proposed change would reintroduce the problems that had occurred with the sudden close of rolls in 1983, when many potential voters were turned away at polling booths because they had been deleted from the roll, or were enrolled at the wrong address (JSCEM, 2005a: 360). A State Labor MP clearly stated the argument: It is a travesty of our democratic system that the Commonwealth committee pushed these changes through. It is nothing more than a deliberate attempt by a Liberal conservative government to disenfranchise further all of those people who may be Labor voters.

A number of academics and human rights activists also mounted campaigns against this reform. For example, former Australian Electoral Commissioner Colin Hughes and Melbourne academic Brian Costar were vocal in opposing the change. Criticising the JSCEM arguments that the change would prevent fraud and ensure proper scrutiny of enrolments, Hughes and Costar argued that there was minimal evidence of fraud, and that the Australian National Audit Office had reviewed the roll and found it to have ‘high integrity’. They also noted that the AEC had confirmed that all enrolments were equally scrutinised, irrespective of timing (Hughes and Costar, 2005). The Democratic Audit of Australia, based at the Australian National University, also promoted debate on the reform (for example, see Sawer, 2006b).

At the 2004 election, during the seven-day period following the issuing of the writs, the AEC processed 78,816 new enrolments and 345,159 updated enrolments (AEC, 2004). A further 150,000 people applied after the seven-day period, but prior to the election,

27 There are exceptions for 17-year-olds who will turn 18 before election day, and for people who are to be granted citizenship before election day. People in these categories are to have three business days from the issuing of the writs to enrol.

125 and therefore were restricted from being placed on the roll, or having their enrolment amended, in time for the election (Campbell, 2006: 5). The Howard Government provided the AEC with additional funding for a public awareness campaign to alert people to the earlier cut-off date. As a result, for the 2007 election, there was a combined total of 279,469 new enrolments and enrolment updates between the calling of the election and the close of the roll (AEC, 2008: 8), a reduction of 34 per cent from the 2004 figures.

The AEC has adopted varying positions on this issue. In a submission to the JSCEM in 2000, it noted that neither it nor the JSCEM had uncovered any organised or widespread attempt to defraud the roll through false enrolments in the previous 15 years (AEC, 2000). The AEC went on to state that the proposed early closure of the rolls would not improve the accuracy of the rolls for an election, and in fact the rolls would be less accurate, as electors would not have enough time to correct their enrolments or to apply for new enrolments. The AEC also stated that the change would have a negative impact on the franchise, that it would delay election results due to an anticipated rise in declaration voting, that it would have a particularly negative impact on young people seeking to vote for the first time, and that it would cause public confusion due to different systems operating at the State and Territory levels.

During questioning before the committee in 2005, the AEC reasserted the above position. At a hearing in March 2006, however, the newly-appointed AEC Commissioner, Ian Campbell, stated that the earlier closure of the roll would not create any difficulties, and would actually ‘make our life easier’ (Campbell, 2006: 8). Under new leadership, the AEC’s focus appears to have switched from accuracy at election time to ongoing accuracy between elections. The change is seen to favour the Coalition electorally, with Australian Election Study figures from 2004 showing that only 43 per cent of young people (under 25 years, the age cohort with the lowest proportion of eligible citizens enrolled) supported the Liberal and National parties, while 49 per cent supported the left-wing parties (Labor –– 32%, Greens –– 17%) (Bean, 2005).

Proof of Identity Requirements

The 2006 legislation introduced proof of identity requirements, both for enrolling and provisional voting. In both cases, citizens are required to provide a driver’s licence or

126 other identifying documentation. For enrolments where the person does not have any documentation, the application needs to be witnessed by two electors who know the person. The provisions add a burden of proof that does not exist for State or Territory enrolments or voting, and can be expected to be particularly onerous for people living in rural and remote areas. The JSCEM recommended these changes in its 2005 report, on the grounds that it was a reasonable measure to counteract possible fraud. The committee repeated the argument of earlier reports that enrolling to vote should not be subject to less scrutiny than ‘joining a video library’ (JSCEM, 2005a: 22–3). The Labor minority on the committee argued that the changes were unnecessary, due to the lack of any evidence of significant levels of false enrolments or electoral fraud (JSCEM, 2005a: 358–9).

The reform requiring proof of identity is seen to favour the Coalition parties (for example, see Brent, 2007). Although it is currently too early to accurately measure the impacts of these changes, evidence from interviews suggests that the reform will produce administrative difficulties. For example, different levels of proof-of-identity requirements now exist between the Federal, and State and Territory levels. This will create serious problems for the bilateral joint roll arrangements between the Commonwealth and other jurisdictions. Prior to the Commonwealth’s reforms outlined above, the joint roll agreements provided for the administration of one roll which could be used for Federal and State or Territory elections, due to identical enrolment procedures. After the change, citizens can enrol at the State or Territory level and, unless they provide proof of identity, will not be enrolled at the Commonwealth level. This is leading to the creation of separate rolls and increased administration costs, as well as possible confusion for voters. One electoral official commented: ‘ We’ve got a few State-only at the moment, an increasing number ’. A State parliamentarian put the position more strongly, stating: It’s clear to me now that when I study the accumulated monthly roll that comes in for my electorate once a month. And it’s clear that something like 20 per cent of the people who are enrolling since the changes, are enrolling as State-only electors.

One indicator that the proof-of-identity requirement may be disenfranchising voters is the rate of provisional vote rejections at the 2007 election. In its 2008 submission to the JSCEM, the AEC states that 27,529 voters lodging provisional votes did not provide

127 proof of identity during the specified six-day period after the election, and were therefore disenfranchised. The AEC has expressed concern over this disenfranchisement and has recommended that the proof-of-identity provision for provisional votes be repealed (AEC, 2008: 46–9).

State and Territory electoral officials also consistently expressed a view that the Federal Government did not enter into genuine consultation with the other jurisdictions, creating tensions between the State and Territory commissions and the AEC. The following comments from four Commissioners indicate the break-down in Commonwealth– State/Territory relations that resulted from the manner in which the provisions were implemented: I don’t know whether it’s a Commonwealth–State thing, or a product of size, or people not thinking, but if you’re in a partnership, you must see it as an equal partnership, and therefore you have an equal say. But this whole thing was heading down the way of we have to invoke the mediation clause of the joint roll agreement, and that wasn’t good.

Whenever the Commonwealth does something [the jurisdiction] might not like, [the jurisdiction] either has to break the nexus and enact something different, and stuff the voter around, or they lump it, and this time they’ve lumped it.

Well, it’s interesting you describe [the proof of identity provisions] as a reform. Some people might regard it as –– you think of a reform as actually improving things.

I think [the Federal commissioner] boxed himself into a position of saying we’ll make it work, and I think he’ll have a lot of trouble in making it work, because the impact is there for all to see. In fact, by putting those hurdles up, enrolment has dropped.

Contrary to the policy of voters having to initiate the enrolment or enrolment update process with proof of their identities, several of the State and Territory commissioners are supportive of a move towards ‘direct update’ of the electoral roll, where enrolments are automatically updated when the commission becomes aware that a citizen has changed address (through motor registry or similar records), or has reached voting age. As one commissioner described it: People who are on the roll and we know they’ve moved, we should write to them and say, you’re on the roll at this address now. We get downloads saying when kids are turning 17, and when they turn 18. If we can match them with a birth or something else, it’s my view that we should proceed with the research to put them directly on the roll, and write to them and say you’re on the roll, send us in a signature

128

This suggests that it will be up to the commissioners to convince their governments to make the legislative changes to allow direct update of enrolments to occur. As the legislation stands, the commissions can contact citizens who they believe should enrol or update their enrolment, but it is then up to the citizen to take action (filling out an enrolment form) before the roll can be updated.

The changes requiring proof of identity reflect the general emphasis in the legislation on transforming the traditional ‘responsibility’ to vote into a ‘right’ to vote. Rather than having electoral legislation that facilitates ease of voting, citizens are now increasingly challenged to justify their entitlement to vote. This shift has occurred in the absence of any evidence of systemic or regular cases of fraudulent voting or enrolment. Unlike the situation for citizens in countries with a voluntary voting system, where there is a greater emphasis on encouraging enrolment and turnout, uninterested Australians may see these more stringent identification requirements as a chance to opt out of the compulsory system.

Peter Brent and Simon Jackman (2007) argue that because of these requirements, it is now easier to be removed from the roll than to be placed on it, resulting in a ‘shrinking’ roll. The barriers to enrolment run counter to the trend in many democracies, where the emphasis is on making it easier for people to enrol. In some countries, this is done through ‘automatic enrolment’, where electoral management bodies use information from other sources (such as driver licensing) to place people on the roll. People enrolled then have the choice to ‘opt out’ of the system, rather than opting in. Hughes and Costar (2005) cite this reform as a reminder of the tenuous nature of citizens’ rights in Australia, where virtually no guaranteed rights are contained in the Constitution in regard to the franchise.

The processes involved in reforming the close of roll and proof-of-identity laws run counter to internationally accepted norms for electoral management and fair elections. The views of the ‘independent’ AEC and of State and Territory commissions have largely been ignored in pursuit of the governing parties’ agenda. This agenda has clear partisan benefits for the Coalition, and it appears that the AEC in particular was ‘captured’ by the Howard Government.

129

Donation Disclosure Thresholds

As mentioned earlier in this chapter, public funding schemes provide an opportunity to introduce accountability and transparency regimes. This is regarded as part of the social contract –– if parties and candidates are to receive public money for campaigning purposes, they need to disclose to the public the amounts and sources of private money donated towards the same purpose. In 1983, the Hawke Labor Government introduced donation disclosure legislation, requiring donors and political parties to disclose donations above a specified amount. The threshold for disclosure was originally set at $1,000, and increased to $1,500 in 1991 (JSCEM, 2005a: 328). The Howard Coalition Government’s 2006 legislation increased the disclosure threshold from $1,500 to $10,000, and introduced annual indexation based on the Consumer Price Index. 28 From the time the Coalition obtained a majority on the JSCEM in 1996, the committee has recommended increases to the threshold –– of $5,000 in 1996 and to $10,000 following the 2004 election. In its submission to JSCEM, the Liberal party noted that it would not be realistic to expect that amounts below $10,000 could create an undue influence on government (JSCEM, 2005a: 329). The Government argued that it is not in the public interest for donations below this amount to be disclosed. The Labor party strongly opposed the increase to the threshold.

Individual but related entities (such as members of a family, or directors of a company) are treated separately under the disclosure laws, as are State and Territory divisions of the same party. For example, a husband and wife could separately donate $10,000 to each of the nine State, Territory, and Federal divisions of a party, a total of $180,000, without any of the donations being disclosed. Although the ability to donate separately was already in the Act, the increased threshold makes this option more attractive to major donors. Countries such as Canada have managed to place stringent caps on total party donations, so the provision for separate donations appears to be a deliberate legislative loophole that advantages Australian parties with a federal administrative structure.

Miskin and Baker (2006) highlight the impact that the increased threshold could have in monetary terms. In the 2004–05 financial year (an election year), $33.1 million in donations to the major parties was disclosed under the existing $1,500 threshold. That

28 Days after the legislation was passed by Parliament, the threshold had already increased to $10,300.

130 figure would drop to $25.2 million under the new regime. That is, an additional $8.1 million would be removed from public accountability, assuming that donating patterns remain the same. It can be expected however, that political parties will adapt their fundraising and donation strategies, and donors who previously donated up to the old cap will be encouraged to donate up to the new threshold. The interaction of the increased threshold with the approval of foreign and corporate donations is also a cause for concern, as is the timing of disclosure and the absence of caps on donations or campaign expenditure. These are all areas where Australia falls far behind the practices of countries such as the United Kingdom and Canada.

The timeline for reporting and disclosing donations (JSCEM, 2005a: 320) is such that donations to registered parties made in 2006–07 for the 2007 Federal election will not be disclosed until 2 February 2009. It is therefore too early for this study to determine the impact that the increased threshold has had on donation patterns. The long period between donation and disclosure is also a concern in terms of accountability and transparency, as media and public interest generally will have moved on from these issues. Furthermore, research into how donating patterns are influenced by disclosure thresholds would require access to parties’ financial records. In terms of fairness, the reform will result in a greater proportion of donations being hidden from voters, which is contrary to the principle of being able to make an informed choice.

Tax-deductibility of Political Donations

Prior to 2006, political donations were tax-deductible to a maximum of $100 per year. The Howard Government increased this limit to $1,500 per year. The Government’s argument for this increase was that it would encourage participation in the democratic process, by providing tax relief. In its report on the 2004 election, JSCEM recommended that the limit be increased to $2,000, arguing that it would encourage small to medium donations, but the only submissions cited by the report as being in support of such an increase were from the Liberal and National parties. Labor was opposed (JSCEM, 2005a: 338–9).

A counter-argument to the JSCEM’s recommendation is that the ability for people to participate in political discourse should not be related to their capacity to pay. An increase to $1500 skews political influence to the wealthier in society, as not only do

131 higher-earning individuals have greater capacity to make donations, they also receive a proportionately higher (taxpayer-funded) subsidy. Based on the tax scales in place at the time of the reform (2006–07 rates), the cost to an individual earning $20,000 per year was $410 more to make a $1,500 donation than for a person earning $100,000 (Kelly, 2006a: 12). In early 2008, the new Rudd Labor Government introduced legislation to remove tax-deductibility for all political donations.

Postal Voting Applications –– Self-Interest vs Fair Elections

The sections above have concentrated on reforms that have primarily been brought about by governing parties using their numbers in parliament to change legislation –– either for a direct partisan advantage over opposition parties or, when opposition numbers are needed for passage of the reforms, for mutual benefit and acting as a party cartel. Usually these reforms have had at least the tacit support of the AEC. There have been times, however, when the Labor and Coalition parties’ cartel is united in opposing the AEC’s efforts to operate a fairer and more independent electoral administration regime. One such area of electoral management that has been subject to legislative reforms throughout the 25-year study period is the administration of postal voting applications (PVAs). The Australian administration of this aspect of the voting process highlights the conflict between internationally accepted norms for fair elections and governing party self-interest.

In recent times postal voting has become more prominent in many modern democracies. It is seen as a way of encouraging and increasing voter turnout, especially among travellers, the elderly, and the infirm. Postal voting therefore promotes fairer elections by increasing the opportunity for participation by all groups in society. Typically, as with other aspects of electoral administration, the electoral management body is responsible for providing postal vote application forms and processing applications. However, a search of the ACE Electoral Knowledge Network database 29 did not reveal any instances where political parties are as intricately involved in the promotion and administration of the postal voting process, as in Australia.

29 http://aceproject.org

132

There has been a steady increase in the use of postal voting in Australian Federal elections during the study period, from 2.83 per cent in 1993, to 5.46 per cent at the 2007 election. 30 There have been a number of reasons for this increase. The 1983 reforms to the Commonwealth Electoral Act 1918 introduced a category of ‘general postal voter’; a 1990 amendment allowed employment reasons as grounds to apply for a postal vote; parties and candidates were allowed to include PVAs in their campaigning material following a 1998 amendment; and further amendments in 2000 and 2001 facilitated the electronic provision of postal voter information to parties (Newman, 2004: 14–5).

Part XV of the Act details the process of applying for a postal vote. As a result of the amendments in the past 25 years, political parties can now actively solicit for electors to lodge postal votes. Such solicitation is commonly carried out by the parliamentary representatives of the major parties, who can use both taxpayer-funded and party resources, such as elector databases and printing allowances, to mail postal vote applications to enrolled voters. It is also standard practice for the mail-outs to include a return envelope to the member or party, which then forwards the application on to the AEC. AEC figures show that of all PVAs received indirectly from electors (that is, through party agents and others), 97.8 per cent were forwarded by the Labor and Coalition parties (AEC, 2008: 36). This is an indication that the governing or major parties have an advantage over less-resourced parties in influencing electors.

Over the past 15 years, the AEC has consistently criticised the practice of such overt party involvement in the conduct of an election. The AEC’s arguments include: • There is a potential for voters to become confused, thinking that it is the parties who are responsible for postal voting, rather than the AEC; • Delays occur, due to applications not being returned directly to the AEC, resulting in possible disenfranchisement (AEC, 1996: 54–5); • Electors often receive more than one PVA, sometimes resulting in multiple applications being received by the AEC, adding to its administrative burden (AEC, 1999: 80);

30 Senate figures. Prior to the 1993 election, postal voting was around 4–6 per cent; however, pre-poll voting was introduced for the 1993 election, resulting in a significant initial drop in the rate of postal voting.

133

• There is unnecessary use of postal voting, increasing costs and delaying the finalisation of results (JSCEM, 2000: 56); • Political parties stockpile PVAs before sending them to the AEC, resulting in processing delays and, at times, disenfranchisement (JSCEM, 2000: 59– 60). Stockpiling is an offence under section 197 of the Act; • The secrecy and integrity of the ballot may be compromised, as votes are cast in an uncontrolled environment (Newman, 2004: 11); • Parties attempt to obscure the fact that the PVA is returned to a party address, by using terms such as ‘Returning Officer’ (AEC, 2008: 35); and • Party officials ‘correct’ details on PVAs before sending them to the AEC (AEC, 2008: 37).

Despite the AEC’s ongoing concerns, the JSCEM and successive governments have chosen to support the continued involvement of parties in the postal vote process. At times, the JSCEM has shifted blame to the AEC for some of the problems, stating that it ‘is possible that some [voters] were disenfranchised as a result of administrative errors by the AEC’ (JSCEM, 2000: 60). The general thrust of JSCEM reports is that the parties are providing a public service in distributing PVA material, for example: The Committee is of the view that distribution of PVAs by candidates provides an important and now well-established service to electors, and that it is important for candidates and political parties to be confident that a service initiated by them has been successfully concluded. The relatively high rate of return experienced in many electorates demonstrates the helpfulness and popularity of the service. Breaking with this practice at future elections may lead to significant voter inconvenience and possibly disenfranchisement. (JSCEM, 2003: 149)

This view was supported by both Labor and Liberal party members during a 2008 JSCEM hearing: We are liaising with a number of voters about their rights to vote. We are making sure that, if they have not got their postal ballot, they ring us –– they do not ring [the AEC]…Their relationship is with their political party, not the AEC. (Jon Sullivan MP, Labor, JSCEM, 2008: 11)

The elector who has dealt with the party does not ring [the AEC], the elector rings us…These [PVAs] go out with big Labor Party and Liberal Party stamps on them. I do not think they are confused where this information is coming from. (Scott Morrison, Liberal, JSCEM, 2008: 14)

134

Thanks to the parties, most voters, within a couple of days of the election being called, have got a PVA in their hot little hands. If the parties have no incentive to send that out they probably will not, therefore voters probably will not have the PVA in their hot little hands in that first week of the election. (Senator Simon Birmingham, Liberal, JSCEM, 2008: 13)

Postal Voting Processes –– The 2007 Federal Election

While there is little in the way of specific direction on this aspect of electoral administration, international best-practice literature consistently refers to the need for parties and candidates not to interfere with election processes (IDEA, 2002: 85). The emphasis is on parties observing rather than being involved in the conduct of an election. Under the Commonwealth system, parties can, either accidentally or deliberately, delay a PVA from reaching the AEC, meaning that eligible voters may be prevented from casting a vote.

Table 6.9 PVAs received by the AEC for the 2007 election –– period between witness signature and AEC receipt of PVA Source Sent Sent Sent via Sent via Other directly to via National Liberal 1 the AEC Labor Same Day 28,186 1,377 355 1,303 324 1 day later 37,089 3,828 844 2,499 323 2 days later 28,534 5,204 976 4,364 242 3 days later 22,575 5,155 1,083 5,305 273 4 days later 17,770 4,886 1,046 6,380 287 5 days later 11,368 4,055 823 6,686 273 6 days later 6,638 3,531 623 5,748 175 7 days later 4,331 3,045 448 4,547 81 8 days later 2,485 1,956 226 3,007 73 9 days later 1,428 1,237 177 1,829 26 10 days later 1,155 1,080 128 1,535 29 11 days later 922 882 114 1,119 21 12 days later 743 694 87 846 34 13 days later 678 653 95 726 26 14 days later + 5,395 5,740 543 4,869 108 Total 169,297 43,323 7,568 50,763 2,295 1 Includes Country Liberal Party Source: AEC, 2008: 36.

During JSCEM’s inquiry into the 1998 election, the AEC identified 174 electors who were disenfranchised through apparent delays in parties returning PVAs (JSCEM, 2000:

135

59). The AEC has raised its concerns since at least 1993 about the delays created by parties stockpiling PVAs, and evidence in its 2008 submission to JSCEM suggests the practice is continuing, with a surge in the number of applications being received more than 14 days after the witness signature on the PVA (see Table 6.9). The proportion of all PVAs received was far higher for the parties (Labor –– 13.25%; Liberal –– 9.59%) for this category (more than 14 days after signing), than for those sent directly to the AEC (3.19 per cent of all PVAs). This indicates delays from the double mailing involved in sending a PVA to the AEC via a party address, and possibly from the parties stockpiling PVAs prior to sending them to the AEC. In either case, the possibility of voters being disenfranchised increases.

A test was conducted during the 2007 election to gauge the service from the parties, and their attitude to dealing with PVAs. The author received PVAs in the mail from the Labor and Liberal parties, as outlined below.

Labor The Labor package (see excerpt at Figure 6.4) was received in an envelope marked ‘Important Voter Information Enclosed –– From Bob McMullan MP Federal Member for Fraser’. On one side of a folded A3-size pamphlet was the PVA and early voting information from the AEC. The other side contained party campaign material (including three photos of Bob McMullan and Labor mentioned 11 times). On three panels, reference is made to a ‘Postal Voting Hotline –– 6257 7575’. The number is that of McMullan’s campaign office. A reply paid envelope was enclosed for the voter to send the PVA: it was addressed to Bob McMullan MP.

Liberal The Liberal party’s Senator Gary Humphries sent a personally addressed A4-size campaign letter, at the end of which was the text: ‘PS: I enclose a postal vote application in case you are unable to vote on polling day’. The enclosed PVA (see excerpt at Figure 6.5) was on a slightly-larger-than-A4 folded pamphlet. The PVA and associated AEC information was displayed on seven of the eight panels, with the remaining panel containing party campaign material. At the bottom of one panel of AEC material was a message in large font –– ‘Need more information? Ph (02) 6247 6444’. This is Senator Humphries’ electorate office number. A reply paid envelope was enclosed, addressed to the Canberra Liberals.

136

Figure 6.4 Labor Postal Voting Information Leaflet (excerpt)

137

Figure 6.5 Liberal Postal Voting Information Leaflet (excerpt)

138

In both cases, the parties provided official AEC information, including the AEC’s phone number (in small font) while promoting their own phone numbers in a much larger font. To gauge how willing the parties were to give out AEC contact details, a call was made to each of the printed party numbers, asking for the address to send the PVA directly to the AEC. The responses were: Labor –– You want to send it straight to them, OK. It’s GPO Box 2867, Canberra City. Liberal –– I’ll just grab it for you. Here it is. It’s GPO Box 2867, Canberra City.

A call was made to the AEC information line, also seeking their postal address: AEC –– What is your postcode? NK –– 2602 AEC –– So you’re in New South Wales. NK –– No, I’m in the ACT AEC –– There’s only one electorate in Canberra, isn’t there? NK –– [explained that there are two electorates in the ACT] After a pause, the postal address was given. The call took three minutes.

The records of these calls may give the impression that the parties are better equipped to respond to electors’ inquiries. It appears, however, that the parties’ proficiency in dealing with PVA enquiries has developed in response to the fact that a significant number of electors believes that it is the parties’ responsibility to deal with these matters –– a perception that is encouraged by the parties themselves. In terms of fairness and equity, it is apparent that it is only the Coalition and Labor parties that have the financial and database resources to engage in mass mail-outs to electors. With the incumbency benefit of having electronic databases of electors, these mail-outs can be designed in a personalised format. The databases and tracking systems that have been developed by the major parties also include information that indicates whether an elector is a possible party supporter. Although there is no evidence that this practice occurs, it is conceivable that a party might delay forwarding a completed PVA to the AEC if the elector is identified as a non-supporter.

Despite continuing evidence of voters being disenfranchised due to the delays in parties returning PVAs to the AEC, the AEC has been reluctant to initiate prosecutions against

139 the parties (JSCEM, 2000: 59–60; 2003: 147–8). This reluctance may be due in part to the AEC’s need to keep the parties ‘on side’, as the parties’ support is needed if legislative change is to occur. While the AEC is limited by its legislation, the recent history of the PVA issue indicates that the AEC does not operate as an independent body, and that it falls short in terms of Wall et al.’s criteria on institutional arrangements, implementation, and powers (2006: 9).

A Change of Government –– More Reform

The election of the Rudd Labor Government in November 2007 has provided fresh impetus for electoral reform. Labor has called not only for the repeal of some of the Howard Government’s 2006 reforms, but has gone further in proposing new transparency measures and considering issues such as the length and variability of terms. In March 2008, the new Special Minister of State, Senator John Faulkner, announced that the Government would publish two green papers on electoral reform, one on political finance issues and the other on broader matters of electoral administration (Faulkner, 2008b). Originally due to be released in mid-2008, their publication has been delayed until later in the year. 31 In the meantime, the JSCEM is conducting its inquiry into the 2007 election, with a particular emphasis on political finance matters.

While the Rudd Government also moved quickly to remove tax-deductibility for political donations by introducing legislation, 32 this was referred to the JSCEM, and despite the committee recommending that the Bill be passed, it did not progress any further in the Senate. The Government re-introduced the legislation in August 2008. 33 The Labor Government may find it difficult to pass reforms through the current Senate, where it is in the minority. In June 2008 (when the Coalition had a Senate majority), the Senate referred the Government’s package of reforms to JSCEM for inquiry. 34 These reforms sought, among other things, to reduce the donation disclosure threshold from $10,500 to $1,000. Labor is seeking support from the Coalition opposition for its

31 Green papers are discussion papers issued by the government, indicating possible reform measures. At the time of writing (December 2008), neither green paper had been publicly released. 32 Tax Laws Amendment (2008 Measures No. 1) Bill 2008. 33 Tax Laws Amendment (Political Contributions and Gifts) Bill 2008. 34 Commonwealth Electoral Amendment (Political Donations and Other Measures) Bill 2008.

140 reforms, but due to the Coalition’s opposition to Labor’s electoral legislation so far, it has been suggested that the Government might wait until the next parliament in the hope that there may be a Labor/Greens majority in the Senate (Millar and McKenzie, 2008).

Summary

This chapter has identified the self-interest of the major governing parties as a significant driver behind electoral reform. This factor is consistent irrespective of whether the reform has occurred under a Labor or Coalition government. The three changes of government since 1983 have all been quickly followed by reform activity. In Labor’s case in 1983, there was a degree of bipartisan support from the Coalition opposition for many of the changes. When it suited, however, Labor was happy to form alliances with the Nationals and Democrats to pass its reform agenda. Perhaps despite Labor’s expectations, the increase in the size of parliament does not appear to have favoured Labor, and the increased disproportionality of the Senate in recent years is most likely reflective of changing voter behaviour rather than a direct result of the reform.

The introduction and increasing use of Senate above-the-line ticket voting raises issues of fairness to voters. It can be argued that the reduction in informal voting is beneficial to voters, as it means more votes are being counted towards the election result. It can also be argued, however, that the increased power given to parties in determining list positions and preference flows actually disenfranchises voters from being able to register their preferred vote with ease. That partisan interest has over-ridden fairness to voters can be more readily argued when alternative solutions, in this case including optional preferential voting, were considered and dismissed.

A natural outcome of the introduction of public funding is that the major parties are far more highly-funded than smaller parties, increasing the disparity between the resource bases of competing parties. Despite the obvious benefits the scheme brings to parties of a significant size, the reform has also supported the development and success of smaller parties, and is generally in line with international norms of best practice in encouraging electoral competition. However, the inadequate regulation of private funding (such as

141 having no limits on the size or source of donations, combined with a high disclosure threshold) undermines the principle of voters being able to make an informed choice, including being aware of potential conflicts of interest. Public funding also raises broader questions of party regulation (considered more fully in Chapter 8) and the mix of funding sources that should be available to parties and candidates (for example, see Young and Tham, 2006).

Compared to the Hawke Labor Government, the Howard Coalition Government was far less willing to accommodate other parties’ concerns after it came to power in 1996, and therefore had to wait until it gained a Senate majority to implement the major components of its reform package. It has been argued that the Coalition’s reforms, such as creating greater barriers to enrolment and reducing the transparency of donations, run counter to accepted norms for fair elections. The inability of the Howard Government to gain sufficient parliamentary support for its initiatives prior to 2006 indicates the extreme nature of the provisions, and the moves by the current Rudd Government to repeal these reforms is indicative of the lack of broad public support for the Howard reforms.

The history of the administration of postal voting throughout the study period is an excellent example of party cartelisation at work, in this case by Labor and the Coalition. Despite concerns repeatedly raised by the AEC, these parties have continued to put partisan self-interest ahead of the internationally-accepted best practice of maintaining a clear delineation between electoral contestants and electoral administrators. Labor and the Coalition have exploited their positions in the hierarchy of incumbency, and in doing so have created an unfair advantage over other candidates and parties.

What appears to be clear from the analysis of the reforms discussed in this chapter is that partisan interest is a far more significant driver for reform than considerations of fairness to voters, or the goal of improving the electoral system in accordance with internationally-accepted norms for electoral management and fair elections. While some reforms may create negligible, if any, partisan advantage, and may be consistent with international norms, such neutral or positive effects appear to be incidental to the use of legislative power to consolidate and improve the governing parties’ respective positions. The following chapter considers the driving motivations for specific reforms

142 that have occurred in each of the States and Territories during the study period, and analyses their impacts on the fairness of elections.

143

Chapter 7

Reform Processes II –– States and Territories

Why is it in Western Australia that our Labor politicians are so loathe to tackle genuine electoral reform when it has been achieved so remarkably by Labor Governments in South Australia, New South Wales and Victoria?...I never thought it possible that I would hear such a betrayal of basic Labor principles from people whom I believed were my comrades. Arthur Tonkin, former Western Australian Minister for Parliamentary and Electoral Reform, 1968 (cited in Phillips and Robinson, 2006: 40)

This chapter discusses some of the significant electoral system reforms that have occurred in the State and Territory jurisdictions during the study period. Other reforms that have been broadly adopted across Australia are examined in Chapter 8. Based on an analysis of legislation, interviews, and debates, specific issues and events in each jurisdiction were selected for assessment. Issues have been selected for their uniqueness to a jurisdiction or their presence in only a minority of jurisdictions, and for their importance in relation to principles of democratic fairness and the reform process.

The sections below use a case study approach to explore how and why reforms have occurred, the impacts these reforms have had on electoral democracy, and whether such reforms are consistent with internationally accepted norms for electoral management and fair elections. Analysis in the previous chapter demonstrated that governing parties instigate electoral reforms to advantage themselves; the analysis presented in this chapter identifies instances in which this has not necessarily occurred.

144

New South Wales

Two factors that impact on fairness and equity are the level of competition, and ballot paper design. While competition is an essential component of democratic elections, the need to choose between a large numbers of candidates can create voter confusion. In this context, the number of candidates is excessive when it becomes difficult for voters, in attempting to make an informed choice, to be aware of what all the parties and candidates represent, and to differentiate between them. In designing ballot papers, it is the role of the electoral administration to minimise levels of informal voting. It therefore must seek to ensure that electors are not accidentally disenfranchised by a ballot paper design that is confusing to follow.

New South Wales Legislative Council elections during the study period provide an excellent case study of the impact that party registration and ballot design rules can have on the conduct and outcomes of elections, including the proportionality of results. The recent history of New South Wales electoral law, as will be discussed below, indicates that if party registration rules are weak, they will be used by political opportunists seeking electoral advantages. This has been noticeable on occasions in New South Wales, and to a lesser extent, in South Australia, two States that have used group voting tickets for their upper house elections.

Too Many Parties –– the 1999 Legislative Council Election

The March 1999 NSW Legislative Council election produced one of the largest ballot papers ever used in Australia (and possibly the world), with 81 groupings (including 78 parties) comprising 264 candidates. The ‘tablecloth’ ballot paper measured 102 centimetres by 72 centimetres. Its size created major logistical issues for the election, requiring the construction of wider voting booths and the use of larger planes for transporting papers (Green, 2003: 1). The election also produced some intriguing results, including the election of an Outdoor Recreation Party candidate, Malcolm Jones, who polled 0.19 per cent of the primary vote (a quota being 4.5 per cent). Jones was elected with the support of preferences from 21 other parties, including eight that had received a higher primary vote than Outdoor Recreation (Green, 2003: 27). The

145 problems associated with the 1999 Legislative Council election represented the culmination of a series of earlier reforms and provided the impetus for further reforms.

From the mid-1980s, reforms were introduced by both Coalition and Labor governments, initially by Liberal Premiers Nick Greiner and John Fahey and later by Labor Premier Bob Carr. Some Coalition reforms also required approval by referendum, with successful referenda being held in conjunction with the 1991 and 1995 elections. Reforms affected many aspects of elections, including: the number of members to be elected and the length of terms; ballot paper design, including the use of above-the-line ticket voting (single vote and preferential) and group voting tickets; and party registration requirements (Simms, 2006: 116–21; Green, 2003: 4–9).

Table 7.1 presents a summary of Legislative Council election results since 1984, and highlights several issues relevant to democratic principles of equity and access. As the table shows, the increase in parties and candidates at the 1999 election is obvious; however significant increases in both are apparent prior to this (in 1995). The success of A Better Future For Our Children candidate, Alan Corbett, in being elected in 1995 on a low 1.28 per cent primary vote and a favourable flow of preferences (requiring 4.5 per cent to achieve a quota) obviously encouraged the formation of many parties for the 1999 election. The formation of new parties was facilitated by the electoral laws in place at the time. Although one of the party registration criteria was a requirement for 200 members, a person could be a member of more than one party for registration purposes and it appears that several parties used petition lists to meet their 200-member requirement (Green, 2003: 8).

Table 7.1 NSW Legislative Council election results 1984–2007 Year Groups/ Candidates Informal Gallagher’s Least Parties 1 vote (%) Squares Index 2 1984 8 43 6.66 3.77 1988 13 56 8.08 4.84 1991 12 54 5.67 4.68 1995 28 99 6.11 9.47 1999 81 264 7.17 8.69 2003 16 284 5.34 4.88 2007 20 333 6.11 5.47 1 Includes ‘Independent’ groupings and ‘Ungrouped’ column. 2 Data at Appendix E. Sources: Green, 2003; NSWEC; AGPD.

146

The Carr Labor Government responded quickly to the negative publicity received by the tablecloth ballot paper. The Parliamentary Electorates and Elections Amendment Act 1999 was passed in November 1999, driven by a desire to avoid the complexities and difficulties of the 1999 election. The Act’s amendments were targeted at ballot design and party registration, to reduce the number of parties contesting Council elections. The main amendments were: • Abolishing group ticket voting (removing the ability for parties to direct preferences); • Allowing voters to record preferences above the line; • Increasing the membership requirement for party registration from 200 to 750; • Removing the opportunity to register a party based on having a member of parliament (rather than having a minimum number of members); • Removing the capacity for a person’s membership to be used to meet the membership requirement of more than one party; • Requiring a $2,000 registration application fee; • Parties to be registered one year ahead of an election for party identification on ballot papers; and • Increasing the powers of the Electoral Commissioner to investigate whether party membership is genuine.

The impacts of these reforms can be clearly seen in Table 7.1. There was a significant reduction in the number of parties contesting the 2003 election (16, down from 81); however there was an increase in the number of candidates (284, up from 264). This was due to the combined effect of removing ticket voting, introducing optional preferential voting above the line, and the pre-existing requirement for a voter to indicate preferences from 1 to 15, in order for a vote to be counted as formal. That is, parties needed to field at least 15 candidates for a vote that is recorded as a ‘1’ or a tick for the party to be counted.

The final two columns of Table 7.1 provide evidence of other impacts that the system, and its reforms, have had. In regard to informal voting, it can be argued that reforms have had little impact, as the rate has remained consistently high throughout the study period. However, reforms have included the introduction of above-the-line ticket voting

147

(from the 1988 election), being able to use a tick to record a formal vote (from 1995), and optional preferential voting (from 2003) –– all innovations that would be expected to lower the informal rate. One of the reasons why these interventions have not reduced the informal vote is that the size of the ballot paper, especially since the 1995 election, has been intimidating and confusing to voters. This may be especially so as voters are confronted with a vastly different (and smaller) ballot paper for Legislative Assembly elections, and have to deal with different above-the-line rules for Federal Senate elections. In addition, the changes that have occurred throughout the period have potentially made it difficult for voters to become familiar with any one format, or for voter education programs to have made a long-term impact.

Using Gallagher’s Least Squares Index (LSI), the impact of the reforms is evident for the 1995 and 1999 elections, with disproportionality in these polls at its worst out of the seven elections during the study period. Proportionality has improved at the two most recent elections, but remains at a level higher than pre-1995. This may, however, be due to the general increase in legitimate minor parties in more recent times. Another factor is the methodology in calculating the Index, which accentuates the impact of very small parties being grouped together. 35

In terms of fairness, the ultimate results of these reforms are positive for candidates and parties, with competition now more genuine. In addition, representation now more closely reflects voters’ choices, with an average proportionality Index value of 5.18 for the two most recent elections, compared to an average 9.08 for the 1995 and 1999 elections. However, while it may now be easier to make an informed choice, the continuing high level of informal voting remains a serious concern.

Victoria

It is not often that a governing party intentionally implements an electoral reform that will disadvantage the party in the future. The Bracks Labor Government’s introduction of proportional representation for the Victorian Legislative Council in the early 2000s did exactly that. While the move to proportional representation has had a positive

35 For these calculations, all parties receiving less than one per cent were grouped together under ‘Independents and Others’. See Appendix C for calculations.

148 impact on the representativeness of the Council, there have been unintended negative consequences, including an increase in rates of informal voting.

The Move to Proportional Representation

In the 1980s, Labor’s long-held policy of abolishing the Legislative Council was moderated due to factional differences within the party –– primarily, concerns that the Socialist Left would lose its upper house power base (Costar, 1992: 204–6). As an alternative, Labor proposed the introduction of proportional representation, a move made to secure a preference deal with the Australian Democrats, who were expected to benefit from a proportional system (Strangio, 2004: 44). The Cain Labor Government made repeated attempts to reform the upper house, but each reform Bill was defeated in the Council. It was not until the election of the Bracks Government in 1999 that Labor had another opportunity to pursue its reform agenda. However, the Bracks Government in 1999–2002 was a minority Government relying on Independents to stay in power, and was consequently unable to gain sufficient support to implement its agenda.

The Bracks Government was re-elected in 2002, this time with a majority in the Assembly, and a historic first-ever majority in the Council. These majorities provided Premier Steve Bracks with the opportunity to fulfil Labor’s desire to abolish the upper house. Despite the double majority, however, Bracks maintained his first-term position of reform rather than abolition, and legislation was passed in 2003 for Council elections to be based on eight five-member regions, with all Councillors serving fixed four-year terms. As Labor had polled less than 50 per cent at the previous two Council elections, 36 this could be expected to have a severe impact on the party’s ability to win majorities in the future. The negative impact for the party was demonstrated at the 2006 election, where Labor won only 19 of the 40 seats. 37 As Victorian political scientist Paul Strangio notes, ‘having waited a century for a majority in the Council, by introducing proportional representation upon finally claiming that prize in 2002, Labor has, in all probability, diminished its chances of exercising future control over the Upper House’ (2004: 48–9).

36 In 1999, Labor won 8 of 22 seats with 42.19% of the vote. In 2002, the party won 17 of 22 seats with a 47.49% vote. 37 41.45% of the vote.

149

The democratic impacts of this reform are shown in Table 7.2. These include a substantial improvement in proportionality. This is an obvious and expected improvement in the move from a majoritarian system, but it is also important to consider the potential impacts of having regions instead of a State-wide electorate. In Victoria’s case, this has resulted in required quotas of 16.7 per cent to win a seat, rather than 2.44 per cent for a State-wide electorate of 40 seats (or 4.76 per cent if half the Council was elected at every second election, as occurs in New South Wales and South Australia). Theoretically, the higher quotas make it more difficult to attain proportionality, and yet the 2006 LSI figure of 6.16 is comparable with New South Wales (5.47 for its 2007 election), which has quotas of 4.5 per cent.

Table 7.2 Victorian Election Results 1999–2006 Year Gallagher’s Council Assembly Least Squares informal informal Index voting (%) voting (%) 1999 10.82 3.37 3.02 2002 27.09 3.67 3.42 2006 6.16 4.28 4.56 Source: VEC. Calculations at Appendix F.

A further democratic impact illustrated in Table 7.2 is the increase in informal voting for both houses at the 2006 election. The 2006 election heralded the use of different voting systems for the two houses, and it appears that this change has unintentionally disenfranchised some voters in both the Legislative Assembly and Legislative Council. The Victorian Electoral Commission (VEC) undertook an analysis of a sample of informal ballot papers to determine why the increases had occurred. The outcomes of the VEC study suggest that the increase in the Assembly informal vote (the highest ever for an Assembly general election) was due to confusion brought about by the introduction of the new Council voting system; a quarter of all Assembly informal votes used a ‘1’ only (formal for an above-the-line Council vote, but informal for an Assembly vote) (VEC, 2007: 93). The increase in informal Council votes was partly attributed to voters misunderstanding the direction to ‘vote “1” above the line’; many voters used only a ‘1’ below-the-line. More than 27 per cent of the sample declared informal was made so by this error (VEC, 2007: 96).

150

The ability of the VEC to initiate reforms is limited. While the Commissioner can authorise modifications to the design of ballot papers, any change to the instructions on a ballot paper which could assist voters requires amendments to the Electoral Act 2002 . Similarly, any systemic changes, such as the introduction of optional preferential voting to reduce informal voting, would require legislative action. Consequently, the VEC has made a commitment to investigating design changes to the Council ballot paper design (VEC, 2007: 118). However, rather than recommending a specific change to the voting system (such as introducing optional preferential voting), the VEC has suggested that parliament first conduct research into the issue (VEC, 2007: 126). The Parliament’s Electoral Matters Committee is currently investigating this issue, and is expected to report in mid-2009.

Queensland

Queensland is the only Australian State operating under a unicameral system, having abolished its Legislative Council in 1922. The State’s single house, the Legislative Assembly, is elected from 89 single-member electorates using preferential voting. During the study period, the only major wave of electoral reform occurred in 1991, with a move to ‘one vote, one value’ and the re-introduction of optional preferential voting (replacing compulsory preferential voting, which had operated since 1962). These changes flowed from the outcomes of the Fitzgerald Inquiry, which reported in 1989. The Inquiry was centred on government corruption, and found instances of illicit political donations influencing government decisions. G. E. (Tony) Fitzgerald also pointed out the general unfairness of the Queensland zonal system which provided undue weighting to country electorates, stating that: The fairness of the electoral process in Queensland is widely questioned. The concerns which are most often stated focus broadly upon the electoral boundaries, which are seen as distorted in favour of the present Government, so as to allow it to retain power with minority support. (Fitzgerald, 1989: 127)

Fitzgerald recommended setting up an Electoral and Administrative Review Commission (EARC) (Fitzgerald, 1989: 348), which was duly established by the Cooper National Party Government prior to the 1989 election. Part of the Commission’s remit was to investigate and report on the Queensland electoral system.

151

The Goss Labor Government subsequently acted on the EARC’s recommendations, passing the reforms referred to above. By 1993, the EARC’s work was considered to be completed, with a parliamentary committee –– the Parliamentary Committee for Electoral and Administrative Review, which had been established in 1990 –– taking over responsibilities for inquiring into electoral matters. This represents a relatively short timeframe for the reform process to be concluded, with the major reforms being passed less than three years from the release of Fitzgerald’s report (1989 to 1991). The speed at which the process took place was assisted by the intense media scrutiny of the Inquiry process and subsequent political consequences, in which the Coalition parties lost power for the first time in 32 years.

Examination of the election results during the study period provides an indication of the impacts of the reforms, with three elections held prior to the 1991 changes, and six elections held between 1992 and 2006. As argued by Fitzgerald, the abolition of the zonal system was required to provide fairer outcomes and prevent the Nationals from gaining power with minority support (when compared to its Coalition partner, the Liberal party). Up until the 1980s, the National Party won more seats than the Liberal Party at elections, despite the Liberal Party winning a higher vote. Since the 1980s, the National Party has regularly outpolled the Liberals.

Table 7.3 Queensland Election Results 1983–2006 Election Gallagher’s Least Informal Squares Index voting (%) 1983 9.34 1.47 1986 12.85 2.17 1989 12.50 3.00 1992 12.85 2.25 1995 9.31 1.75 1998 13.79 1.45 2001 20.11 2.27 2004 19.89 1.99 2006 16.85 2.08 Note: 82 seats in the 1983 election, 89 seats from 1986 to 2006. Data and calculations at Appendix G. Sources: ECQ; AGPD.

Table 7.3 details the proportionality of results over the period. The removal of the zonal system is a positive outcome in relation to the fairness of elections. However, it is evident that election results have become less representative of voters’ choices. The

152 three most disproportional results –– in 2001, 2004, and 2007 –– coincide with Labor’s three most successful elections in terms of seats won (see Appendix G). An indication of the benefit that Labor has received from the reform can be seen by comparing the 1989 (pre-reform) and 2001 elections (post-reform). In 1989, Labor won 54 seats with 50.32 per cent of the vote, yet in 2001 won 66 seats with only 48.93 per cent support. The reform has had little impact on the Liberal Party, with the party being under- represented in terms of percentage seats to percentage votes for every election since 1983.

The other aspect of the reform package –– the introduction of optional preferential voting –– appears to have had little impact on the rate of informal voting, which has consistently remained at a low level throughout the study period. Optional preferential voting is usually expected to reduce rates of informal voting (as discussed above in the Victorian case study). There are two reasons why this has not occurred in Queensland. First, the rate of informal voting was already low prior to the change; second, voters only fill out one ballot paper on election day, removing the confusion experienced by voters in bicameral jurisdictions.

A significant unforeseen impact of the reform is the change in how political parties run their campaigns. Prior to the 2001 election, the major parties advocated filling out all preferences. From 2001, however, parties have increasingly advised voters to ‘just vote one’, with all major parties now advocating this option on their how-to-vote material (ECQ, 2007: 9, 21). As a result, approximately two-thirds of voters used the ‘1 only’ option at the 2006 election (ECQ, 2007: 8).38 This is leading to concerns that the system is turning into a de-facto first-past-the-post system, where seats are determined by a minority of votes, and voters are not appropriately informed of their options (for example, see Wanna, 2004). Consequently, while the reform is an improvement on fairness in regard to ease of voting, voters’ ability to make an informed choice is diminished due to the limited information provided in parties’ how-to-vote material.

38 Based on an Electoral Commission of Queensland survey of 11 seats.

153

Western Australia

The recent history of electoral reform in Western Australia provides examples in which partisan considerations have been a more powerful driver for reform than considerations of fairness principles. During the study period, two waves of significant reform have occurred, both under Labor governments. The first was instigated by the Burke Government during the 1980s and the second by the Gallop Government in the 2000s. Both sets of reform have included significant changes to the Legislative Council system, and moves to ‘one vote, one value’. On both occasions, the reforms have represented the triumph of political pragmatism over accepted norms for fair elections.

The Burke Government Reforms

Western Australia has historically had a highly malapportioned system for its two houses of parliament (see Phillips and Robinson, 2006; Kelly, 2004). When the Burke Government was first elected in 1983, voting for the Legislative Council was based on provinces electing two members each, with one member elected every three years for a six-year term. There was a high degree of vote weighting in favour of non-metropolitan seats. The Burke Government moved to introduce proportional representation for the Legislative Council based on a single State-wide electorate, including the requirement for ‘one vote, one value’ for the Legislative Assembly (Phillips, 1991: 234). However, the initial legislation 39 was unsuccessful, and after further failed attempts at reform, legislation was finally passed in 1987. 40 Although proportional representation was introduced for the Legislative Council, Labor retained a level of zonation, or malapportionment, based on three metropolitan and three non-metropolitan regions. 41 Vote weighting for the Assembly was also retained.

The retention of malapportionment was necessary to secure the support of the Nationals in the Council, who continued to be major beneficiaries of weighting in favour of rural regions. While the introduction of proportional representation had a positive impact on fairness and equity, the reform was tempered by the political realities of needing support

39 Acts Amendment (Constitution and Electoral) Bill 1983. 40 The Acts Amendment (Electoral Reform) Act 1987 . The legislation was introduced in July 1986, and passed in June 1987. 41 In the Council of 34 members, 17 were elected from metropolitan regions and 17 from non- metropolitan regions. At the 1989 election, approximately 74% of voters lived in the metropolitan area.

154 from a non-governing party. As Jim McGinty, a recent Labor Minister for Electoral Affairs, described in an interview: It was a deal. It was a significant improvement in the electoral system, as it took out the grossest of malapportionments, and it provided for proportional representation, which I think is a fair way for making representation in the upper house…It was a significant improvement but there was nothing principled about it. It was just a deal that was done…It just knocked some of the harsher degrees of malapportionment out, which then makes it harder to achieve the principle.

McGinty went on to explain the political motivations for then Premier Brian Burke’s position on electoral reform: Brian Burke in ’87 had the opportunity to achieve ‘one vote, one value’, but he didn’t want to do it. Brian Burke and Julian Grill pulled the plug on the Labor party’s policy commitment to ‘one vote, one value’ in order to maximise Labor’s chances of winning key marginal electorates in country cities and large towns. Arthur Tonkin resigned from the ministry over it –– he was Minister for Electoral Affairs –– because Burke had ratted on party policy.

The Gallop Government Reforms

When Labor won the 2001 election, it renewed its attempts to introduce reform to achieve ‘one vote, one value’. However, following a convoluted saga of parliamentary debates, a committee inquiry, and Supreme and High Court challenges, the Gallop Government’s legislation was eventually unsuccessful. When Labor was re-elected on 26 February 2005, it had a brief opportunity to pass legislation with the support of the Greens and Alan Cadby –– a former Liberal MLC who was sitting as an Independent and due to retire on 21 May 2005. As in 1987, Labor needed to moderate its plans, this time by making concessions acceptable to the Greens and Cadby.

Labor yielded to the Greens’ demand that non-metropolitan representation not be weakened for the Council in ensuring that ‘one vote, one value’ could be introduced in the Assembly. In addition, the only exception from the equality principle for the Assembly that the Greens would support was for geographically large electorates. In order to meet its own election commitment to retain the five Assembly seats in the Mining and Pastoral Region, the Gallop Government increased the size of the Assembly from 57 to 59 seats. This allowed the five seats to be retained using the Greens’

155 criteria.42 The legislation was passed in May 2008, in the week prior to Cadby and three of the Greens members leaving their seats in parliament.

The outcome was a significant improvement in voting equity for the Assembly, with only five seats impacted by the ‘geographically large’ criterion. However, the reforms also increased the Council from 34 to 36 members, by creating six six-member regions (previously from five- or seven-member regions). This led to an increase in the level of malapportionment for the Council, from 4.1:1 to 4.6:1 in the worst-case scenario (Kelly, 2006b: 425). Furthermore, the quota to win a Council seat increased in the two regions where the Greens were successfully elected in the 2005 election, indicating that the Greens’ motives may have been more altruistic than partisan.

Table 7.4 Western Australia –– Gallagher’s Least Squares Index Election Legislative Legislative Assembly Council 1983 4.48 10.92 1986 7.87 11.29 1989 11.79 7.04 1993 7.22 5.69 1996 10.22 6.29 2001 16.61 5.96 2005 12.64 6.83 2008 12.18 8.62 Sources: WAEC; AGPD; Jaensch, 1986; Western Australia Legislative Council, 1983. See Appendix H for calculations.

Examining the effects of the 1987 and 2005 reforms, it is clear that accepted norms of fair elections, particularly in terms of equal votes, have been a secondary consideration after the more pressing constraints experienced by a government that does not have sufficient numbers in its own right to pass legislation. An assessment of proportionality using the Gallagher Least Squares Index is useful in evaluating the impact of these two sets of reforms. It can be seen from Table 7.4 that, prior to the 1987 reform, the Legislative Assembly achieved greater proportionality than the Legislative Council.

This supports the view of Buxton (previously discussed in Chapter 2) that malapportioned systems can still produce proportional results based on the spread of votes (1985: 57). As expected, the Council became more representative of voters’

42 For fuller accounts of these reforms, see Kelly, 2004 and 2006b; and Phillips and Robinson, 2006.

156 choices with the introduction of regional proportional representation for the 1989 election. It is interesting to note, however, that there is a trend of decreasing proportionality in recent elections for both Houses. This is due to the growth in support for minor parties, particularly the Greens, and in 2001, One Nation. The decreasing proportionality also partly reflects the over-representation of Labor in the Assembly in 2001, 2005 and 2008 (see Appendix H).

South Australia

In comparison with the other Australian jurisdictions, South Australia is distinctive in having introduced a relatively small number of major electoral reforms during the 25- year study period.43 This fact alone is interesting and invites inquiry into the reasons for the lack of reform. In addition, despite the apparent absence of legislative activity, the State is unique in Australia in having laws that prohibit inaccurate and misleading statements in election campaign advertising. These two aspects of South Australia’s electoral system is the focus of this case study.

Challenges in Implementing Reform

It has been difficult in recent years to implement electoral reform in South Australia, with the most recent amendments to the Electoral Act 1985 occurring prior to the 1997 general election. Despite the State Electoral Office (SEO) recommending legislative action on a variety of issues that emerged in its past three election reports (for the 1997, 2002 and 2006 elections), the principal Act remains unchanged (SEO, 2007a: 51). One reason for this was explained by an electoral official: There was a Bill in the parliament in about 2001 that just fell over because, they were very trivial type things, and you had McEwen in the lower house who was an independent member –– he stood as an independent Liberal, and one of the proposals that the government of the day put in was that they wanted to get rid of the use of the word Labor or Liberal connected to Independent. You could stand as an Independent. It passed the upper house. McEwen said, ‘stuff you, if you want that to go through, I won’t support you in government’. Basically it fell over at that point before the 2002 election. It just stalled and went nowhere. It’s never been introduced back in the house since.

43 South Australia’s Electoral Act 1985 included a major re-write of the previous legislation, the introduction of ticket voting for the Legislative Council, and a party registration regime. The Constitution (Electoral Redistribution) Amendment Act 1991 introduced a ‘fairness’ clause for redistributions. As explained in Chapter 3, redistribution processes are beyond the scope of this thesis.

157

While the SEO may recommend changes to electoral law, it is careful in its wording when advocating for reform. Typically, in its election reports the SEO recommends that certain matters should be ‘considered’, or ‘may need reviewing’ (SEO, 2007b: 59–65), rather than advocating a specific course of action. This allows the SEO to push for reform, without being constrained if its recommendations are not adopted. As explained by one official: A simple example in where we draw the line is, in an election report, or two election reports, and about to be a third one since the legislation last changed, there might be a recommendation that [says] ‘yep, here’s some trends. New South Wales had a ballot paper that size. Massive amounts of political parties. Difficulties in processing it. The parliament might want to consider the ability to register political parties. What the requirements are –– should they have to register every year?’

That’s about where it goes. Then the Attorney-General and advisors might come up and go, ‘should we make it 500 members, should we make it a thousand, will we make it $2,000 a year to do it?’ They’re policy decisions. We wouldn’t say ‘here’s how you should do it’. We’d say ‘you need to look at this issue’.

Even when the SEO has been more definitive, however, recommending that some measure be ‘inserted’ or ‘amended’, there remains a lack of significant action on the part of legislators. The legislation has not changed, for example, despite repeated recommendations that the party registration process should be reformed, since the growth in parties at the 1999 election (see for example, SEO, 2007b: 61). Amending legislation was passed by the Legislative Council in 2002, but lapsed (SEO, 2002: 29), and has not subsequently been re-introduced. The lack of legislative amendments demonstrates the inability of the SEO to instigate electoral reform in South Australia. As a result, electoral laws are determined by political considerations rather than principles of fairness. South Australia’s party registration issues are examined in more detail in the following chapter.

Truth in Political Advertising

There have been several attempts in recent years to introduce provisions for greater honesty in election campaigns. 44 The principle underpinning such laws is that they

44 For example, Australian Democrats Senator Andrew Murray’s Electoral Amendment (Political Honesty) Bill 2003.

158 bring election campaign advertising in line with provisions that exist for businesses under the various Trade Practices Acts (Murray, 2003: 10324–5). Reform efforts are also aimed at preventing candidates with significant campaigning resources from obtaining an unfair advantage, especially when the veracity of the advertising may not be determined until after the election. From the voter’s perspective, the ideal is to be able to make an informed decision based on fact. The fairness of an election comes into question when the laws are weak in protecting that ideal.

South Australia is the only Australian jurisdiction that currently has ‘truth in political advertising’ legislation,45 with section 113 of its Electoral Act 1985 making it an offence to publish an advertisement that ‘contains a statement purporting to be a statement of fact that is inaccurate and misleading to a material extent’. However, it has proved difficult for the SEO to enforce this provision, or to achieve prosecutions. An example is the case of Sam Bass, who lost his seat at the 1997 general election. It took five years for this case to be finally decided in the High Court. 46 The process was described by one electoral administrator: There were a couple of Labor workers…who were prosecuted about putting in a poster about Sam Bass who was a Liberal member for Florey, who had been on a parliamentary junket…There was…a cartoon on it: Sam sitting back with a cocktail in one hand and with his feet up with the cigar…saying to the people in Florey ‘how are things back there?’. Labor won the seat, so he lost his job.

It went to court, they pleaded guilty. He then took civil action against them, got awarded sixty grand. They appealed it to the full Supreme Court. They appealed the $60,000 and the judge said…‘we’ll make it 100’. It went to the High Court and the High Court threw it out … because federally there’s no truth in advertising stuff, there’s Federal cases that say you can really say anything about anyone during a political campaign.

While it can be argued that South Australia has established best practice for ‘truth in political advertising’ legislation, the limited success in prosecuting the Bass case indicates weaknesses in the law. The general reluctance in recent years of both Labor and Liberal governments to implement electoral reforms suggests that a major embarrassment, such as occurred with the creation of the ‘tablecloth’ ballot paper for the New South Wales 1999 Legislative Council election (discussed earlier in this chapter), may be required before significant reforms occur.

45 The Commonwealth Electoral Legislation Amendment Act 1983 introduced a prohibition on untrue, misleading or deceptive advertising. However this provision was repealed in 1984 (see Williams, 1997). 46 Roberts v Bass [2002] HCA 57.

159

Tasmania

In the past 25 years, there have been three major instances of electoral reform in Tasmania. In 1985, the Electoral Act 1985 was passed, replacing the Electoral Act 1907 . The main purposes of this Act were to consolidate amendments that had occurred in the previous seven decades, to modernise the language, and to introduce a party registration regime. However, it is the two more recent reforms, in 1998 and 2004, that provide interesting examples of the catalysts, drivers, and processes of electoral reform. The first, motivated by partisan interests, was the Parliamentary Reform Act 1998 , which reduced the size of the House of Assembly from 35 to 25 members, and the Legislative Council from 19 to 15 members. The second, the Electoral Act 2004 , was driven by the desire for improved efficiency and professionalism in the electoral administration, with neglible, if any, partisan advantage. As will be discussed, these two sets of reforms resulted in vastly different impacts on the fairness of elections for the Tasmanian parliament.

Reform for Political Advantage

The Parliamentary Reform Act 1998 was the culmination of almost two decades of public debate on reducing the size of the Assembly. From the early 1980s –– virtually from the time the Greens first won representation in the Tasmanian parliament in 1982 – – there had been suggestions that the House of Assembly should be reduced in size from 35 members. Associated with this debate were calls for a complementary reduction in the size of the 19-member Legislative Council, in keeping with the tradition of having the upper house roughly half the size of the lower house. Table 7.5 details the various attempts to bring about change during this period. It is interesting to note that the Greens held the balance of power in the Assembly from 1989 to 1992, and again from 1996 to 1998. It can be seen in the table that it was during this second period of balance of power, when the Liberal party was in minority government, that efforts to reduce the size of the Assembly increased.

160

Table 7.5 Reduction in Size of Tasmanian Parliament –– Reform Process Reform Proposal Election outcome Year action and (Assembly) outcome Liberal–Labor–Greens 1982 General election 19 – 14 – 1 (Independent Green) (+ 1 Democrats) 1983 Advisory Committee – Recommends against any Ogilvie Report (1984) reduction 1986 General election 19 – 14 – 2 (Independent Greens) 1989 General election 17 –– 13 –– 5 1992 General election 19 –– 11 –– 5 1993 Liberal govt legislation Recommends bicameral parliament with reduction from 35 to 30 Assembly MHAs. Proposal lapses. 1994 Board of inquiry –– Recommends against a reduction Morling Report but, if size is to be reduced, it should be a unicameral parliament of 44 – 4 x 7-member electorates; 16 x single-member electorates 1995 Labor opposition Recommends bicameral legislation parliament of 25 MHAs and 15 MLCs. Proposal lapses 1996 General election 16 –– 14 –– 4 (+ 1 Independent) 1997 Tasmanian Chamber of Recommends unicameral Commerce and parliament of 40 –– 5 x 5-member Industry proposal electorates; 15 single-member electorates 1997 Commonwealth-State Recommends unicameral Inquiry –– Nixon parliament of 27 –– 3 x 9-member Report electorates 1997 Liberal govt propose Recommends unicameral referendum parliament of 40 –– 4 x 7-member electorates; 12 single-member electorates. Proposal lapses 1997– Various attempts at Nixon report not supported –– 98 reducing the size of Liberal govt wants 28 MHAs, parliament based on Morling report, Labor opposition wants 25 MHAs May Labor opposition Proposal for bicameral parliament 1998 introduces Bill –– –– 25 MHAs and 15 MLCs. Bill Liberal Premier Rundle passes in two days with Labor and recalls parliament in Liberal support July to debate Bill 1998 General election 10 –– 14 –– 1 Assembly size reduced to 25 members 2002 General election 7 –– 14 –– 4 2006 General election 7 –– 14 –– 4 Note: Greens were first elected in 1982. Sources: Tasmanian Parliament, 2007; ABS, 2002.

161

The reduction in Assembly electorates in 1998 from seven to five members increased the required quota from 12.5 per cent to 16.7 per cent, making it far more difficult for minor parties such as the Greens to win seats. This was evident in the election held a month after the reform was implemented, in which the Greens managed to retain only one of their four seats. However, in the subsequent elections of 2002 and 2006, the Greens recovered to win four seats on both occasions. Political commentary at the time of the reform made it clear that the change was primarily designed by the two major parties to remove, or at least reduce, the influence of the Greens. Comments from interviewees support this popular view. Comments from Tasmanian politicians included: [NK –– Was that simply a matter of diminishing the role of the Greens?] Yep, that’s my view –– to increase the quota from about 11 per cent to 16. It worked the first time and didn’t work the second.

It was politically expedient at the time.

The real reason was that some idiot had told [Premier] Rundle and [Opposition Leader] Bacon that this would be the way to make it very difficult for the Greens to win seats. We’ve got a ridiculously small parliament. It’s not working. It hasn’t got the numbers to work.

The impact of the reduction in size of the Assembly can be observed using Gallagher’s Least Squares Index, as illustrated in Table 7.6. The proportionality Index for the five general elections prior to the reform averaged 4.14, while the average increased to 6.48 at the three elections since the reform. This indicates that the reform has had a negative impact on representation.

Table 7.6 Tasmania –– Gallagher’s Least Squares Index, Legislative Assembly Election Gallagher’s Least Squares Index 1982 6.71 1986 4.25 1989 3.02 1992 2.82 1996 3.88 1998 9.93 2002 3.79 2006 5.73 Source: TEC. See Appendix I for calculations.

162

While voter behaviour can also have an impact on proportionality, an increase in the quota makes it more likely that such negative impacts will occur. Furthermore, as noted in an interviewee’s comment earlier, this partisan-driven reform has reduced the capacity for the parliament to effectively carry out its functions of legislative review and executive scrutiny, and to act as a channel for community participation –– a result of the 26 per cent overall decrease in the size of the parliament from 54 to 40 members.

Reform for Efficient Electoral Administration

More recently, Tasmania’s Electoral Act 1985 has undergone a comprehensive review process. This process included two rounds of public consultation, resulting in the repeal of the 1985 Act and the introduction of the Electoral Act 2004 . The review process, depicted in Figure 7.1, provides an excellent model for legislative reviews that could be used by other jurisdictions. In 1999, the Chief Electoral Officer, David Farrell, and his Deputy (now Commissioner), Bruce Taylor, initiated a review of the Act. After gaining Cabinet support for a new Act to be drafted, a lay draft was developed by the Tasmanian Electoral Office (TEO) and a discussion paper released for public comment in December 1999. The proposed changes had three main aims: to utilise contemporary legislative drafting; to adopt modern electoral practice; and to remove detailed procedures and forms from the principal legislation. The reform did not seek to make changes to the main components of the voting system –– that is, five five-member electorates using the Hare-Clark system for the Assembly, and single-member electorates for the Council. The intention was to modernise, simplify, and clarify (Farrell, 2000), as one administrator stated, we went from 350 pages down to 213 pages .

Twelve submissions were received in response to the first paper, and a second discussion paper, which detailed potentially contentious matters, was released in December 2000. Following further public input, the formal drafting process commenced. There was a hiatus in activity due to the 2002 general election, and a transition in the Chief Electoral Officer position, before the Government established an all-party parliamentary working group in April 2004 to consider a draft bill. The new legislation was introduced into parliament in August 2004, and easily passed in late 2004, with the Liberal opposition only expressing concern over two issues: prisoner voting rights; and the increase in candidate nomination fees (Hodgman, 2004: 72–81).

163

Figure 7.1 Tasmanian Reform Process (1999–2004)

Tasmanian Electoral Office (TEO) initiated review (1999) ↓↓↓ Cabinet approved drafting of new Act ↓↓↓ TEO released discussion paper (December 1999) ↓↓↓ Public comment –– 12 submissions ↓↓↓ 2nd discussion paper released (December 2000) ↓↓↓ Further public comment ↓↓↓ Bill formally drafted (2001–2004) ↓↓↓ Draft bill considered by a parliamentary working group (April 2004) ↓↓↓ Bill introduced to parliament (August 2004) ↓↓↓ Passed by parliament (November 2004) ↓↓↓ Receives Royal Assent (17 December 2004) ↓↓↓ Act commences Tasmanian Electoral Commission established (16 February 2005)

These two instances of Tasmanian electoral reform arguably provide the most extreme examples of worst practice and best practice in all of the Australian jurisdictions. Although the short time period between the two reforms occurring is indicative of the resilience of the system, it is worth noting that in both cases, it was the major parliamentary parties that have benefited –– in the first instance, by restricting an unwanted competitor; in the second, by protecting the status quo.

Australian Capital Territory

It is an uncommon occurrence for a new self-governing jurisdiction to be created in Australia and, when this does occur, it presents an excellent opportunity to establish democratic best practice based on accepted norms for electoral management and fair

164 elections. The Australian Capital Territory (ACT), the newest of Australia’s legislative jurisdictions, 47 provides Australia’s most convincing example of community participation in determining a jurisdiction’s electoral system –– a non-binding referendum held in 1992 to decide the voting system, and a binding referendum in 1995 to entrench the system. In addition, the ACT is Australia’s leader in the use of electronic voting. Both of these issues are assessed in this case study.

The ACT has always used a proportional representation voting system for its 17- member Assembly, as had been the case for its advisory assemblies prior to self- government. Initially, the Legislative Assembly elections used a single Territory-wide electorate but since the 1995 election, the Territory has been divided into three electorates: two of five members each; and one of seven members. The 1989 and 1992 elections used a modified d’Hondt system, before changing to a Hare-Clark system for the 1995 and subsequent elections. In conjunction with the 1995 election, a referendum was held to entrench aspects of the Hare-Clark system, including: compulsory voting; electorates having an odd number of members and a minimum of five members; full preferential voting; Robson Rotation; 48 and no ticket voting. 49 The referendum was successful, meaning that any future changes to these aspects of the system would require either a simple majority of the Assembly and a majority of electors at a referendum or a two-thirds majority of the Assembly.50 Reforms have refined the system and generally appear to have been driven by administrative reasons, rather than partisan advantage. Major electoral reforms, results and issues for the ACT are shown in Table 7.7.

47 The ACT held its first self-government election for the Legislative Assembly on 4 March 1989. There have been a further six elections, in 1992, 1995, 1998, 2001 and 2004 and 2008. 48 Robson Rotation is the method of printing multiple ballots for an electorate, with candidates listed in different orders to minimise any advantage a candidate listed first may receive. The method was designed by Neil Robson, a Tasmanian parliamentarian. 49 Section 4, Proportional Representation (Hare-Clark) Entrenchment Act 1994 . 50 Section 5, Proportional Representation (Hare-Clark) Entrenchment Act 1994 .

165

Table 7.7 Australian Capital Territory –– Significant Electoral Events Date Action Election outcome Liberal – Labor – Others October –– Legislation providing for ACT self-government November passed by the Federal parliament. The 1988 Australian Capital Territory (Electoral) Act 1988 stipulates three-year terms and fixed date elections, using proportional representation with a single electorate of 17 members. 4 March 1st general election using modified d’Hondt 4 –– 5 –– 8 1989 15 February 2nd general election (modified d’Hondt). A 6 –– 8 –– 3 1992 referendum held on the same day supported a change to the Hare-Clark proportional system (in preference to single-member electorates). 18 February 3rd general election, the first under the Hare- 7 –– 6 –– 4 1995 Clark system, shifting from a single electorate to three electorates of five, five and seven. Referendum held to entrench aspects of the Hare-Clark system: passed by 56.8% of all electors entitled to vote. 1 1997 The Electoral (Amendment) Act 1997 moves the date for Assembly elections from February to October, to take effect from the 2001 election. 21 February 4th general election 7 –– 6 –– 4 1998 20 October 5th general election 7 –– 8 –– 2 2001 2003 The Electoral Amendment Act 2003 extends terms from three to four years, from the 2004 election. 29 October 6th general election: the first time a party wins 7 –– 9 –– 1 2004 government with an absolute majority. 18 October 7th general election. 7 –– 6 –– 4 2008 1 To be successful, the referendum required a majority of all electors entitled to vote. Of those who actually voted, 65.0% supported entrenchment.

In the ACT’s first election in 1989, there were 117 candidates, with 22 parties and nine independent candidates. There were several reasons for the large number of parties, including the fact that there was no requirement for a minimum number of members. In addition, any one person was allowed to be the registered officer for several parties (one individual was the registered officer for six parties). Parties (but not independent candidates) were able to lodge voting tickets (JSCEM, 1989: 80, 89). As a result, a large number of new parties contested the election. Many of these were considered ‘frivolous’ and accordingly attracted low percentages of votes. Examples included Home Rule OK (with 0.04 per cent of the vote), Sleepers Wake (0.12 per cent), Sun- Ripened Warm Tomato (1.17 per cent), Party! Party! Party! (0.69 per cent), and the

166

Surprise Party (0.12 per cent). Reflecting the community’s mixed feelings about the move to self-government, four of the 17 elected members were from the No Self Government Party (11.5 per cent of the vote) and the Abolish Self-Government Coalition (7.5 per cent). The ballot paper measured 35 centimetres by 102 centimetres.

Table 7.8 Participation in ACT elections Election Candidates Parties Independents 1989 1 117 22 9 1992 1 83 11 6 1995 2 74 6 12 1998 2 108 10 20 2001 2 94 11 17 2004 2 77 10 10 2008 2 86 8 11 1 Modified d’Hondt voting system 2 Hare-Clark voting system Source: ACTEC. Note: Independent groups (e.g. Rugendyke Independents, Osborne Independent Group) are regarded as parties

As Table 7.8 shows, there has been a reasonably consistent level of electoral competition in the ACT over six elections. However, in the 2001 and 2004 elections, there was a reduction in the number of minor party and independent candidates successfully elected, to the current average of 2.3 (for 2001, 2004 and 2008)51 from the previous average of 4.8 (1989–98).

Since the inception of the ACT Assembly, there has been regular debate on its size. Partisan advantage is at the heart of these debates on whether, and to what level, the size of the Assembly should be increased. A 1998 governance inquiry, the Pettit review, recommended that the Assembly be increased from 17 to 21 members. Democratic arguments put forward in support of an increase included that ‘an increase in the size of the Assembly is important just to provide a greater pool of people for the Ministry and the non-executive element’ (Pettit, Keady and Blick, 1998: 39). In the following year, an Assembly committee inquiry recommended against an increase, noting that the proposed change was unpopular with the public (SCRRG, 1999: 11). Proposals to increase the size of the Assembly –– implying more politicians and therefore increased costs –– always tend to be unpopular with the wider public, and the tone of the committee report indicates concern for a voter backlash if such an increase occurred.

51 Four Greens members were elected at the 2008 election.

167

The dominance of partisan advantage is more clearly expressed in a dissenting report from a 2002 inquiry: ‘It is painfully obvious that Members who support this minor increase [to 21 members] do so from an ill conceived intention to protect personal and party interests’ (SCLA, 2002: 63). The Australian Democrats, Greens and the Liberal party all supported a model of 21 members (three electorates of seven members). Labor, however, supported an increase to 25 members (five electorates of five members). There is a clear advantage for Independents and minor parties such as the Greens and Democrats in having electorates with a greater magnitude, as this increases proportionality and improves their chances of winning seats. Labor sees advantages in having five-member electorates. Reasons for this were explained by one interviewee: Labor thinks it has a much better chance of getting three out of five seats than it has of getting four out of seven seats, so it thought it had a better chance of getting majority government with a five-by-five than three-by-seven and the Liberals probably thought that, more so, and thought that’s why they’d rather have three-by-seven than five-by-five. The Greens and Democrats would prefer seven than five because they’re more likely to get seats so that’s where it potentially bogged down, because they couldn’t agree.

Due to the constraints of the Commonwealth self-government legislation, any increase in the size of the Assembly needs to be approved by the Federal government.52 This has been a restriction for the Stanhope Labor Government. Despite enjoying an absolute majority in the Assembly from 2004 to 2008, the Stanhope Government was unable to get support from the Liberal Federal Minister, possibly due to the Liberal party’s concerns that an increase could advantage Labor. The recommendation of the 1998 Pettit review, and the acknowledgement that the current size of the Assembly creates logistical problems for Cabinet and committee structures, appears to be overlooked in the concern about partisan advantage.

Electronic Voting

The use of electronic voting can have significant democratic benefits, including increased accessibility for people with a disability and voters living in remote areas. It also enables more rapid and accurate counting of votes. Electronic voting was introduced for the 2001 ACT general election. The process for initiating the enabling legislation provides insight into the importance of personal relationships in achieving

52 Section 8, Australian Capital Territory (Self-Government) Act 1988 .

168 reforms, irrespective of whether there are partisan implications. The process was described by the ACT’s Electoral Commissioner as follows: Electronic voting and counting was a huge coup…The main reason we succeeded with that one was that after the 1998 election we had to do a recount in Molonglo, when we had three or four votes difference between two candidates and we did a full recount, and actually worked out we got it wrong the first time around, and we actually changed the results as a result of the recount. I was wanting to do something that was more accurate and Kate Carnell got in as Chief Minister at that election.

At the declaration of the poll we started talking about electronic voting and counting. I said I would love to do electronic voting and counting but its going to cost a bomb. I can’t imagine you would be prepared to pay that. She said, ‘yes, we would –– do it’. That’s essentially where all that came from.

At the 2004 election, 13.4 per cent of all votes were lodged electronically, an increase from 8.3 per cent at the 2001 election.53 While the electronic voting system allows voters to make an informal vote, accidental informal voting is almost entirely eliminated, as the system issues a warning to the voter that their vote would be informal if they continued (ACTEC, 2005: 15). The impact of electronic voting can be seen in the reduced level of informal voting in the 2001 and 2004 elections compared to votes cast by paper ballots (Table 7.9).

Table 7.9 Use of electronic voting and informality rates in the ACT Election Paper Informal Electronic Informal votes rate (%) votes rate (%) 2001 182,162 4.3 16,559 0.6 2004 181,580 2.9 28,169 1.1 Sources: ACTEC, 2002, 2005.

The reduction of informal voting has democratic implications for voters –– with unintentional informal voting virtually being eliminated, fewer voters are accidentally disenfranchised. It could be argued that the particularly low level of informal electronic voting in 2001 (0.6 per cent) was due to electronic voters in 2001 being a self-selecting sample of voters who opted to use the new system, who were possibly more committed to lodging a formal vote than other voters. In 2004, electronic voting was the more common form of voting at booths where this was possible: the informal rate in the 2004

53 The use of electronic voting at the 2008 election was approximately 20 per cent. Exact figures and informality rates were not available at the time of writing.

169 election therefore more accurately reflects the impact of electronic voting on the general population.54

It can be seen from these examples that the ACT has been quite innovative in its electoral reform processes. One vital reason for this is the role of path dependency –– that is, the importance of the original design of a system in determining future directions (as discussed in Chapter 2). In contrast, path dependency has taken the Northern Territory along a different road of reform.

Northern Territory

Of all Australian jurisdictions, the Northern Territory has probably the most challenging demographic and geographic factors to deal with in an electoral sense: a small but highly mobile population dispersed over a large area (apart from urban concentrations in Darwin, Alice Springs, and to a lesser extent, Katherine); a large Indigenous population, especially in remote areas; and a high percentage of people whose first language is not English. From its establishment, the unicameral Northern Territory Legislative Assembly has been elected by a majoritarian system. From the foundation of this system, path dependency has resulted in government rarely changing hands, and an equally low level of electoral reform. Of particular interest in this case study is the reform process that resulted in the Territory’s introduction of a party registration regime.

The Northern Territory was granted self-government in 1978 55 and the first general election under self-government was held in 1980. The Territory’s Assembly originally consisted of 19 members, increasing to 25 members at the 1983 election. The Country Liberal Party 56 won government at the 1980 election, and retained government at a further five elections, until being defeated by Labor at the 2001 election. The electoral laws during this time were based on the Commonwealth Electoral Act 1918 and

54 Personal correspondence, ACT Electoral Commission, 10 November 2008. 55 By the Commonwealth’s Northern Territory (Self-Government) Act 1978 . The Territory had a fully- elected Legislative Assembly from 1974; however, powers largely remained with the Commonwealth until 1978. 56 The Country Liberal Party only exists in the Northern Territory, and is an amalgam of the Liberal and National parties.

170 remained largely unchanged for two decades, apart from some minor amendments in 1995 (MEC, 2003: 7). The Labor party had a long-held position calling for electoral reform and central to this was the establishment of an independent electoral administration. On its election in August 2001, the Martin Labor Government committed to a review of electoral law, which commenced in April 2003 with the appointment of Minter Ellison Consulting. As an electoral official explained: They had an electoral office which was effectively in the Chief Minister’s department. One of the things that Clare Martin wanted to do was to have an independent electoral commission because there was a perception that the electoral office was a bit too cosy with the government of the day. It’s not a good look.

The review team included two former Federal electoral commissioners, Bill Gray and Colin Hughes, and the now current Australian Capital Territory commissioner, Phillip Green. The review was conducted over six months, and included three consultation phases: a discussion paper (30 submissions); public meetings (37 attendees at five meetings); and a draft report (23 submissions) (MEC, 2003: 10, 15–7). The final report was presented to the Government in October 2003 and new legislation was introduced to the Assembly in November 2003. In February 2004, this legislation was largely passed without amendment, the main exception being a late change to the party registration regime, which is assessed below.

Determining the Northern Territory’s Party Registration Regime

As discussed more fully in Chapter 8, political party registration schemes have developed in the various Australian jurisdictions since the early 1980s. Registration schemes can have a significant impact on the promotion of democratic principles and the conduct of fair elections. They can be used to assist or hinder electoral competition –– registration schemes rarely have a neutral affect. Prior to the 2004 reforms, political parties were not recognised in the Northern Territory’s Electoral Act. Reasons for this included the lack of a public funding scheme and the fact that party names were not shown on ballot papers prior to the 2005 election (candidates’ photographs were shown; see Figure 7.2).

171

Figure 7.2 Northern Territory Ballot Paper (from Katherine by-election, 2003)

The 2003 Minter Ellison review recommended that, as with all other Australian jurisdictions, ballot papers should show party affiliations, and therefore a party registration regime should also be introduced (MEC, 2003: 69, 85). The review considered the question of a reasonable membership level, drawing on examples in other Australian jurisdictions. In its report, the review team recommended that the threshold be 20 members, noting that they had received evidence that a higher threshold requirement of 50 members may prevent some of the larger parties from registering (MEC, 2003: 85). While the figure of 20 members may appear to be very low, this was at a time when there were only about 112,000 registered voters in the Territory.

172

When the Government introduced its legislation in November 2003, it stipulated a 50- member registration requirement. The day before debate on the legislation commenced in February 2004, the Government lodged an amendment, changing the figure to 200 members, a 10-fold increase on the original number recommended by Minter Ellison. The only explanation that the Government gave for increasing the proposed threshold was the need to protect access to the electoral roll (Martin, 2004), which was available to registered parties but not to other parties or candidates (except by viewing at electoral offices). The late change resulted in party registration being the most hotly-debated issue of the new electoral framework (McNeill, 2004: 4), as the following Hansard extracts (NTLA, 25 February 2004) attest: [N]ow the Labor Party has come to power, the power has seduced so that, by the utilisation of this mechanism with 200 required to formulate an official party, it basically takes it out of the reach of community groups and puts it largely into only the domain of more established parties such as the ALP or CLP. (Terry Mills, CLP)

This is a deliberate attempt to obliterate opposition by the Labor government (John Elferink, CLP)

[T]his has been brought in to kill off political opposition. I just think this is a crying shame…This is anti-democratic…supporting this bill means that I will sign the death warrant for any small party that wants to start in the Northern Territory…This is putting politics ahead of principle. (Gerry Wood, Independent)

The high threshold required to register parties in the Northern Territory is reflected by the low number of registered parties, with only three –– the Country Liberal Party (CLP), Labor, and the Greens –– contesting the 2005 and 2008 elections. It can be argued that this registration scheme restricts citizens from being able to organise into political groups to access the political process. However, statistics from the Northern Territory elections held prior to the party registration legislation being passed in 2004 do not support this (see Table 7.10). Before party registration, an average of 3.8 parties contested each general election from 1983 –– similar to the three parties contesting the two most recent elections. The number of candidates ranged from 63 to 85 during the earlier period, compared with 80 and 66 in 2005 and 2008 respectively. To date, the

173 legislation has not resulted in any significant difference in terms of electoral competition by parties and candidates. 57

Table 7.10 Candidates and parties contesting Northern Territory elections Election Seats Candidates Parties 1983 25 66 3 1987 25 85 3 1990 25 80 4 1994 25 63 3 1997 25 66 4 2001 25 86 6 2005 25 80 3 2008 25 66 3 Source: NTEC; AGPD.

With no evidence to the contrary –– apart from Chief Minister Martin’s claim regarding access to the electoral roll –– it would appear that Labor’s move in 2003/04 to ignore the review’s 20-member recommendation, and increase the threshold first to 50 members and then to 200 members, was purely a manoeuvre to damage its major opponent –– the CLP. This is borne out by the CLP’s strong criticism during debate on the legislation, and is supported by the comments of one electoral official: It’s a bit subjective. There was some suggestion that even the CLP was struggling to get 200. If they were governing for 27 years, and they’ve only been out of power for a few years, you’d have to say if that was the case, you should probably err on the side of not being too demanding.

While the Territory’s reform process appears to have been quite comprehensive and broadly consultative, it was restricted due to the powers that the Commonwealth retains over certain aspects of the Territory’s electoral system. Electoral issues beyond the scope of the Minter Ellison review included the manner of representation –– the Commonwealth’s Northern Territory (Self-Government) Act 1978 stipulates that there are to be single-member electorates, each having no greater than a 20 per cent variance (higher or lower) from the average enrolment. In addition, it is not possible for the Territory to specify the extent of the franchise (section 14 of the Commonwealth Act) or extend the period between elections (section 17). Whether the Territory can legislate for fixed-date elections is unclear, with the Territory’s Solicitor-General being unable to give a definitive position in 2003 (MEC, 2003: 45). There also appears to have been a

57 It is also interesting to note that two candidates were elected unopposed at the 2008 election (the seats of Arnhem and MacDonnell), a rarity in modern Australian elections.

174 desire for the Minter Ellison review to satisfy the Labor Government. As one review team participant noted: [The Government] ended up adopting pretty much everything that was recommended, but the consultant team from Minter Ellison was very keen to put up proposals that would be smiled upon, so whenever we met anything that was likely to be contentious we said ‘here are the options, you choose’.

In the Northern Territory, the primary catalyst for electoral reform in the past 25 years has been the 2001 change of government. Once Labor achieved government, it immediately set about reforming the system (and particularly the administration) which was viewed as being too close to the former Country Liberal party. A year after the reform process was completed, Labor won the 2005 election with an increased majority. Whether the reforms serve to entrench Labor in power is hard to determine. As the reforms were largely administrative, it is doubtful that there is any significant partisan advantage, except for the party registration regime, which is contrary to the participation principle of fairness.

Summary

This chapter has identified a number of drivers behind electoral reform. First, as was also shown in the previous chapter on Commonwealth reforms, the self-interest of parties, either acting alone or as a cartel, is a significant driver. The clearest example of this is the cartel behaviour of the Labor and Liberal parties in reducing the size of the Tasmanian parliament in 1998. While it appears that the Greens have been able to overcome this barrier at the two most recent elections, they (and any other minor parties) will continue to have to achieve a quota more than four per cent higher than was the case at elections prior to 1998. It is also interesting to juxtapose the Tasmanian move to a smaller parliament with the ACT’s debate on enlarging its Assembly. In both cases, partisan advantage has been the major reason for action (in the first instance) and inaction (for the latter).

While it can be argued that Western Australia’s moves to proportional representation for its Legislative Council is consistent with accepted norms for fair elections, the partisan imperatives at play during both waves of reform in that State have produced less than ideal outcomes. Labor–National party cartelisation was evident in the 1987 reform, but

175 was less pronounced in the 2005 Labor–Greens–Cadby agreement. The latter reform actually created an electoral disadvantage for the Greens. In comparison, the Bracks Labor Government’s introduction of proportional representation for Victoria’s upper house has disadvantaged the governing party, thereby providing a rare example of a governing party acting in order to create fairer elections rather than seeking a partisan advantage.

The Northern Territory’s party registration regime presents an interesting example of how a reform process based on an independent expert review with community consultation can be corrupted by the self-interest of the governing party. The Martin Labor Government’s overriding of the independent review’s recommendation for a 20- member threshold for party registration has created a major barrier for the creation of new parties in the future.

The ability of electoral management bodies to initiate reform is limited. There are good examples of proactive electoral commissions recommending and driving change, such as the major review that was initiated in Tasmania in 1999 and the ACT’s introduction of electronic voting for the 2001 general election. At other times, commissions need to quietly encourage change, as the Victorian and South Australian commissions have done in their reports to parliament. Alternatively, commissions can be reactive, responding to problems produced by existing electoral law such as the multitude of parties contesting the New South Wales 1999 Legislative Council election.

Queensland’s Fitzgerald Inquiry provides the best example of a driver for reform coming from outside the institutional structure of electoral management. Unfortunately, the Inquiry was brought about by major corruption in the State and, while there have been some positive outcomes from the reform process, the combination of single- member electorates with a unicameral system regularly delivers highly disproportional outcomes. The growth of ‘1 only’ voting is another worrying trend.

The importance of path dependency in shaping electoral systems is evident in the case studies of the newer jurisdictions, where patterns of reform are clearly based on their origins. In the case of the ACT, minority government brought about by proportional representation has been conducive to regular reform in the 20 years of self-government. In comparison, the Northern Territory’s pattern of majority governments, based on

176 single-member electorates, has stifled opportunities for reform. While the CLP was continuing to win elections –– six in a row since the first self-government election in 1980 –– there was no imperative to change the system. In the same way, Labor had no need to accommodate concerns about implementing its restrictive party registration regime following the change of government in 2001.

As discussed in Chapter 5, Australian electoral management bodies generally do not have independent powers to develop their own regulatory frameworks, and are therefore heavily reliant on a supportive government if change is to be achieved. In this sense, institutional design does not provide protection against governing party control, as was seen in Chapter 6 in regard to party involvement in postal voting processes at the Commonwealth level. The ACT’s process, in which voters were asked to determine the voting system by referenda (albeit with limited choices), is encouraging in terms of democratic participation. This process also stands in stark contrast to reform processes at the Federal level. Except for the Australian Constitution’s requirements that parliamentarians be ‘directly chosen by the people’,58 that equal Senate representation for the original States be maintained,59 and that there be a minimum number of seats for each State in the House of Representatives,60 the Federal parliament is virtually free to devise an electoral system of its own choosing.

Chapters 6 and 7 have taken a case study approach in discussing the impacts of reforms in each of Australia’s nine jurisdictions in relation to fairness principles. The following chapter continues this approach by addressing four aspects of electoral systems that have contemporary relevance across a majority of jurisdictions.

58 Sections 7 and 24. 59 Section 7. 60 Section 26.

177

Chapter 8

Reform Processes III –– Cross- Jurisdictional Issues

The previous two chapters have applied a case study approach to assess reform processes and impacts in each of the nine Australian jurisdictions. This assessment highlighted reform issues specific to each jurisdiction. In contrast, this chapter focuses on a selection of electoral management matters that have broader relevance across all jurisdictions. Based on an analysis of the interview data, four issues –– voter turnout; prisoners’ voting rights; party registration; and informal voting –– were selected for comparative analysis (Table 3.6), due to their ongoing importance in contemporary electoral administration. These issues are assessed across the jurisdictions to determine whether Australian practice is consistent with internationally accepted norms for electoral management and fair elections, in terms of fairness to voters and parties. In addition, the drivers for reform in these areas are identified to determine the roles of partisan self-interest and party cartelisation in the reform process. While aspects of these four issues have been discussed earlier, this chapter adopts a comparative analysis in assessing best practice and identifying areas where a uniform approach to reform may be merited.

Voter Turnout

The level of voter turnout in democratic elections is an indicator of the health of a democracy. An election is the primary interaction between citizens and government, and a decline in turnout levels can undermine the legitimacy of a democratic system

178

(Ellis et al., 2006: 12). Compulsory voting is in place for all Australian jurisdictions,61 and it may be expected that voter turnout would not be a significant issue. An assessment of turnout in the 66 general elections that have been conducted in Australia during the 25-year study period demonstrates this is the case, except in the Northern Territory. As Table 8.1 shows, turnout at Northern Territory elections is consistently more than 10 per cent lower than all other jurisdictions, which regularly have a voter turnout above 90 per cent. It is also important to note that turnout figures compare the numbers of people who actually voted with the number of people enrolled, and not with the number eligible to be enrolled, which is likely to be greater.62 With otherwise consistently high rates under compulsory regimes, a drop of a few per cent can be a cause for concern. The figures therefore indicate that there is a serious issue to be addressed in the Northern Territory.

Table 8.1 Turnout Rates, Australian Elections 1983–2007 Turnout (%) Average Jurisdiction 1983 2007 turnout (%) Commonwealth 94.6 - 94.2 - 93.8 - 95.3 - 95.8 - 95.8 - 95.0 - 94.9 - 94.3 - 94.8 94.8

NSW 92.5 - 93.6 - 93.6 - 93.8 - 93.1 - 91.9 - 92.6 93.0

Victoria 93.2 - 92.4 - 95.1 - 94.1 - 93.2 - 93.2 - 92.7 93.4

Queensland 91.7 - 91.3 - 91.2 - 91.5 - 91.4 - 92.9 - 92.6 - 91.4 - 90.5 91.6

Western Australia 89.0 - 91.4 - 90.7 - 93.5 - 90.0 - 90.6 - 89.8 90.7

South Australia 93.5 - 94.4 - 93.6 - 91.8 - 93.6 - 92.3 93.2

Tasmania 94.2 - 93.1 - 95.6 - 96.0 - 95.0 - 93.7 - 95.0 94.6

ACT 88.8 - 90.3 - 89.5 - 92.6 - 90.9 - 92.8 90.8

Northern Territory 81.6 - 71.2 - 81.6 - 80.7 - 79.0 - 80.6 - 80.1 79.2 Sources: Electoral commission figures; AGPD.

61 The only significant difference in the compulsory regimes occurs in South Australia, where it is not compulsory for citizens to enrol, but once enrolled, it is then compulsory to vote. All other jurisdictions have compulsory enrolment and voting. 62 The ability to get eligible people onto the electoral roll is a related issue. A discussion of this is included in Chapter 5 – The ATSIEEIS Case. See also Brent and Jackman, 2007.

179

The Northern Territory –– Indigenous Enrolment and Turnout

A comparison between Northern Territory Assembly election turnout and voting in the Northern Territory at Federal elections (Table 8.2) shows that turnout is consistently lower at Assembly elections. However, there has also been a notable drop in Federal election turnout figures in the Northern Territory during the 2000s (average 85.6 per cent), compared to the 1990s (average 89.4 per cent). The Northern Territory has the highest proportion of Indigenous people in Australia (31.6 per cent), significantly higher than the second-highest-ranking jurisdiction, Western Australia (3.8 per cent) (ABS, 2007). One possibility for the drop in turnout at Federal elections in the Northern Territory could be the abolition of the ATSIEEIS (discussed in Chapter 6). 63

The ATSIEEIS program was concentrated in the Federal electorate of Lingiari, which takes in all areas of the Northern Territory outside of Darwin. Lingiari, established when the Northern Territory was split into two seats for the 2001 Federal election, has the highest concentration of Indigenous voters in the country (43.5 per cent). This is more than double the next highest concentration –– 18.2 per cent in Kalgoorlie, Western Australia (Nelson, 2007: 33). Since Lingiari’s creation, it has recorded turnout figures of 80.6 per cent (2001), 77.7 per cent (2004) and 81.3 per cent (2007). Lingiari’s are the lowest turnout rates in Australia.

Table 8.2 Northern Territory Voter Turnout (%) 1983–2007 64 Year Federal House of NT Representatives Assembly 1983 81.4 81.6 1984 85.5 1987 79.9 71.2 1990 89.4 81.6 1993 88.8 1994 80.7 1996 89.1 1997 79.0 1998 90.3 2001 86.1 80.6 2004 84.3 2005 80.1 2007 86.5 Sources: AEC; NTEC; AGPD.

63 Although the ATSIEEIS was abolished in 1996, it can be argued that the positive impact of its programs remained for the 1998 election. 64 Turnout at the 2008 Assembly general election was 75.7 per cent.

180

An analysis of elector turnout at Northern Territory elections presents a similar story, and with its smaller electorates (around 4,000–5,000 enrolled), provides a more nuanced understanding of the issue. Figure 8.1 shows the correlation between Indigenous population and voter turnout at the 2005 general election. The six electorates with the lowest Indigenous populations (average 6.6 per cent Indigenous)65 have an average turnout of 84.2 per cent, while turnout for the six highest Indigenous population electorates (average 70.7 per cent Indigenous)66 averages only 64.4 per cent turnout. The lowest voter turnout was in the Stuart electorate, with an 84.3 per cent Indigenous population and only 59.3 per cent turnout. Generally, voter turnout at by-elections tends to be lower than that for general elections, and at a by-election for the seat of Stuart in 2006, turnout was down to 53.9 per cent.

Figure 8.1 Turnout at the 2005 Northern Territory Assembly general election

Northern Territory Turnout 2005 Election 100.0

90.0 80.0

70.0 60.0 % 50.0 40.0

30.0 20.0

10.0 0.0 1 2 3 4 5 6 7 8 9 10111213141516171819202122232425 Electorate

Indigenous Population % Turnout Rate % Informal Rate %

Note: See Appendix J for electorate data.

65 Port Darwin, Wanguri, Nelson, Nightcliff, Fannie Bay, and Casuarina. 66 Stuart, Arnhem, Arafura, MacDonnell, Barkly, and Nhulunbuy.

181

It is an added concern in the context of the legitimacy of election results if low turnout figures are compounded by high levels of informal voting, resulting in a significant gap between the level of formal votes and the number of people eligible to vote. It can be anticipated that there would be a correlation between literacy levels in remote Indigenous populations and the number of informal votes, and there is a trend over time towards higher informality in high-Indigenous electorates. The average informality rate at the 2005 general election for the six low-Indigenous electorates was 3.3%, compared to an average 5.3% for the six high-Indigenous electorates. Another factor in high- Indigenous areas may be the number of candidates contesting seats. Results from the Assembly seat of Stuart emphasise this. In 2005, when only two candidates stood, the informal rate was 4.6 per cent. The following year six candidates contested the Stuart by-election, and the informality rate increased to 13.6 per cent.

Administrative Response to Indigenous Enrolment and Turnout

As discussed in Chapter 6 in relation to the abolition of ATSIEEIS, the manner in which Indigenous voter enrolment and turnout is dealt with by administrators and legislators provides a good insight into the drivers for reform: that is, whether reforms are driven by the interests of governing parties, or are based on accepted norms for good electoral management. Interviews in jurisdictions with high proportions of Indigenous people consistently identified Indigenous enrolment and turnout as major and ongoing issues. This is especially so in remote areas with low accessibility, which are among those most likely to have high proportions of Indigenous people. A 2003 JSCEM report noted that the AEC, the Aboriginal and Torres Strait Islander Commission, and the Labor Party had all expressed concerns about the under-enrolment of Indigenous people (JSCEM, 2003: 94). The Labor submission to the inquiry estimated that Indigenous enrolment was only 54 per cent of those eligible, compared to 95 per cent of non-Indigenous Australians (ALP, 2002). 67 JSCEM stated in its 2003 report that it would pursue this issue, but the Committee’s report into the 2004 election did not refer to it.

A number of reasons have been suggested to explain low Indigenous turnout. Alport and Hill (2008) cite antipathy, largely brought about by a lack of efficacy in terms of the relevance and impact of votes in determining election outcomes. They argue that the lack of enforcement of compulsory voting in remote areas is another key factor (Alport

67 The ALP estimated 140,000 enrolled out of a possible 260,000 Indigenous people.

182 and Hill, 2008: 3). It is important to note that seats with high Indigenous populations tend to be safe Labor seats. For example, the six high-Indigenous Northern Territory seats referred to earlier are held by Labor with margins of 62 to 76 per cent of the two- party preferred vote. 68 There is therefore no electoral advantage (in fact a disadvantage) for Country Liberal Party governments to encourage greater levels of Indigenous enrolment and participation, and in fact there is a disadvantage. One electoral administrator expressed a view on this issue: I said here’s where we think we should be going with mobile polling places, what do you think? Well the conservative side of politics don’t like that idea, but they’re not going to come out and say, ‘we don’t want to see these people voting’.

Aspects of the Howard Government’s 2006 reforms, especially the earlier close of rolls and the need for proof-of-identity for enrolment and provisional voting, were seen as additional barriers to higher levels of participation among Indigenous people. As Alport and Hill point out, ‘Many Aboriginal people do not carry a wallet, therefore there is nowhere to keep any cards and they are easily lost’ (2008: 6). 69 Furthermore, because these are such safe seats, there is little incentive for Labor governments to take action to encourage participation.

Low voter turnout can also be a reflection of protest against government authority (see Mansell, 1993 and Alport and Hill, 2008: 3) and can further explain low participation for this social group. This view was supported by one interviewee: You would have to work out how much of it was the abolition of the [ATSIEEIS] program and people going around, and how much was a general disillusionment with a particular section of the community with the government of the day. And certainly in the first six to eight years of [the Howard] government there’s been quite a bit of tension between the Indigenous community and the government.

Northern Territory Assembly Elections –– Timeframe

As shown in Table 8.2, voter turnout for Northern Territory Assembly elections is consistently and significantly lower than that for Federal elections in the Northern Territory. A reason for this difference appears to be the short election campaign period between the calling of an Assembly election and polling day. The Territory’s Electoral Act 2004 (section 28) requires that the polling day must be 18 days after the issuing of

68 All except MacDonnell (62%) have at least 71 per cent of the two-party preferred vote. 69 However Alport and Hill go on to suggest that photo ID cards may be useful to encourage enrolment and turnout (2008: 12).

183 the writ. This is the shortest timetable in operation in Australia and as a result, all procedural aspects of the election need to meet tight deadlines. For example, rolls close two days after election writs are issued, nominations close a further four days later, and postal and mobile voting commences three days after that (NTEC, 2008).

In addition to the short timetable for campaigns, elections can be called at any time after the first three years of an Assembly, 70 enabling the Chief Minister to call an election with little warning. Together, these two factors would make it logistically difficult to conduct an election in any jurisdiction, but this is especially so in the Northern Territory which has a large geographic area of dispersed settlement. The 18-day timeframe is more than a week shorter than other Australian jurisdictions without fixed election dates. The Minter Ellison report commissioned by the Labor Government in 2003 was opposed to this short election period, and recommended extending the timetable to 33 days (NTEC, 2007: 5). In addition, the Northern Territory Electoral Commissioner recommended that either fixed-date elections be introduced or the election timeframe be lengthened by at least seven days (NTEC, 2007: 50).

It is evident that the short length of an election campaign period can impact on voter turnout. Problems include accessing people in remote areas for enrolment, especially when they may only have one chance to vote at a mobile polling booth less than two weeks later, and having enough time for postal voters to receive and return their completed ballot papers. In some instances, election dates are known well in advance based on comments from the government. However, the 2008 general election held on 9 August was considered to be a ‘snap’ election, with the announcement made only 19 days earlier. Turnout at the election was 75.65 per cent, the lowest for more than 20 years. Unfortunately, at this stage there has been no action by the Northern Territory Government to extend the election period. As one administrator noted in an interview, the ALP in opposition were very critical of the cramped timetable…certainly a government of any colour prefers to have control.

70 Section 23 of the Electoral Act 2004 .

184

The Prisoner Franchise

Supporters of prisoner enfranchisement argue that voting is a fundamental right, and denial of the vote is racist, as Indigenous Australians are imprisoned at 12 to 15 times the non-Indigenous rate. Also, it is argued that denying a right to vote is counter- productive to the rehabilitative aspect of incarceration (Orr, 2007: 2). Detractors say that prisoners have broken the ‘social contract’ by committing serious crimes, and therefore deserve ‘civil death’ by disenfranchisement (Hill, 2000: 204–5). 71 There are quite divergent voting entitlements for prisoners in democracies around the world, as Massicotte, Blais and Yoshinaka’s (2004) comparative study of 63 countries reveals. In 16 of the 63 countries in the study (25 per cent), all prisoners have a right to vote. This group includes Denmark, Germany, South Africa and Sweden. A further 16 countries (25 per cent), such as the Netherlands, New Zealand, and Spain, allow a partial franchise, dependent on the length of sentence or the type of offence committed. In 24 of the countries studied (38 per cent), including Brazil, India, and the United Kingdom, all prisoners are disenfranchised (Massicotte, Blais and Yoshinaka, 2004: 18–25). 72

A detailed analysis of the philosophical arguments for different countries’ approaches to the prisoner franchise is outside the scope of this thesis. Instead, it focuses on the variation that exists within Australia, and the processes undertaken to determine prisoners’ entitlements at the Commonwealth, State and Territory levels. From this, an assessment is made of the motivating forces driving Australian prisoner franchise laws – – the participation principle of fairness, partisan self-interest, or a mixture of both.

Changes to Commonwealth Laws

As noted in Chapter 6, the Hawke Labor Government extended the prisoner franchise in its 1983 reforms. The franchise had remained largely unchanged since Federation, with voting rights extended only to those imprisoned for an offence with a maximum penalty of less than one year. Labor extended the franchise to those imprisoned for offences with maximum penalties of less than five years (Davidson, 2004: 1). These provisions were difficult to administer, as State Controllers-General of Prisons needed to determine the maximum possible penalties for offences, irrespective of prisoners’ actual sentences.

71 For discussion on the arguments for and against prisoner voting, also see Orr, 1998; Hasen, 2004; and Fitzgerald, 2005. 72 No information was provided for the remaining seven countries (11 per cent).

185

Subsequently, the Keating Labor Government changed the provision in 1995 so that all prisoners sentenced to terms of less than five years could vote. This had the effect of further extending the franchise (Orr, 1998: 58–9).

When the Coalition came to power in 1996, there was an immediate push to not only to restrict the prisoner franchise, but to remove it entirely. In 1997, the Coalition- dominated JSCEM recommended that no prisoners should be entitled to enrol or vote (JSCEM, 1997: 48). The Coalition’s attempt to legislate for this change in 1999 was defeated in the Senate (Norberry and Williams, 2002: 24). However, the Coalition was successful in restricting the prisoner franchise in 2004, reducing the ‘less than five years’ provision to ‘less than three years’. While the Coalition’s legislation called for total prisoner disenfranchisement, Labor argued that a three-year qualification was a more suitable restriction as it coincided with a full term of the Australian Parliament (Faulkner, 2004: 25138). The Coalition agreed to this compromise in 2004 because it lacked the necessary numbers in the Senate to achieve its preferred outcome.

After the Coalition gained a Senate majority in 2005, it returned to its previous position of removing the franchise for all prisoners in full-time detention –– a change estimated to affect approximately 20,000 prisoners (Costar, 2006: 4). While the Coalition has consistently argued against any prisoners having the right to vote, analysis of the interview data indicated some opposition to this stance within the Coalition parties. As one Liberal parliamentarian stated: I was a minority in my own party. The majority of my party had the John Howard, Eric Abetz line that if you were in prison, you didn’t have a right to vote. I don’t take that view. I think that, as the law said previously, if you’re doing more than three years, you didn’t have the right to vote, which is probably fair enough.

But I was dead against taking the right away from someone who was doing a short sentence. I felt that was completely unjust. They’re in prison paying their debt to society. It’s not as though it’s a difficult matter to go and take their vote. They shouldn’t have lost [the vote].

In its 2000 report, the JSCEM stated that the prisoner franchise should not be removed ‘until there is sufficient and widespread public support for a change’ (JSCEM, 2000: 90). The Government asserted that its 2006 move was supported by the Australian electorate; this is not, however, borne out by the submissions received by the Senate Finance and Public Administration Committee’s inquiry into the Bill. Of the

186

39 submissions that addressed the prisoner franchise issue, only three expressed support for the coalition’s proposed change (from the Liberal Party, Nationals, and Festival of Light 73 ). While it would be erroneous to argue that these figures are representative of the general view of the Australian electorate, it is significant that there was minimal explicit support for the change outside the Coalition parties.

There is very little information about public opinion on this issue; however, one newspaper poll in 2007 put public support for prisoners’ voting rights at 62 per cent (albeit of only a small, unrepresentative sample) (The Age, 2007). While such an unscientific survey may be inconclusive, some of the arguments put forward by the Coalition are even more so. During one parliamentary debate on the issue, senior Liberal Senator Nick Minchin argued that the Government position stood up to the ‘pub test’ –– that is, the majority of people in a hotel bar would support the Government’s position –– suggesting perhaps that, apart from its ideological position on the issue, the Coalition also supported the change as a populist (and partisan) measure (Minchin, 2004: 25151).

Many of the submissions to the JSCEM inquiry opposing the change cited Article 25 of the International Covenant on Civil and Political Rights (ICCPR), which states that all citizens shall have the right and opportunity to vote at elections. This argument is consistent with developments in Canada and the United Kingdom in recent years. In Canada, no prisoners have been disenfranchised since the Supreme Court ruled in 2002 that disenfranchisement was in breach of the country’s Charter of Rights and Freedoms. While the United Kingdom disenfranchises all prisoners, in 2004 and 2005 the European Court of Human Rights found this to be in contravention of the European Convention on Human Rights (Sawer, 2006b: 8–9).

The ICCPR also refers to the purpose of prison being a place to rehabilitate offenders back into society (Kerr, 2005: 30). Some JSCEM submissions argued that the denial of voting rights removed the benefits of preparing prisoners for their transition back into society through the restoration of their responsibilities as members of the community. Furthermore, they pointed to the disproportionate impact on young males (the prison population is 93 per cent male) and Indigenous Australians.

73 The Festival of Light is a lobby group promoting ‘Christian values’ and ‘family values’. It has recently changed its name to Family Voice Australia – see www.fol.org.au.

187

The 2007 High Court Challenge

The Howard Government’s law disenfranchising all prisoners was challenged in the High Court in 2007 by Vickie Lee Roach (HCA, 2007a; 2007b), an Indigenous woman who had been sentenced to prison for a total of six years on five burglary-related offences. The Roach case was based on a number of arguments, but primarily that disenfranchisement of all prisoners was contrary to the constitutional requirement that parliament be ‘directly chosen by the people’ (Orr, 2007: 3). In addition, it was argued that disenfranchisement denied prisoners their implied right to freedom of political expression and communication.

In September 2007, the High Court determined (by a 4–2 majority) that the blanket disenfranchisement was unconstitutional. However, it upheld the law previously in effect, that prisoners serving sentences of three years or more would be disenfranchised (HCA, 2007c). The reasoning of the High Court was that the blanket ban was an arbitrary disenfranchisement of a particular class of citizen, without regard to the circumstances in which those citizens became part of the prisoner class.

Justices Gummow, Kirby and Crennan’s joint lead judgement refers to the 2004 legislative change which restricted the franchise to prisoners serving less than three-year sentences. They argued that such a restriction still gave regard to the seriousness of the offence as a measure of fitness to participate in the electoral process (HCA, 2007c: para. 98). This view was supported by Chief Justice Gleeson, who also suggested that a more restrictive disenfranchisement based on a lesser term (for example, prisoners serving terms of one year or more), would not necessarily be invalid (HCA, 2007c: para. 19). However, Gummow, Kirby and Crennan appear to be less positive towards the prospect of a more restrictive disenfranchisement, noting that the three-year sentence threshold mirrors the three-year electoral cycle that is entrenched in the Constitution (HCA, 2007c: para. 102).

Prisoner Franchise at the State and Territory Levels

At the Australian sub-national level, there is a similar divergence of entitlements as occurs internationally. Three jurisdictions (Victoria, Queensland, and the Northern Territory) adopt the Commonwealth standard –– that is, following the High Court’s

188

2007 Roach decision (HCA, 2007c), that prisoners serving terms of three years or less are entitled to vote. The Australian Capital Territory had adopted the Commonwealth standard of disenfranchisement up until 2006. An aspect of the Howard Government’s reforms however, was that while prisoners were disenfranchised, they remained on the electoral roll. The ACT’s Electoral Act 1992 states that all enrolled electors are entitled to vote, 74 and the Howard reforms, perhaps unintentionally, actually extended the vote to all prisoners in the Territory. As one administrator noted: The ACT made it known to the Commonwealth when they were talking about taking the vote away from prisoners, that the ACT thought, given their human rights stance, that all prisoners should have the right to vote…I don’t know who came up with the idea…which was that everyone in prison can enrol, but if you’re in prison full-time you can’t vote for Federal elections, which gives the ACT what it wanted.

The effect of the 2007 High Court decision was to revert to the previous Commonwealth law (including the removal of prisoners from the roll). This therefore disenfranchised all prisoners for ACT elections. In response, the Stanhope Labor Government introduced its own legislation to enfranchise all prisoners. This was the first occasion on which the ACT administered an ‘ACT only’ category of enrolment separate to the Commonwealth (Corbell, 2008: 3).75

Tasmania independently enfranchises prisoners serving sentences of less than three years (section 31(2), Electoral Act 2004 ). New South Wales (section 21, Parliamentary Electorates and Elections Act 1912 ) continues to enfranchise prisoners serving sentences of less than one year, as did Western Australia (section 18, Electoral Act 1907 ) until it amended its Act in 2007 to completely disenfranchise all prisoners, in line with the Commonwealth change. In May 2008 the Carpenter Labor Government in Western Australia introduced new legislation to extend the franchise to prisoners serving sentences of less than three years, but the Bill lapsed on the prorogation of Parliament ahead of the 2008 general election. South Australia has long-established laws enfranchising all prisoners, irrespective of the length of their sentences (Orr, 2004: 10). 76

74 Section 128(1). 75 The Electoral Legislation Amendment Act 2008 was passed in May 2008. 76 Sections 29(4)–(5) and 69 of South Australia’s Electoral Act 1985 .

189

In light of the High Court decision, and particularly Justices Gummow, Kirby and Crennan’s comments linking three-year sentences to the Australian Constitution’s requirement for three-year electoral cycles (HCA, 2007c: para. 102), there may be some scope to argue that New South Wales’ and Western Australia’s one-year sentence threshold, alongside four-year electoral cycles, could provide conditions for a Constitutional challenge. This said, there do not currently appear to be any moves in legal or human rights circles to mount such a challenge.

Partisan Impacts of Prisoner Voting

There is a belief, repeatedly expressed in the interviews, that allowing prisoners to vote provides a distinct advantage to the Labor party. As one parliamentarian noted: Obviously that does benefit Labor. I was a scrutineer at a Federal election at the remand centre and there were 14 people voting and I’m sure they all voted Labor or Green. One voted Green and the rest voted Labor from what I could see, but you’re not talking a lot of people.

While results from recent Federal elections do tend to suggest that this is the case, the advantage does not appear to be substantial, nor beyond what could be expected in the general population. As Table 8.3 illustrates, the number of prisoners voting is small, but their votes generally favour Labor.

Table 8.3 Prisoner Voting at Federal Elections, Two-Party Preferred Electorate and 2001 2004 2007 jurisdiction Labor Liberal Labor Liberal Labor Liberal Fraser, ACT 10 4 3 2 5 2 Bass, Tasmania –– –– –– –– 3 1 Franklin, Tasmania –– –– –– –– 3 3 Kalgoorlie, WA 90 16 144 46 20 13 O’Connor, WA 1 0 1 11 13 12 Canning, WA –– –– 7 2 6 4 Lingiari, NT 77 10 a 90 9 1 –– –– Note: These are the only electorates in which separate prisoner voting statistics are available. 1 Country Liberal Party candidate Source: AEC election results.

Two electorates in which Labor appears to have obtained a large advantage are in the seats of Kalgoorlie and Lingiari in the 2001 and 2004 elections. These two seats have the highest proportions of Indigenous people in the country and, combined with the high imprisonment rate of Indigenous Australians, these are the likely reasons for such high Labor support in these seats. The 91 per cent support for Labor from prisoners in

190

Lingiari in 2004 is consistent with the non-prisoner support for Labor in remote booths of Lingiari that have high Indigenous populations, such as Ti Tree Station (98.35 per cent Labor, two-party preferred), Titjikala (93.33 per cent), Kintore (86.99 per cent) and Willowra (84.19 per cent).

Analysis of interview data and parliamentary debates suggests that, although Labor may receive some electoral advantage from entitling prisoners to vote, their support for the prisoner franchise is based on genuine philosophical grounds rather than the prospect of partisan advantage. Likewise, the Liberal Party may see some small electoral advantage in denying prisoners the vote, but also tends to base its position on ideology. Comments by Liberal politicians such as Senator Minchin suggest, however, that populist positioning could also be a factor in the Liberal Party’s policy. The participation principle of fairness supports the broadest possible franchise. However, Australia’s mix of prisoner enfranchisement laws appears to be consistent with the diversity of laws applied internationally.

Party Registration

Political parties play a critical role in healthy democracies. One of their main functions is to provide an organised way of developing policy that represents societal cleavages (Reilly, 2002: 702). In addition, parties cultivate a democracy’s future leaders, they give a sense of order and stability to parliamentary organisation and debate, they bring together disparate groups and individuals into processes of democracy, they recruit political activists, and they provide defined choices in election campaigns (Reilly, Nordlund and Newman, 2008: 1). In modern democracies, it would be hard to imagine an organised and stable political environment without some form of political party structure.

While there may be a tendency for parties to naturally form to embrace new or existing social views, the institutional design of electoral systems allows for a form of ‘political engineering’ to occur –– either encouraging the formation of new parties (typically where a democracy’s party system is weak) or creating barriers to the continuation of existing parties or the formation of new parties. Barriers are typically erected in multi-

191 party democracies where an excess of parties has created confusion for voters at elections, or where there may be instability in government or parliament. Barriers can also be used as an anti-competitive measure in democracies where established parties seek to thwart potential new participants (or to deter non-genuine competition, as shown in Chapter 7 in relation to the 1999 New South Wales election). A healthy democracy would normally provide a balance between encouragement and restriction in its party regulation regime.

Organised political parties and other political bodies have existed in Australia for more than 100 years, but it is only in the past 25–30 years that significant reforms have occurred to recognise parties in a formalised sense for electoral purposes. These include ballot paper design, funding and disclosure requirements, and the distribution of voter preferences. In addition to influencing the capacity for citizens to coalesce into political groupings, such reforms impact on the ability of candidates to compete on a fair and equal basis. The favoured status given to registered political parties –– for example, in accessing public funding and identification on ballot papers –– also impacts on the ability of political organisations to deliver messages to voters during election campaigns. Such messages are essential to enable voters to make an informed choice.

Registration Criteria

The formal registration of political parties in Australia commenced in New South Wales in 1981, and has generally been viewed as a necessary corollary to other reforms such as public funding, party identification on ballot papers, and above-the-line ticket voting (JSCER, 1983: 182). All jurisdictions impose conditions concerning the name of a party. Consistent across all jurisdictions are requirements that: • Party names are to have a maximum of six words; and • Obscene names are prohibited. In addition, in every jurisdiction, party names are not to: • Resemble the name of another, unrelated party; • Be likely to cause confusion with another party; or • Contain the word ‘independent’ or ‘independent party’. 77

77 In the ACT, the use of the word ‘Independent’ or ‘Independent Party’ is only prohibited if it is used in conjunction with the name of another party, or is likely to be confused with another party – Section 93(2), Electoral Act 1992 .

192

Commonly, the following information is a condition of application for registration: • The name of a person to be the party’s registered officer (in Western Australia, the party secretary); • An abbreviated form of the party name, for ballot paper purposes; and • A copy of the party’s constitution (except Tasmania).

In Queensland, the constitutions of registered parties must include rules stipulating that preselection ballots are to be based on ‘principles of free and democratic elections’. 78 In addition, four jurisdictions require a fee for registration: $500 for the Commonwealth, Victoria, and the Northern Territory; and $2,000 for New South Wales.

All Australian jurisdictions also require a minimum membership size before a party can be registered. The minimum number of members differs widely, especially when compared to the total number of people enrolled in the jurisdiction. As Table 8.4 illustrates, the variation currently extends by a multiple of approximately 45, from the Commonwealth’s one member per 27,293 enrolled to the Northern Territory’s one member per 599 enrolled. Apart from these two outlier jurisdictions, however, the remaining seven jurisdictions are reasonably consistent, with a ratio of around 2,400 to 7,000 enrolled per party member. The capacity of citizens to form into political groupings is evident from Table 8.4. The Commonwealth, with a relatively low threshold, has the highest number of parties; conversely, the Northern Territory has the lowest number of parties contesting elections.

Table 8.4 Party Membership Requirements for Registration Jurisdiction Minimum Enrolled at most Enrolled / Parties required recent election required contesting most members (Year) members recent election 1 Commonwealth 500 13,646,539 (2007) 27,293 26 New South Wales 750 4,374,029 (2007) 5,832 16 Victoria 500 3,353,845 (2006) 6,708 12 Queensland 500 2,484,479 (2006) 4,969 6 Western Australia 500 1,330,399 (2008) 2,661 10 South Australia 150 1,055,347 (2006) 7,036 10 Tasmania 100 341,481 (2006) 3,415 5 ACT 100 243,471 (2008) 2,435 8 Northern Territory 200 119,814 (2008) 599 3 1 Separate divisions of a party (for example, Liberal NSW, Victoria; Labor, Country Labor, etc) or co- operative alliances (e.g. Australian Greens, Greens NSW) are counted as one party. Sources: Electoral commission election data.

78 Section 73A, Electoral Act 1992 (Queensland).

193

The Commonwealth has a relatively low party membership requirement, given the size of the Australian electorate. Understandably, there have been suggestions that the threshold should be increased, possibly to 1,000 members (JSCEM, 2005a: 91). Based on the ‘members to electors’ ratio, and given the number of parties contesting recent elections, this suggestion appears to be reasonable. However, it is not unusual for parties to have a distinct geographical base. Indeed, for Commonwealth registration, the major parties register separate parties based on State and Territory divisions.

A 500-member requirement assists the formation of parties based on local or regional issues, an example being the Save the ADI Site Party (SAS) in the 2001 and 2004 elections. SAS was registered in October 2001 to campaign for the retention of government-owned bushland in Sydney’s western suburbs. The land had previously been used by Australian Defence Industries (ADI) and the government was proposing to sell it for residential development. In 2001, SAS fielded candidates in four House of Representatives’ seats, achieving a substantial vote of 3.29 per cent (in Chifley). In the 2004 election, SAS contested three lower house seats (receiving a strong vote of 2.67 per cent in Lindsay) and stood two candidates for the Senate. Once the ADI bushland was sold to developers, SAS’s reason for existence was eliminated, and the party was voluntarily de-registered in August 2005. An increased threshold for registration would limit local issues groups having a democratic voice via mobilisation as a political party in this way.

Parliamentary Representative Alternative for Party Registration

As an alternative to meeting the membership requirement to register a party, four jurisdictions allow that a party may be registered if it has a parliamentary representative. South Australia adopts a relaxed approach with its parliamentary representative rule, requiring only that the representative be a member of a parliament or assembly of any of Australia’s nine jurisdictions ( Electoral Act 1985 , s.36). The other three jurisdictions – – the Commonwealth, Queensland, and Western Australia –– specify the parliamentary representative must be a member of that jurisdiction’s parliament.

In the case of Western Australia’s amending legislation in 2000, the Electoral Commissioner, Ken Evans, recommended that Queensland’s registration procedure be

194 followed (Moore, 2000: 8530). This procedure included a provision that parties with current parliamentary registration did not have to satisfy the 500-member rule. As a preventative measure to preclude independent members of parliament from subsequently setting up their own separate party, the Western Australian legislation (Electoral Act 1907 , s.62I) specifies that the ‘parliamentary representative’ rule only applies to pre-existing parties that had a parliamentary representative on 14 June 2000, when the legislation was introduced to parliament. At the time, there were five independent parliamentarians, of whom four were originally elected as either Labor or Liberal party representatives, and this provision was considered to be a measure with those members specifically in mind. The provision also allowed at least one party, the Australian Democrats, to contest the general election six months later as a registered party at a time when it would not have satisfied the 500-member rule.

The Commonwealth, Queensland and South Australian Acts do not contain any provisions to prevent a member of parliament from establishing his or her own party after being elected either as an independent or as a representative of another party. At the Commonwealth level, this loophole was used by Senator Meg Lees to establish the Australian Progressive Alliance party after she resigned from the Australian Democrats in 2002.

Where the legislation provides two separate ways to qualify for party registration –– either by membership level or parliamentary representation –– there is a question of fairness based on equal opportunity. The membership level criterion requires identified membership support within the community, while the parliamentary representation criterion is based on having sufficient voter support for an individual to be elected. When a party is registered because it has an existing parliamentary representative, there is no requirement to demonstrate any minimum level of party membership support. It could be argued that by having a representative elected demonstrates that such parties have a significant level of electoral or community support. However, this argument does not hold where the representative had been elected as a member of another party, and later resigned (as in Meg Lees’ case).

The lack of equity in the registration of parties is compounded in the three jurisdictions that allow parliamentarians to form new parties after they have been elected. In those cases, the representative does not have to show that there is any level of support for the

195 party, either in terms of membership or public support. The Northern Territory’s Minter Ellison report did not recommend a parliamentary representative criterion due to the inequitable effects of such a provision: It is arguable that a political party that only has one member does not fall within the meaning of “political party”, which is usually taken to denote a collection of people sharing some common political principles or goals. (MEC, 2003: 85)

South Australian Registration Regime

As mentioned in Chapter 7, South Australia has difficulties with its party registration regime, which was introduced as part of the State’s 1985 electoral reforms. Parties are able to lodge a voting ticket for the direction of preferences for both Legislative Council and House of Assembly elections. This provides an incentive, as has occurred in New South Wales, for multiple parties to register and to then trade preferences for the ‘harvesting’ of votes. This incentive also appears to have driven the growth in registered parties in South Australia since 1985.

Table 8.5 South Australia –– Registered Parties 79 Election Registered Parties standing candidates parties Council Assembly 1985 5 5 4 1989 10 7 7 1993 14 10 9 1997 24 12 8 2002 30 19 10 2006 28 10 8 Sources: SEO, 2007b: 31; House of Assembly, 1986.

As Table 8.5 illustrates, a larger number of parties competes in Council elections, where preference flows are especially important due to the proportional representation voting system. The incentive of directing preferences was described well by one legislator: Once again, the rules are written for one regime, but someone gets smart about it, and I think what happened, a large number of parties were registered when someone was going for an upper house seat.

79 SEO (2007b: 31) states that four parties contested the 1985 Council election but the House of Assembly (1986: 5) lists five parties as having contested the election (Australian Democrats; Call to Australia, Christian Alternative; National Party; Australian Labor Party; and Liberal Party) in addition to three named independents (Independent Nuclear Disarmament Platform; Independent for Peace and Survival; and Independent Marijuana Legalization Group).

196

Someone twigged that if they could set up enough parties, there’d be enough people who would vote for each named party, that by the time those preferences flowed on –– they came about because someone with a bit of political nous figured out that I might pick up about one per cent if I called a party this, and half a per cent for this, and I might actually get over the line simply because of the preferences of those particular groups.

In South Australia, the legislation also enables a person to be counted as a member of more than one party, for the purpose of meeting the registration requirement. This provision underpins the groupings of parties that occur. The Australian Labor Party (which has also registered the New Labor Party and Country Labor Party) and the National Party (and its Young National Party) probably have a sufficiently high membership to have distinct members for each party. However, the most extreme over- use of this provision is the Over-Taxed Motorists, Drinkers, Smokers Association, which registered five other parties (four of which were registered on the same day): the Smokers Rights Association; the Over-Taxed Smokers Association; the Over-Taxed Drinkers Association; the Over-Taxed Motorists Association; and the Over-Taxed Pokies Party. Despite repeated calls by the SEO for reform, the ability to use the same individuals to register multiple parties remains unchanged. As one administrator explained: From time to time electoral commissioners have made recommendations –– 1997 election, 2002 election, and now the 2006 election. There was a recommendation made that the same voters should not be used in reaching the membership thresholds of more than one party, and that’s perfectly sensible.

Similarly, the SEO has called for a restriction on the use of frivolous party names (SEO, 2007b: 61), again without success. This recommendation appears to be directed at parties such as the Stormy Summers Reform Party and Albert Bensimon’s No Hoo Haa Party.

Three other jurisdictions –– Queensland, Tasmania, and the Australian Capital Territory –– also allow the same members to be used to register more than one party. However, in these jurisdictions the incentive to use multiple parties to direct preference flows is absent. In Queensland (six parties at the last election), the system of single-member electorates and optional preferential voting removes this incentive, while in Tasmania (five parties) and the Australian Capital Territory (10 parties, including two ‘Independent’ groupings), the Hare-Clark voting system precludes ticket voting. It is

197 therefore more likely in these jurisdictions, compared to South Australia, that electoral outcomes will more accurately reflect voters’ wishes.

Party Registration –– Achieving a Balance

A well-constructed party registration regime may assist the conduct of fair elections by organising election candidates into clearly identifiable groupings, providing genuine competition and allowing voters to make informed choices. If the regime does not strike the right balance, it may unfairly restrict competition or allow excessive competition, as seen in the 1999 New South Wales election. Generally, however, when asked about their party registration procedures, electoral administrators were satisfied with existing provisions, particularly in regard to membership requirements. Responses included the following, from three commissioners: If they’ve got any worthwhile support, they can register quite easily. The reality is that after an election or two, if they don’t get many votes, they fade away. So we don’t get many, but it’s not that there’s a problem –– it’s difficult enough so that you’re not going to get nutty parties too often.

If you’re serious about contesting, then you should have…people who are prepared to say I support you, publicly. I think if you can’t find…members, how are you going to win seats?

I don’t think it’s a limiting factor. I think any party would probably need to have that many members anyway to have any great purpose, I would have thought, but it hasn’t been an issue.

An argument supported by several administrators is that systems with single-member electorates create a disincentive for registration, as small parties do not see a great opportunity to win seats. This was particularly the case in Queensland and the Northern Territory, the two jurisdictions without any form of proportional representation. These jurisdictions had only seven (Queensland) and three (Northern Territory) parties contesting their most recent elections. This is consistent with research conducted by Duverger (1951), Rae (1967) and others, who have argued that majoritarian systems support two-party systems.

The 2004 Northern Territory parliamentary debate on party registration raised a further issue in relation to the benefits of registration ––access to the electoral roll. It is common for members of parliament and registered parties to receive details of the electoral roll in electronic form. The main reasons cited for this are the need to promote

198 their policies and to converse with the public. However, access to the roll in this form is denied to potential Independent candidates prior to an election being called, and this puts those candidates at a disadvantage, as they are not able to compete on an equal basis with party-nominated candidates.

Deregistration of the Liberals for Forests

The ability of a government to use legislation to structure electoral competition is evident in the case of the Liberals for Forests party. One component of the 2006 Howard Government’s reform package was deregistration of all parties that did not have existing or past parliamentary representation. This resulted in the deregistration of 19 parties in December 2006. 80 It is believed that the key motivation for this aspect of legislation was to prevent the Liberals for Forests party from continuing to use ‘liberal’ in its name. The legislation was based on a recommendation by the JSCEM in its 2004 election inquiry report (JSCEM, 2005a: 100–1). The JSCEM recommendation reflected previous Liberal Party concerns about the use of ‘liberal’ and the Richmond electorate result in 2004 (see Chapter 4).

The Liberals for Forests had originally been registered in 2001, following an Administrative Appeals Tribunal of Australia (AATA) ruling against the AEC’s decision not to register the party (which was in part based on objections from the Liberal Party). In support of its decision, the AATA referred to words such as ‘liberal’ labour’, ‘progressive’, ‘national’, ‘socialist’ and ‘democrat’ as being generic and therefore not owned by one particular entity. 81 The Liberals for Forests had registered its name in the lower case, to identify its ideological position as ‘small ‘l’ liberals’, and to differentiate itself from the Liberal Party. The Howard Government legislated 82 to prevent the use of names that could be confused with existing parties, but could not apply this condition to parties already registered. A solution was to deregister all parties without parliamentary representation, requiring a re-application under the new law. Of

80 For a list of deregistered parties, see the AEC media release at 81 See par 40, Woollard and Australian Electoral Commission and Liberal Party of Australia (WA Division) Inc [2001] AATA 166. 82 Electoral and Referendum Amendment (Enrolment Integrity and Other Measures) Act 2004 , amending section 129 of the Commonwealth Electoral Act 1918 .

199 the 19 parties deregistered, eight have re-registered.83 Possible reasons why other parties have not re-registered include an inability to meet the 500-member test, or a lack of ongoing activity.

While the Liberal Party has been successful in removing registration of the Liberals for Forests at the Federal level, it has not had the same success at the State level. In Western Australia, Liberals for Forests has been a registered party since July 2001, and has one member of parliament in the Legislative Assembly (since 2001). However, the process of deregistering the party at the Commonwealth level raises two main concerns. First, it is through legislation that governing parties have an enormous power to limit electoral competition. This reflects the importance of the institutional structure of Australian electoral administration, a key theme in this thesis.

Second, the AATA has raised the important question of whether existing parties should be able to control the use of names that are based on ideology or history, such as those mentioned above. The possibility of voter confusion is, however, a real concern. The AATA made its position clear: ‘It is unlikely that any elector, seeing the two names on a ballot paper, will draw the conclusion that "liberals for forests" is a political party related to the Liberal Party of Australia’ (AATA, 2001: par 40). The only other debate on the issue has occurred in the highly-partisan JSCEM inquiry. There does not appear to be any systematic assessment of whether voters are confused by subtle variations in names.

This analysis of party registration regimes has shown that the development of party law has been heavily influenced by partisan interests. Some reforms have been initiated by governing parties for the purpose of eliminating opponents, such as the Coalition’s actions in deregistering the Liberals for Forests, and Labor’s high threshold for registration in the Northern Territory. Other developments in this area have resulted from party cartelisation, as in Western Australia’s 2000 reform. It can be seen from the examples discussed that the participation principle of fairness, under which citizens should have equal opportunities to form political parties, is diminished by the partisan behaviour of governing parties.

83 As of December 2008, the parties that have re-registered are the Christian Democratic Party (Fred Nile Group), Citizens Electoral Council of Australia, Non-Custodial Parents Party, One Nation Western Australia, Queensland Greens, Socialist Alliance, The Australian Shooters Party, and The Fishing Party.

200

Informal Voting

Any level of informal voting diminishes the legitimacy of election results and alienates voters through disaffection. The pattern and extent of informal voting provides insight into the health of an electoral system on a number of levels. Informal voting can either be deliberate (when voters choose not to make their vote count, for example to make a protest), or accidental (as when voters inadvertently make a numbering mistake). Both deliberate and accidental informal voting is a concern for electoral administrators, the former because it signifies citizens’ disconnection and disillusionment with the democratic system, and the latter because it indicates that voters do not understand the voting method, or that the process is too confusing or burdensome. Accidental informal voters are disenfranchised because the system has failed to provide an adequately simple method of voting.

In their study of informal voting in the 1987 and 1990 elections, McAllister, Makkai and Patterson (1992) note that among established liberal democracies, Australia has one of the highest rates of informal voting. Their study identified Australia’s social structure as being a key determinant of informal voting. In particular, ethnicity is a significant indicator of informal voting levels, with a lack of English language proficiency increasing the likelihood of informal voting (1992: 29–30). In recent years, the AEC has published a number of reports assessing informal voting rates. The reports on the 2001 (Medew, 2003) and 2004 elections (Dario, 2005) support McAllister, Makkai and Patterson’s argument that proficiency of the English language is a significant determinant of informal voting. In addition, Dario identifies two other important factors that influence levels of informal voting: the use of optional preferential voting at the State level (for NSW and Queensland) (2005: 15), and the number of candidates contesting an election (2005: 19). These factors are discussed in relation to Federal, State and Territory elections.

Informal Voting at Federal Elections

The high levels of informal voting in Senate elections were identified as a concern by the JSCER, which published the results of two informal vote surveys in its first report in

201

1983 (JSCER, 1983: 284–98). These surveys, of the Senate elections in 1977 (Australia-wide) and 1983 (14 selected divisions), showed that more than 50 per cent of all informal Senate votes were a result of incorrect numbering caused by either a break in the sequence, or duplicating numbers. A further 25 per cent of informal votes resulted from voters not numbering all the squares. The Hawke Government’s introduction of above-the-line Senate voting, which was used by more than 85 per cent of voters at the 1984 election, was instrumental in lowering the rate of informal voting (see Table 8.6).

Table 8.6 Senate and House of Representatives Informal Voting Rates (%) 1980 1983 1984 1987 1990 1993 1996 1998 2001 2004 2007 Senate 9.7 9.9 4.7 4.0 3.4 2.6 3.5 3.2 3.9 3.8 2.6 HoR 2.4 2.1 6.3 4.9 3.2 3.0 3.2 3.8 4.8 5.2 4.0 Sources: Reilly and Maley, 2000; AEC; AGPD.

The relaxation in the requirement to number all boxes for those voting below-the-line was another intervention that further lowered the informal rate. From 1984, Senate ballot papers were formal if they clearly identified a sequence of preferences against at least 90 per cent of the candidates listed (Farrell and McAllister, 2006: 44; section 103 of the Commonwealth Electoral Legislation Amendment Act 1983 ). For example, where there are 50 candidates on the ballot paper, numbering the ballot paper from 1 to at least 45 constitutes a formal vote.

While the 1983 reforms were successful in lowering the Senate informal rate, informal rates for the House of Representatives increased markedly following the same reforms. The AEC identified in its 1988 report that first preferences were clearly indicated in more than 62 per cent of all informal House of Representatives votes cast in the 1987 election. A relaxation of the rules to allow these votes to be added to the count would have reduced the informal rate from 4.9 per cent to 1.9 per cent (AEC, 1988).

As previously mentioned, the number of candidates contesting an election is a determinant of informal voting. The 2004 election figures show a positive correlation between candidate numbers and informal votes, with a steady increase in the informal rate as the number of candidates increases (Table 8.7).84 An examination of the highest and lowest informal rates for each category (of number of candidates) highlights the

84 This correlation between the number of candidates and informal voting is also discussed in this chapter in reference to Northern Territory elections.

202 influence of other factors, such as ethnicity. For example, the high rate of 11.71 per cent of informal votes with eight candidates contesting occurred in the seat of Reid, an inner metropolitan electorate in Sydney. Reid has the highest proportion of people born overseas of all Australian electorates (49.5 per cent), and the second-highest proportion of people from non-English speaking countries (35.5 per cent) (Nelson, 2007: 35, 39). This relationship is further illustrated in Table 8.8 which presents electorates with the highest informal rates per candidate by location and ethnicity.

Table 8.7 Informal Voting Rates –– 2004 House of Representatives Election No. of No. of Highest Lowest Average candidates seats rate (%) rate (%) rate (%) 4 3 5.53 3.61 4.31 5 18 9.11 2.87 4.44 6 29 9.24 2.76 4.68 7 39 9.10 2.77 4.89 8 30 11.71 3.40 5.64 9 18 8.43 4.22 5.70 10 6 7.45 4.49 5.81 11 5 8.53 5.89 6.64 12 1 7.41 7.41 7.41 13 0 –– –– –– 14 1 11.83 11.83 11.83 Total 150 11.83 2.76 5.19

Table 8.8 2004 House of Representatives High Informal Rates –– Ethnicity No. of Highest Electorate NESC candidates informal ranking rate (%) 4 5.53 Throsby 62 5 9.11 Fowler 8 6 9.24 Prospect 12 7 9.10 Watson 1 8 11.71 Reid 2 9 8.43 Kingsford Smith 25 10 7.45 Lindsay 65 11 8.53 Parramatta 16 12 7.41 Dobell 124 13 –– –– –– 14 11.83 Greenway 48 Note: NESC means ‘Born in a Non-English Speaking Country’. Ranking is out of 150 electorates, with 1 indicating the highest proportion of NESC.

Of the 10 seats with the highest rates of informal voting per number of candidates, more than two-thirds (seven) are ranked in the top third of electorates for numbers of people born in non-English speaking countries. It is also significant that all of these

203

10 electorates are in New South Wales, where an optional preferential voting system is used at State level elections. In an assessment of informal votes in New South Wales, Dario found that 35.65 per cent were made informal by the voter numbering ‘1’ only, with a further 10.71 per cent of voters having incorrectly used a tick or cross (2005: 15– 16). In all, 46.36 per cent, or more than 116,000 voters in the State attempted to vote correctly but were accidentally disenfranchised. All of these votes would be formal under State election voting rules. It is important to note that the informal rate at the 2003 New South Wales Legislative Assembly election was 2.6 per cent (Hann, 2006: 3), less than half of the 6.1 per cent the State recorded at the 2004 Federal House of Representatives election in New South Wales. This disparity is further evidence that confusion between different voting rules, and making voting unduly burdensome, contributes to informal voting.

Proposals to introduce optional preferential voting have lacked support at the Federal level. Parliamentarians have raised concerns about the potential impacts of relaxing the rules to allow ‘1 only’ votes, or votes using ticks and crosses. Concerns are largely based on the difficulty in discouraging voters from intentionally voting this way. The JSCEM report on the 1990 election highlighted this difficulty, with the committee stating that a change to the rules would encourage optional preferential voting, as well as ‘Langer’ style voting (JSCEM, 1991: 41–2). 85 As discussed in Chapter 7 in relation to Queensland’s voting system, such changes could lead to a form of de facto first-past- the-post outcomes.

Informal Voting at State and Territory Levels

Analysis of informal voting at State and Territory elections provides an insight into the impacts of the voting systems in use. Using data for the lower house (or single house in unicameral systems) from the past three elections (Table 8.9), it is evident that the optional preferential voting system in place in New South Wales and Queensland consistently produces lower rates of voting informality.

85 Langer style voting was promoted by Albert Langer so that voters could avoid preferencing one major party over the other by duplicating a number in their preferences, for example by numbering the ballot paper 1, 2, 3, 4, 5, 5.

204

Table 8.9 State and Territory Informal Rates –– Lower/Single Houses Voting Informal voting rate (%) Average System Jurisdiction system 1996 2007 informal average rate (%) (%) NSW OPV 2.51 2.62 2.77 2.63 2.37 Queensland OPV 2.27 1.99 2.08 2.11 Victoria FPV 3.02 3.42 4.56 3.67 W. Australia FPV 4.39 4.54 5.24 4.72 4.26 N. Territory FPV 5.17 4.27 3.75 4.40 S. Australia FPV-Ticket 4.04 3.12 3.60 3.59 3.59 Tasmania Hare-Clark 3.91 4.87 4.49 4.42 4.04 ACT Hare-Cark 4.32 3.97 2.68 3.66 OPV= Optional preferential voting; FPV = Full Preferential Voting; FPV-Ticket = A ‘1’ only vote is formal, with full preferences distributed according to a ticket lodged by the candidate.

South Australia is the only Australian jurisdiction that uses ticket voting as an option for single-member electorates. Where voters make a ‘1 only’ vote, preferences flow according to a ticket lodged by the candidate of their first (and only) choice. This may partially explain why South Australia’s informal rate is lower than the other jurisdictions with full preferential voting. However, ticket voting makes up less than five per cent of all votes in South Australian Assembly elections, so its effect on informal voting cannot be substantial. The two jurisdictions operating under the Hare- Clark system of proportional representation, with no ticket voting, also record relatively high levels of informality. An interesting aspect here is the steadily decreasing rate of informal voting in the Australian Capital Territory, partly a result of the introduction and increasing use of electronic voting (as discussed in Chapter 7).

Analysis of Cross-Jurisdictional Issues

This chapter has assessed four contemporary electoral issues –– voter turnout, prisoners’ voting rights, party registration and informal voting –– that have relevance across all Australian jurisdictions. These issues were selected for analysis based on interview data and the indicators for comparative analysis (Table 3.6). The applicable indicators are: • Voter turnout –– levels of participation, ease of voting, and accessibility; • Prisoners’ voting rights –– inclusiveness; • Party registration –– barriers to entry, and membership threshold; and • Informal voting –– inclusiveness, ease of voting, and levels of informality.

205

Voter turnout is not usually considered to be a significant issue in democracies with compulsory voting, and the analysis shows that there is generally strong turnout around Australia. The Northern Territory is, however, an exception, and analysis of the data indicates that more can be done to encourage and assist Indigenous people in non-urban and remote Australia to participate in the voting process. Indigenous engagement also appears to be a matter for concern in areas outside the Northern Territory, such as parts of Western Australia, Queensland and northern South Australia. Indigenous turnout has not improved to any great degree during the 25-year study period, indicating that Indigenous citizens are the ‘forgotten people’ in Australian elections.

Professional electoral administrators expressed concerns about low Indigenous turnout, and have suggested ways of improving participation, such as targeted education programs and longer election periods. However, their desire to improve participation has not been supported by either Coalition or Labor governments. This is due to the partisan effects increased participation is likely to have: a negative impact for the Coalition; and, for Labor, the lack of a positive impact.

The participation principle for fairness is that all citizens should have an equal opportunity to participate in the voting process (Mercurio and Williams, 2004: 367–8), and this is clearly relevant to voter turnout. Indigenous people living in remote areas are disadvantaged by the current system, which, if left unaddressed, will compound Indigenous people’s lack of connection with their democratic role as citizens. The failure to improve participation levels through the study period, combined with the paucity of government action, ensures that Indigenous participation in elections will remain an important issue into the future.

There is no generally-accepted international practice for prisoner voting. However, Article 25 of the ICCPR suggests that a full prisoner franchise is consistent with international human rights principles. Currently, the variation in laws across Australia mirrors the diversity in approaches to the prisoner franchise internationally. The 2007 High Court decision to uphold a limited prisoner franchise provides a minimum standard, and it is likely that future reforms will move towards extensions rather than restrictions of the franchise in this area. While there is an observable advantage to Labor in giving prisoners the vote, the partisan interests on the issue appear to be

206 secondary to genuine ideological positions. There is, however, some incongruity in the Liberal Party’s position, given that it champions ‘the inalienable rights and freedoms of all peoples and in those most basic freedoms of parliamentary democracy’.86

Comparatively, South Australia and the Australian Capital Territory exemplify best practice for this aspect of electoral law. Their full prisoner franchise is consistent with the participation principle requiring all citizens to have an equal opportunity to vote. Worst practice currently applies in Western Australia, with its total disenfranchisement of prisoners. In response to the High Court decision, amending legislation is currently before the Western Australian Parliament, which would enfranchise prisoners serving less than a one year sentence. 87 If passed, this would still be Australia’s worst practice, along with New South Wales, which currently has the one-year provision. Apart from the South Australian case, there is reluctance among governments of all persuasions to move toward giving the franchise to all prisoners. This is generally due to concerns that such a move would be portrayed in the media as giving ‘murderers and rapists’ the vote. Consequently, significant further changes to the prisoner franchise are unlikely.

Party registration regimes are a mechanical functioning of electoral law that can encourage or impede competition. Barriers to entry and membership thresholds are the comparative indicators used in this study. Both indicators are directly relevant to the participation principle that citizens have an equal opportunity to form parties and compete in elections. Party registration has provided a structure necessary for modern Australian elections, complementing public funding, financial disclosure and ballot paper design laws. However, it remains an evolving field of electoral law, and there are difficulties in providing the right balance of regulation that encourages genuine and ample competition without descending into frivolity or the establishment of parties as ‘fronts’ for the harvesting of votes. The wide variance in member thresholds for party registration (particularly the Northern Territory system), and the ability for individual parliamentarians to establish parties in some jurisdictions, is indicative of the partisan motives at play in the formation of these laws.

86 From the Liberal Party website, What Does The Liberal Party Stand For . 87 The Electoral Amendment (Miscellaneous) Bill 2008 was introduced on 11 November 2008.

207

The ability for governing parties to act alone, or within a party cartel, to control party registration law and therefore limit its electoral competition, is a serious threat to the conduct of free and fair elections. The requirement to provide evidence of a reasonable membership level is internationally-accepted practice. However, it is also an internationally-accepted norm that such standards be set by independent electoral authorities, rather than by the parties that stand to benefit from limiting competition. Tasmanian laws are best practice in this area of electoral legislation. The State has a reasonable threshold of 100 members for its more than 340,000 electors, and parties are unable to circumvent this requirement by having a parliamentary representative. Furthermore, Tasmania does not have a party registration fee. The Commonwealth represents worst practice, enabling any parliamentarian to separately register a new party, and imposing a registration fee (although at $500 this could be considered reasonable). Although the Commonwealth has an encouraging threshold of 500 members for more than 15.5 million electors, its willingness to legislate to override an independent tribunal’s decision is a serious affront to fairness.

The final issue analysed in this chapter, informal voting, provides a good indication of voters’ feelings of connection with elections, and of the efficacy of the voting system. Informal voting rates are of concern in most jurisdictions, with only Queensland and New South Wales having ‘acceptable’ levels. The improvement in the ACT’s informal rate is encouraging and suggests one benefit of extending electronic voting. At the ACT’s most recent election in October 2008, however, the informal rate increased to 3.8 per cent. Further analysis is necessary to understand the reasons for this change. It appears that a move to optional preferential voting in Victoria, South Australia, Western Australia and the Northern Territory would lower the informality rate in those jurisdictions. Optional preferential voting would enfranchise voters who have been accidentally disenfranchised in the past. Such systems do, however, also raise questions about the legitimacy of results if parties encourage voters to ‘just vote ‘1’’, as has occurred in Queensland.

The diversity of voting rules across Australia creates voter confusion, with the result that many voters are accidentally disenfranchised. A solution would be for parliaments to cede their rule-making powers to independent electoral management bodies, thereby removing partisan or self-interest motivation. These independent bodies could then agree on uniform rules for formality (taking into account the different voting systems)

208 targeted at making voting easier and lowering the rate of informality. Such action would not only reduce accidental informal voting, but may also decrease intentional informal voting because citizens feel more empowered.

Summary

Analysis of the four issues discussed in this chapter suggests governing parties act in a partisan manner to override principles of fairness. Australian electoral management bodies lack the power to independently develop a regulatory framework based on fair and non-partisan practices. Applying IDEA’s typology of electoral management powers, Australian commissions fit into the governmental model, defined as: ‘powers are limited to implementation…Does NOT have independent regulatory powers’ (Wall et al., 2006: 9, 13). The management of the four issues discussed here is therefore dependent on government policy and legislation, and falls short of fulfilling internationally accepted norms for fair elections.

209

Chapter 9

Concluding Discussion

At a time when electoral reform is a high priority on the policy agenda of the Rudd Labor Government at the Federal level, and of several State governments across Australia, this research makes a significant contribution to understanding the reform process and the impacts that reforms have on principles of fairness. From this thesis, academics, professional electoral administrators and policy makers will be better equipped to understand the policy forces at play, such as the partisan interests of governing parties and others within the hierarchy of incumbency. Importantly, this study also provides an assessment of the impacts that partisan motives have on international norms of electoral fairness.

The main aim of this research was to ascertain whether electoral reforms in the past 25 years have enhanced or diminished fairness in the conduct of Australian elections. In answering this question, the research considered whether reforms were consistent with international norms for electoral management and fair elections; and whether the institutional structure of Australian electoral management protected the electoral system against governing party control. The forces driving electoral reform were identified, providing insight into the reform process. Prominent among these forces is the role of governing parties within the institutional structure of electoral management.

It was hypothesised that reforms have had a mixed result in terms of fairness, with improved equity in electoral outcomes but diminished access to political participation for candidates and parties. In addition, it was considered that party cartelisation had intensified during the 25-year study period, benefiting the major governing parties to the

210 detriment of minor parties. It was hypothesised that this cartelisation was at the expense of opportunities to adopt ‘best practice’ for electoral management and fair elections.

Summary of the Research Process

The methods used in the research drew on three approaches to democracy and fairness assessment: discussion-based models; comparative models; and ratings-based models. These models provided the themes and indicators that formed the basis of the research. Data was drawn from four sources: existing literature; electoral legislation; parliamentary inquiries and debates; and interviews. The collection of interview data was a critical phase of the research process. The data were used to identify reform issues that warranted case study attention in the subsequent stages of the research, using Creswell (1998) and Stake’s (1995) feedback model of data collection and analysis. Importantly, the interviews generated a better understanding of the conditions under which electoral reform occurs, including the importance of personal relationships in progressing reforms at the administrative level.

Dependent on the theme being analysed, the study drew on a mix of the three models of assessment. Generally, a combination of the discussion-based and comparative assessment models was used in relation to the electoral reforms presented in Chapters 4 to 8. The Elklit and Reynolds ratings-based model was also a valuable tool in guiding the assessment of the independence of electoral management bodies, as discussed in Chapter 5. Elklit and Reynolds’ original model was designed for assessment of both developing and developed democracies. For the purpose of this thesis, I modified the Elklit and Reynolds model in order to distinguish between similar legislative and administrative arrangements and to identify best practice in developed democracies. In addition, Gallagher’s Least Squares Index quantitative measure, designed to gauge the impact that voting systems have on the fairness of representation, was used as an assessment tool in analysing Commonwealth, State and Territory reforms in Chapters 6 and 7.

In answering the question of whether reforms have enhanced or diminished fairness, various definitions of ‘fairness’ were considered. The definition considered most

211 pertinent to this study was that put forward by Mercurio and Williams (2004). They define freedom and fairness in terms of four underlying principles: participation; free- conscience; election outcome; and knowledge. These principles, especially participation (having an equal opportunity to participate) and knowledge (understanding the electoral process and being able to make an informed choice), provided consistent reference points throughout the study.

The thesis concentrated on two primary aspects of electoral systems. The first assessed was the institutional design of Australian electoral administration, with discussion of the institutional structure of the system (Chapter 4), followed by an assessment of the independence of electoral management bodies (Chapter 5). The second aspect was analysis of reform processes and the impacts of reforms at the Commonwealth (Chapter 6), and State and Territory (Chapter 7) levels. Selected reform issues that are of current relevance across multiple jurisdictions were considered in Chapter 8. At the Commonwealth level, the study focused on the two major waves of reform –– in 1983 and 2006 –– that occurred in the 25-year period. At State and Territory levels, reforms were selected for their uniqueness to the jurisdiction. Taken together, these assessments of reforms and analyses of reform processes add to the knowledge of the Australian institutional structure and political environment in which reforms occur.

Summary of Findings

This research adds to knowledge about Australia’s electoral administration structure and about how reform processes can increase fairness even when prompted by partisan motives. The findings also identify areas where partisan interests override principles of fair elections, such as the maintenance of an independent electoral administration.

First, it was found that the institutional structure for electoral administration in Australia –– primarily the relationships between parliaments, political parties, and commissions – allows governing parties to control the electoral system to their own benefit. This power to control arises from the ability to legislate, and from having majorities on committees inquiring into electoral matters. Legislation is typically heavily

212 prescriptive, providing commissions with little scope to independently adopt reforms based on non-partisan ideals of fairness and equity.

In analysing selected Australian electoral reforms during the study period, the thesis tests two political science theories: voting and electoral system design theory; and party cartelisation theory. The case-study findings make a contribution to the academic debate in these theoretical areas. The findings appear to discount Duverger’s (1951) theory that it is the electoral system that shapes the party system. Instead, the results support Rokkan’s (1970) hypothesis that it is the governing party system that shapes the voting and electoral systems. While Duverger’s theory helps to explain the formation of the party system that occurred in Australia more than 100 years ago, Rokkan’s theory is more appropriate for the 25-year study period of the present research. The findings also provide evidence for the existence of party cartelisation in Australia. However, results from this research highlight the fluidity of cartelisation, with cartel membership shifting in accordance with the power balance in place at the time of reform.

Using IDEA’s criteria for electoral management, it is evident that Australian commissions have little independence from government, and instead conform to a ‘governmental’ model in relation to powers to regulate. On broader aspects of authority, Australian commissions meet IDEA’s ‘mixed’ model, having a degree of institutional independence, but still within the direction and control of government (Wall et al., 2006: 9). The analysis of the independence of electoral management bodies in Chapter 5 demonstrates that Australian electoral commissions are non-partisan, or neutral, rather than being truly independent in the internationally-accepted meaning of the concept. While Australian electoral commissions have a good degree of autonomy in carrying out their functions of enrolling voters, registering parties, and conducting elections, they lack the true independence that would enable them to structure the regulatory framework in accordance with international standards for fair elections and independent electoral management.

A frequently occurring theme in interviews conducted with commissioners was that they lacked the power to publicly criticise weaknesses in their governing legislation. This was largely due to the ongoing need to carry out the functions of their commissions as dictated by parliament. Electoral commissions are, therefore, only able to make suggestions for reform, rather than being advocates for specific changes to occur.

213

Accordingly, the commissions are servants to their political masters –– administering the laws they are given –– rather than being truly independent and ensuring the electoral system is regulated on the basis of fairness principles.

The emerging body of inquiry by parliamentary committees is important in providing a forum for public debate of electoral matters issues. However, the value of the committees’ work is severely limited by the partisan approach that is often taken by committee members, particularly at the Commonwealth level. This is evident in the approach the JSCEM has consistently taken on postal voting processes, despite the protestations of the AEC. Although not as extreme as the Commonwealth example, evidence of partisan behaviour on committee inquiries was also found at the State and Territory levels.

The findings of this research provide evidence of party cartelisation. While the Labor/Coalition cartel is the dominant relationship, formation of other cartels occur when it is in the interests of the participants to collude. Evidence of the dominant cartel is apparent in JSCEM’s approach to postal voting, and the lack of action from both Labor and the Coalition to address the concerns of the AEC. These findings support the party cartelisation theory, as promoted by Marsh (2006), Ward (2006), Young (2003) and others, that Labor and the Coalition collude to prevent other parties and candidates from being able to compete on equal terms. At the State level, Labor/Coalition cartel behaviour was clearly apparent in reducing the size of the Tasmanian Parliament to limit growth of the Greens party.

Interestingly however, the Australian party cartel is not a strict relationship between only Labor and the Coalition. The Australian party cartel is more fluid, with members joining and leaving the cartel on an issue-by-issue basis. This thesis provides numerous examples of variations in cartel membership. For example, the National party joined with Labor to further its own interests by agreeing to the increase in seats for the Federal House of Representatives in 1983, and to the introduction of proportional representation in Western Australia in the 1980s. Another example of shifting cartel affiliations is seen in the behaviour of the Democrats at the Federal level in 1983. On the issue of optional preferential voting, the Democrats combined with the Liberal party to oppose its introduction by Labor because of the impact it would have on the Democrats’ ability to negotiate preferences for Senate ticket voting. However, the

214

Democrats combined with Labor to support the introduction of public funding, against opposition from both Coalition parties. Similarly, we saw Labor in Victoria in the 1980s move to introduce proportional representation to garner support from the Democrats (outside parliament), who were expected to benefit from its introduction, while the Liberal-National Coalition opposed the reform.

Is this cartelisation at work, or is it better described as shifting ‘marriages of convenience’, with partisan self-interest acting as the primary driver of temporary coalescence? The evidence over the past 25 years suggests that where possible, parties will act alone in instigating reform in their own interests. An example is Labor’s treatment of party registration in the Northern Territory, where it increased the membership threshold from the recommended 20 to its preferred 200, to limit electoral competition. The Coalition’s reform agenda at the Federal level, which was supported by JSCEM’s acquiescence and passed in 2006, provides another example of parties acting in their own self-interest.

It has been uncommon in recent years, however, for any party to have control of government without having to accommodate the interests of other parties. There are several examples of parties governing as a minority government, or where the governing party is in the minority in an upper house. Cartels of convenience in these instances have been established to implement reforms that further the interests of all cartel members. It has been shown that these cartels, or ‘marriages of convenience’, create inequitable electoral systems that reflect and reinforce the hierarchy of incumbency (described in Chapter 2). This hierarchy provides electoral advantages to governing parties and their cartel partners that are not available to non-cartel parties or Independents.

Although it can be argued that reforms such as public funding assists smaller parties to mount election campaigns, the reluctance of the Labor and Coalition parties to limit corporate donations –– which are frequently directed towards parties in power –– significantly disadvantages smaller parties. It has also been noted that some jurisdictions have reduced effective thresholds for entry to parliament, as with the introduction of proportional representation in Western Australia and Victoria. However, when a serious threat to shared Labor and Coalition interests appears, the

215 cartel can re-mobilise to act decisively in response to the threat. An example is the Labor/Liberal cartel’s response to the Greens gaining the balance of power in Tasmania.

A key finding in this study is the importance of personal relationships in determining the final outcomes of reform processes. As Phillip Green (1986) has noted, in 1983 the Hawke Government and the JSCER were committed to achieving reform based on consensus wherever possible. In contrast, the story of the Howard Government’s reform attempts was characterised by a fiercely adversarial approach, in which the Government co-opted the JSCEM to provide supporting arguments for the Coalition’s pre-determined policies. The less-public relationships, such as those between commissioners and senior officials from other Departments (particularly from Treasury and Premiers’ or Prime Minister’s departments), and between commissioners and their governing Ministers, also played a critical role in the reforms. The outcomes may not always be in the best interests of the governing party, but are often consistent with commissioners’ desires for fair regulation and smoothly-functioning systems.

Fierce partisan divisions exist on a number of electoral issues, such as the timing of roll closures, donation disclosure, and the prisoner franchise. The lack of an independent decision-making body with the power to implement reform means that such issues will continue to be fought on partisan lines, with fairness remaining a secondary consideration. In the context of democratic outcomes, partisan advantage has been shown to be a far stronger driver for reform than the desire to improve fairness of elections. Any advantages to voters or non-governing parties tend to be an incidental, or even accidental, outcome of reform rather than a motivating consideration.

While party self-interest is the force driving most electoral reform in Australia, there are examples of reform based on principles of fair elections, notably the shift to proportional representation for Victoria’s upper house. In that case, reform has given voters having a fairer system, which provides them with more equitable outcomes relative to their vote. As a result of the reform, Labor lost its majority in the Legislative Council; under the old majoritarian system, Labor probably would have maintained its majority, at least in the short term.

As outlined, the findings of this study generally support the dominant electoral system theory put forward by Rokkan (1970), Tsebelis (1990), Benoit (2004), Colomer (2005)

216 and others –– that it is the governing parties that shape voting and electoral systems. This position does not necessarily contradict Duverger’s (1951) theory that party systems are determined by the electoral system in which they act. Duverger’s theory remains relevant to the original formation of parties in Australia, which occurred in the late 19th and early 20 th centuries. This was a critical juncture in the development of Australian electoral systems in path dependency terms. Since that time, path dependency has ensured that there are important continuities between the modern Australian party system and its early formation. The result, a predominantly two-party system, supports Duverger’s theory. However, for the period studied, 1983 to 2007, the dominant electoral system theory outlined above has greater explanatory power, in that governing parties use the reform process to further their own self-interest.

While some reforms have benefited other (minor) parties, this has largely occurred in situations where governing parties have required minor party support to pass legislation. Such reforms provide further evidence for the argument, made by theorists including Benoit, that an ongoing tension exists between voting systems and parties, with each attempting to dominate the other. Lijphart argues that, in response to this tension, proportional voting systems produce stronger electoral institutions. However, this argument is confounded in the Australian setting, which mainly consists of bicameral parliaments using both majoritarian and proportional voting systems. This thesis has shown that majoritarian houses generally dominate proportional houses (with Tasmania being the exception) and, as a result, electoral institutions are left weaker.

Significance of the Research

This research contributes to the academic literature on electoral systems in two notable ways. First, as Massicotte, Blais and Yoshinaka (2004) have noted, academic studies of electoral system design overwhelmingly concentrate on national systems, and very little comparative work on electoral systems at the sub-national level has been carried out. This is certainly true of the literature on Australian systems. The comparative analysis of all Australian jurisdictions conducted in this study helps to address this gap. The case study analysis of the 25-year period captures the reform environment in a critical phase of Australia’s electoral history. An integral component of this research is the

217 interview data, which provides an unusually intimate view of how electoral reforms occur in the Australian setting. In particular, the interview data highlights the tensions at play during the reform process, between the conflicting objectives of fair elections and party self-interest.

Second, the normative assessment conducted in this thesis applied a modified version of the Elklit and Reynolds framework to understand the institutional structure of Australian electoral commissions. This framework provides a useful guide to assessing the independence of Australian commissions. This modified framework developed in the thesis is more relevant for developed democracies than the original Elklit and Reynolds model, and will be useful for future assessments of Australian electoral systems and of the systems in other developed democracies. Ideally, the modified Elklit and Reynolds framework could be used to assess all aspects of a selected theme of electoral systems (detailed in Chapter 3). In this study, the modified framework proved to be useful for assessing the independence of electoral management bodies, but other assessment models were also drawn on to guide assessment of selected electoral issues in Chapters 6 to 8. This approach was taken because case study methodology was applied to selected reforms and reform processes, analysing selected indicators in specific jurisdictions. Apart from the independence of electoral management bodies, a ratings-based comparative assessment of themes and indicators across all jurisdictions was not pursued, and the modified Elklit and Reynolds framework could be fruitfully extended to an assessment of other indicators across all jurisdictions in the future.

Future Research Directions

It is anticipated that the modified Elklit and Reynolds framework developed for this thesis will be a useful tool for further research on electoral systems and electoral management bodies. The five-point scoring scale from best to worst practice provides sufficient separation between similar systems. Furthermore, the indicators for assessment allow for flexibility depending on the democracies being studied. However, the application of the modified framework to other aspects of the electoral system may be problematic. In the present study, indicators are given scores without adding a

218 weighting to reflect the relative importance of the indicator. 88 As noted in Chapter 3, there is a subjective element to such assessments, and further research is needed to identify the relative importance of different aspects of an electoral system in providing free and fair elections. For example, how do you compare the importance of candidate nomination matters with voter enrolment issues –– is one more important than the other?

Political finance regulation continues to be a contentious issue in Australian politics, and this area of electoral law would benefit from analysis using the modified Elklit and Reynolds framework. Political finance regulation was not assessed in the present study due to the excellent work that has been conducted by Sally Young and Joo-Cheong Tham (2006). Given the dynamic nature of public debate on political donations and public funding, however, a comparative assessment using the modified framework would be opportune, especially with reforms in this area being proposed at the Federal and State levels. Such an analysis could also incorporate the hierarchy of incumbency, assessing the financial resources available to incumbent parliamentarians and parties. As has been demonstrated throughout this thesis, the hierarchy of incumbency tends to create unfair electoral competition.

The assessment of the independence of electoral commissions presented here provides a useful benchmark of current conditions. Research into the commissions’ administrative structures at lower management levels would provide greater insight into how independence is regarded at those levels. Such research could also assist in understanding the role that having a career electoral professional as commissioner plays in creating a culture of independence.

Conclusion

Voters are able to express their opinions about a government’s or a party’s policy direction at the ballot box. However, it is electoral law that determines how these citizens vote, and if they can vote at all. When governing parties have control over such laws, this raises concerns about the legitimacy of a democracy. In Australia, governing

88 This is the case for both the modified framework and Elklit and Reynolds’ original framework.

219 party control highlights the need for greater independence in the management and conduct of Australian elections. Despite the growth of issue-specific parliamentary committees, these do not provide a sufficiently powerful policy process to bring about electoral reforms based on international norms for electoral management and fair elections. An alternative approach would be to establish an agency independent of both government and parliament that has the power to regulate the manner and conduct of elections.

Such a truly independent agency could only be achieved, however, if the governing parties were willing to cede their law-making powers. The evidence provided in this study indicates that such voluntary relinquishment of control is unlikely to occur. Pressure external to the electoral institutional structure, as occurred with the Fitzgerald Inquiry in Queensland, may be required. Evidence of media interest in party donations, incumbency benefits and the political finance system has been particularly noticeable in recent times (for example, the Wollongong scandal in New South Wales), and this may drive change.

At the national level, the Rudd Labor Government has established a significant electoral reform agenda, including legislation already introduced, and a policy of releasing green (discussion) papers on matters such as political finance and enrolment. Unless changes are made to the policy development and implementation processes, however, it would appear that the Rudd Government is destined to follow the examples of Hawke and Howard –– electoral reform largely based on partisan self-interest, but perhaps with shifting alliances between parties in cartels of convenience.

220

221

References

AATA [Administrative Appeals Tribunal of Australia]. 2001. ‘Woollard and Australian Electoral Commission and Liberal Party of Australia (WA Division) Inc [2001] AATA 166 (6 March 2001)’. URL: < http://www.austlii.edu.au/au/cases/cth/AATA/2001/166.html>. Retrieved 24 March 2007.

ABS [Australian Bureau of Statistics]. 2002. ‘1384.6 –– Statistics –– Tasmania, 2002’. URL: . Retrieved 14 March 2007.

ABS. 2007. ‘4705.0 –– Population Distribution, Aboriginal and Torres Strait Islander Australians, 2006’. URL: . Retrieved 23 February 2008.

ACE [ACE Electoral Knowledge Network]. 2007. EMB independence and the origin of independent election administrations . URL: . Retrieved 3 July 2008.

ACTEC [ACT Electoral Commission]. 2002. The 2001 ACT Legislative Assembly Election: Electronic Voting and Counting System Review . Canberra: ACT Electoral Commission.

ACTEC. 2005. ACT Legislative Assembly Election 2004: Electronic Voting and Counting System –– Review . Canberra: ACT Electoral Commission.

ACTEC. 2006. Electoral Compendium 1989–2006 . Canberra: ACT Electoral Commission.

AEC [Australian Electoral Commission]. 1988. Informal Voting 1987, House of Representatives . Canberra: Australian Government Publishing Service.

AEC. 1996. Submission to the Inquiry by the Joint Standing Committee on Electoral Matters: The Conduct of the 1996 Federal Election . URL: . Retrieved 19 June 2008.

AEC. 1999. Submission to the Joint Standing Committee on Electoral Matters: The Conduct of the 1998 Federal Election . Canberra: Australian Electoral Commission. URL: . Retrieved 28 November 2007.

AEC. 2000. Submission to the Joint Standing Committee on Electoral Matters Inquiry into the Integrity of the Electoral Roll . Canberra: Australian Electoral Commission. URL: . Retrieved 24 November 2006.

AEC. 2001. Submission to the Senate Public Finance and Administration Inquiry into Bills Concerning Political Honesty and Accountability . Submission No. 14. Canberra: Australian Electoral Commission. URL: . Retrieved 14 July 2008.

AEC. 2004. Over 13 Million Australians Have the Right to Vote in the 2004 Federal Election . Media Release, 10 September. URL: . Retrieved 24 June 2007.

AEC. 2007. House of Representatives - Two Party Preferred Results 1949 –– Present . Canberra: Australian Electoral Commission. URL:

222

Australian_Electoral_History/House_of_Representative_1949_Present.htm>. Retrieved 5 October 2007.

AEC. 2008. Submission to the Inquiry by the Joint Standing Committee on Electoral Matters into the 2007 Federal Election . URL: . Retrieved 28 June 2008.

ALP [Australian Labor Party]. 2002. ‘Joint Standing Committee on Electoral Matters Inquiry into the 2001 Federal Election: Submission by the Australian Labor Party’. Submission No. 153. URL: . Retrieved 11 August 2007.

Alport, K. and L. Hill. 2008. ‘Voting Attitudes and Behaviour Among Aboriginal Peoples: Reports from 29 Anangu Women’. Refereed paper delivered at the Australian Political Studies Association conference, Brisbane, 6–9 July.

Banducci, S. A. and J. A. Karp. 1999. ‘Perceptions of Fairness and Support for Proportional Representation’. Political Behavior 21(3): 217–238.

Bean, C. 2005. ‘Young People’s Voting Patterns’. Paper prepared for the Youth Electoral Study Workshop, Old Parliament House, Canberra, June.

Beetham, D., S. Bracking, I. Kearton and S. Weir. 2001. International IDEA Handbook on Democracy Assessment . The Hague: Kluwer.

Beetham, D., E. Carvalho, T. Landman and S. Weir. 2008. Assessing the Quality of Democracy: A Practical Guide . Stockholm: International IDEA.

Bennett, S. 2008. Compulsory voting in Australian national elections . Research Brief No. 6, 2005–06 (revised 2008). Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 14 October 2008.

Benoit, K. 2004. ‘Models of electoral system change’. Electoral Studies 23(3): 363–389.

Blais, A., L. Massicotte and A. Yoshinaka. 2001. ‘Deciding who has the right to vote: a comparative analysis of election laws’. Electoral Studies 20(1): 41–62.

Blanc, J., S. Clift, J. Grace, L. Handley and M. Walecki. 2007. Challenging the Norms and Standards of Election Administration . Washington: International Foundation for Electoral Systems (IFES). URL: . Retrieved 14 February 2008.

Blau, A. 2004. ‘Fairness and Electoral Reform’. British Journal of Politics and International Relations 6(2): 165–81.

Bogdanor, V. and D. Butler, eds. 1983. Democracy and Elections: Electoral systems and their political consequences . Cambridge: Cambridge University Press.

Boix, C. 1999. ‘Setting the Rules of the Game: The Choice of Electoral Systems in Advanced Democracies’. American Political Science Review 93(3): 609–24.

Bowler, S., T. Donovan and J. A. Karp. 2006. ‘Why Politicians Like Electoral Institutions: Self- Interest, Values, or Ideology?’. The Journal of Politics . 68(2): 434–446.

Brent, P. 2007. ‘2007 Elections –– Provisional Voting Rejections’. Democratic Audit of Australia, Australian National University. URL: . Retrieved 16 July 2008.

223

Brent, P. and S. Jackman. 2007. A shrinking Australian electoral roll? Democratic Audit of Australia, Australian National University. URL: . Retrieved 18 July 2008.

Buxton, J. 1985. ‘Western Australia, 1984: Some Problems in Representation’. Paper presented at the Parliament and Representation Conference, Nedlands, Western Australia.

Campbell, I. 2006. ‘Senate Finance and Public Administration Legislation Committee hearing’, 7 March, Hansard , pp. 1-26. Canberra: Parliament of Australia. URL: . Retrieved 14 November 2006.

Chamarette, C. 1995. ‘Commonwealth Electoral Amendment Bill 1995 –– Second reading speech’, 11 May, in Hansard , URL: . Retrieved 21 September 2008.

Charnock, D. 1994. ‘Electoral Bias in Australia 1980-1993: The Impact of the 1983 Electoral Amendments’. Australian Journal of Political Science 29(3): 484–500.

Colomer, J. M., ed. 2004. Handbook of Electoral System Choice . Basingstoke: Palgrave Macmillan.

Colomer, J. M. 2005. ‘It's Parties That Choose Electoral Systems (or, Duverger's Laws Upside Down)’. Political Studies 53(1): 1–21.

Copeman, C. and A. McGrath (eds). 1997. Corrupt elections: recent Australian studies and experiences of ballot rigging . Kensington: Tower House Publications.

Corbell, S. 2008. Electoral Legislation Amendment Bill 2007: Government Amendments –– Supplementary Explanatory Statement . Canberra: ACT Legislative Assembly. URL: . Retrieved 30 October 2008.

Costar, B. 1992. ‘Constitutional Change’, in Trials in Power: Cain, Kirner and Victoria, 1982– 1992 , eds M. Considine and B. Costar. Carlton: Melbourne University Press.

Costar, B. 2006. Submission (No.2) to the Senate Finance and Public Administration Committee inquiry . Canberra: Parliament of Australia. URL: . Retrieved 16 June 2007.

Courtney, J. C. 2004. Elections . Vancouver: UBC Press.

Cox, G. 1997. Making Votes Count: Strategic Coordination in the World’s Electoral Systems . New York: Cambridge University Press.

Creswell, J. W. 1998. Qualitative Inquiry and Research Design: Choosing among five traditions . Thousand Oaks: Sage.

Dacey, P. 2005. ‘What do “Impartiality’, “Independence” and “Transparency” Mean? –– Some Thoughts From Australia’. Paper delivered at the Improving the Quality of Election Management Conference of Commonwealth Chief Election Officers, New Delhi, India.

Darcy, R. 1986. ‘Position Effects with Party Column Ballots’. Western Political Quarterly 39(4): 648-662.

Darcy, R. and M. Mackerras. 1993. ‘Rotation of ballots: Minimizing the number of rotations’. Electoral Studies 12(1): 77–82.

224

Darcy, R. and I. McAllister. 1990. ‘Ballot position effects’. Electoral Studies 9(1): 5–17.

Dario, G. 2005. Analysis of Informal Voting During the 2004 House of Representatives Election . Research Report Number 7. Canberra: Australian Electoral Commission. URL: . Retrieved 24 June 2007.

Davidson, J. 2004. Inside outcasts: prisoners and the right to vote in Australia . Current Issues Brief No. 12, 2003-04. Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 14 December 2007.

Denzin, N. K. 1970. The research act: A theoretical introduction to sociological methods . Chicago: Aldine Publishing Company.

Dummett, M. 1997. Principles of Electoral Reform . Oxford: Oxford University Press.

Dundas, C.W. 1994. Dimensions of Free and Fair Elections: Frameworks, Integrity, Transparency, Attributes and Monitoring . London: Commonwealth Secretariat.

Duverger, M. 1951. Political Parties: Their Organization and Activity in the Modern State . New York: Wiley.

Economou, N. 2007. ‘What happened to rural weightage: Malapportionment and the reform of Australian electoral systems under Labor’. Paper delivered at the Australasian Political Studies Association conference, Melbourne.

ECQ [Electoral Commission of Queensland]. 2002. Queensland Election 2001: Ballot Paper Survey . Brisbane: Electoral Commission of Queensland. URL: . Retrieved 17 February 2007.

ECQ. 2005. 2004 Queensland Election Ballot Paper Survey . Brisbane: Electoral Commission of Queensland. URL: . Retrieved 17 February 2007.

ECQ. 2007. 2006 Queensland Election Ballot Paper Survey . Brisbane: Electoral Commission of Queensland. URL: . Retrieved 24 November 2007.

Edwards, K., L. Saha, and M. Print. 2006. Youth Electoral Study - Report 3: Youth, The Family, and Learning About Politics and Voting . Canberra: Australian Electoral Commission. URL: . Retrieved 24 June 2007.

EFA [Election Funding Authority]. 2007. Annual Report 2006-2007 . Sydney: Election Funding Authority. URL: . Retrieved 21 January 2008.

Electoral Matters Committee. 2008. Inquiry into the conduct of the 2006 Victorian state election and matters related thereto: Report to Parliament Electoral Matters Committee . Melbourne: Victorian Government Printer.

Elkins, D. J. 1992. ‘Electoral Reform and Political Culture’. In Comparative Political Studies: Australia and Canada , eds M. Alexander and B. Galligan. Melbourne: Pitman.

225

Elklit, J. and A. Reynolds. 2002. ‘The Impact of Election Administration on the Legitimacy of Emerging Democracies: A New Comparative Politics Research Agenda’. Commonwealth & Comparative Politics 40(2): 86–119.

Elklit, J. and A. Reynolds. 2005. ‘A Framework for the Systematic Study of Election Quality’. Democratization 12(2): 147–162.

Ellis, A., M. Gratschew, J. H. Pammett and E. Thiessen. 2006. Engaging the Electorate: Initiatives to Promote Voter Turnout From Around the World: Including voter turnout data from national elections worldwide 1945–2006 . Stockholm: International IDEA.

Engstrom, R. L. 2005. Revising Constituency Boundaries in the United States and Australia: It Couldn’t be More Different . Democratic Audit of Australia, Australian National University. URL: . Retrieved 18 September 2006.

Evans, G. 2005. Fairness and Unfairness in South Australian Elections. Unpublished Master of Arts Thesis, School of History and Politics, Faculty of Humanities and Social Sciences. Adelaide: University of Adelaide.

Evans, T. 2006. Compulsory Voting in Australia . Canberra: Australian Electoral Commission.

Farrell, D. 2000. Discussion Paper No 2: Modernising Tasmania’s Electoral Act . Hobart: Tasmanian Electoral Office. URL: . Retrieved 18 August 2007.

Farrell, D. M. and I. McAllister. 2003. ‘The 1983 change in surplus vote transfer procedures for the Australian Senate and its consequences for the single transferable vote’. Australian Journal of Political Science 38(3): 479–491.

Farrell, D. M. and I. McAllister. 2005. ‘Australia: The Alternative Vote in a Compliant Political Culture’, in The Politics of Electoral Systems , eds M. Gallagher and P. Mitchell. Oxford: Oxford University Press.

Farrell, D. M. and I. McAllister. 2006. The Australian Electoral System: Origins, Variations and Consequences . Sydney: University of New South Wales Press.

Faulkner, Senator J. 2004. ‘Second Reading Speech on the Electoral and Referendum Amendment (Enrolment Integrity and Other Measures) Bill 2004’, 25 June, Hansard , pp. 25136–40. Canberra: Parliament of Australia. URL: . Retrieved 16 June 2007.

Faulkner, Senator J. 2008a. New Arrangements for Merit and Transparency in Senior Public Service Appointments . Special Minister of State Media Release, 5 February. URL: . Retrieved 15 February 2008.

Faulkner, Senator J. 2008b. Electoral Reform . Special Minister of State Media Release, 28 March. URL: < http://www.smos.gov.au/media/2008/mr_062008.html>. Retrieved 29 March 2008.

Fitzgerald, G. E. 1989. Report of a Commission of Inquiry Pursuant to Orders in Council: Commission of Inquiry into Possible Illegal Activities and Associated Police Misconduct . Brisbane: Queensland Government.

Fitzgerald, S. 2005. Ending Felon Disenfranchisement . Democratic Audit of Australia, Australian National University. URL: . Retrieved 23 September 2007.

226

Gallagher, M. and P. Mitchell, eds. 2005. The Politics of Electoral Systems . Oxford: Oxford University Press.

Garber, L. 1994. ‘Election Commissions: Responsibilities and Composition’. Paper presented at the NDI-sponsored African Election Colloquium, Victoria Falls, Zimbabwe.

Goodin, R. E. 1996. ‘Theories of Institutional Design’, in The Theory of Institutional Design , ed. R. E. Goodin. Cambridge: Cambridge University Press.

Goodwin-Gill, G. S. 1998. Codes of conduct for elections: a study prepared for the Inter- parliamentary Union . Geneva: Inter-parliamentary Union.

Goot, M. 1985. ‘Electoral Systems’, in Surveys of Australian Political Science , ed. D. Aitken. North Sydney: George Allen & Unwin.

Goot, M. 2006. ‘The Australian party system, Pauline Hanson’s One Nation and the party cartelisation thesis’, in Political Parties in Transition , ed. I. Marsh. Sydney: The Federation Press.

Green, A. 2003. Prospects for the 2003 Legislative Council Election. Background Paper No. 3/03, NSW Parliamentary Library Research Service. Sydney: NSW Parliament.

Green, P. 1986. The Australian Labor Party and the Commonwealth Electoral System 1972- 1986. Unpublished Litt. B. Sub-Thesis. Canberra.

Gribbin, C. 2004. Lowering the Voting Age . Melbourne: Victorian Electoral Commission. URL: . Retrieved 24 May 2007.

Grofman, B. 1990. Political Gerrymandering and the Courts . New York: Agamon Press.

Hann, L. 2006. Informal Voting at State and Territory Elections . Research Report Number 10. Canberra: Australian Electoral Commission. URL: . Retrieved 24 June 2007.

Harris, P. 1997. An Electoral Administration: Who, What and Where . Paper prepared at IDEA for the South Pacific Electoral Administrators’ conference in Fiji, October.

Harrop, M. and W. L. Miller. 1987. Elections and Voters: A Comparative Introduction . Basingstoke: Macmillan Education.

Harwin, D. 2007. ‘Select Committee on Electoral and Political Party Funding: Establishment’, 27 June, in Hansard , Parliament of New South Wales website. URL: . Retrieved 27 July 2008.

Hasen, R. L. 2004. Ending Felon Disenfranchisement in the United States: Litigation or Legislation? Democratic Audit of Australia, Australian National University. URL: . Retrieved 23 September 2007.

HCA [High Court of Australia]. 2007a. Roach v Electoral Commissioner & Anor [2007] HCA Trans 275 (12 June 2007) . URL: . Retrieved 29 June 2007.

HCA. 2007b. Roach v Electoral Commissioner & Anor [2007] HCA Trans 276 (13 June 2007) . URL: . Retrieved 29 June 2007.

227

HCA. 2007c. Roach v Electoral Commissioner [2007] HCA 43 (26 September 2007) . URL: . Retrieved 28 September 2007.

Hill, L. 2000. ‘Precarious Persons: Disenfranchising Australian Prisoners’, Australian Journal of Social Issues 35 (3): 203–13.

Hill, L. 2004. ‘Compulsory Voting in Australia: A Basis for a 'Best Practice' Regime’. Federal Law Review 32(3): 479–497.

Hodgman, M. 2004. ‘Second Reading Debate, Electoral Bill 2004 (No. 63)’, 26 October, Hansard. Hobart: Parliament of Tasmania. URL: . Retrieved 14 June 2007.

House of Assembly. 1986. General Elections 1985 . Adelaide: Government Printer.

Hughes, C. A. and B. D. Graham. 1976. Voting for the South Australian, Western Australian and Tasmanian Lower Houses 1890–1964 . Canberra: Department of Political Sciences, Research School of Social Sciences, Australian National University.

Hughes, C. A. 1977. A Handbook of Australian Government and Politics, 1965–1974 . Canberra: Australian National University Press.

Hughes, C. A. 1978. ‘Fair and Equal Constituencies: Australia, Jamaica and the United Kingdom’. Journal of Commonwealth and Comparative Politics 16(3): 256–271.

Hughes, C. A. 1986. A Handbook of Australian Government and Politics 1975–84 . Sydney: Australian National University Press.

Hughes, C. 2001. ‘Institutionalising electoral integrity’, in Elections: Full, Free & Fair , ed. M. Sawer. Sydney: The Federation Press.

Hughes, C. and B. Costar. 2005. ‘Fiddling the ballot books’. The Age , 3 November.

Hughes, C. and B. Costar. 2006. Limiting democracy: the erosion of electoral rights in Australia . Sydney: University of New South Wales Press.

Hughes, C. 2007. The importance of boundaries . Research Paper 1, Democratic Audit of Australia, Australian National University. URL: . Retrieved 5 December 2007.

IDEA [International Institute for Democracy and Electoral Assistance]. 1999. Code of Conduct: Ethical and Professional Administration of Elections . Stockholm: International IDEA.

IDEA. 2002. International Electoral Standards: Guidelines for reviewing the legal framework of elections . Stockholm: International IDEA.

Jaensch, D. 1974. Election Statistics of Colonial South Australia . Bedford Park: Flinders University.

Jaensch, D. 1977. The 1977 Elections in South Australia: A Statistical Analysis. Occasional Monograph No. 6, Discipline of Politics. Bedford Park: Flinders University.

Jaensch, D. 1986. The 1986 Elections in Western Australia: A Statistical Analysis . Monograph No. 2. Klemzig: Polity Publications.

Jaensch, D. 1995. Election! How and Why Australia Votes . Sydney: Allen & Unwin.

228

Jaensch, D. 2002. Community Access to the Electoral Processes in South Australia since 1850 . Adelaide: State Electoral Office.

Johnston, M. 2005. Political Parties and Democracy in Theoretical and Practical Perspectives: Political Finance Policy, Parties, and Democratic Development . Washington: National Democratic Institute for International Affairs.

JSCEM [Joint Standing Committee on Electoral Matters]. 1989. Inquiry into the ACT Election and Electoral System . Canberra: Australian Government Publishing Service.

JSCEM. 1991. 1990 Federal election: Report from the Joint Standing Committee on Electoral Matters . Canberra: Australian Government Publishing Service.

JSCEM. 1994. Financial reporting by Political Parties: Interim Report from the Joint Standing Committee on Electoral Matters on the Inquiry into the Conduct of the 1993 Election and Matters Related Thereto . Canberra: Parliament of the Commonwealth of Australia.

JSCEM. 1997. Report of the Inquiry into All Aspects of the Conduct of the 1996 Federal Election and Matters Related Thereto . Canberra: Commonwealth of Australia. URL: . Retrieved 16 February 2006.

JSCEM. 2000. The 1998 Federal Election: Report of the Inquiry into the Conduct of the 1998 Federal Election and Matters Related Thereto . Canberra: Commonwealth of Australia. URL: . Retrieved 16 February 2006.

JSCEM. 2003. The 2001 Federal Election: Report of the Inquiry into the Conduct of the 2001 Federal Election, and matters related thereto . Canberra: Commonwealth of Australia.

JSCEM. 2005a. The 2004 Federal Election: Report of the Inquiry into the Conduct of the 2004 Federal Election and Matters Related Thereto . Canberra: Commonwealth of Australia.

JSCEM. 2005b. Official Committee Hansard, Joint Standing Committee on Electoral Matters, Thursday 7 July, Tweed Heads . Canberra: Commonwealth of Australia. URL: . Retrieved 16 July 2008.

JSCEM. 2005c. Official Committee Hansard, Joint Standing Committee on Electoral Matters, Wednesday 6 July, Brisbane . Canberra: Commonwealth of Australia. URL: . Retrieved 23 August 2008.

JSCEM. 2008. Official Committee Hansard, Joint Standing Committee on Electoral Matters, Friday, 27 June, 2008, Canberra . Canberra: Commonwealth of Australia. URL: . Retrieved 18 September 2008.

JSCER [Joint Select Committee on Electoral Reform]. 1983. First Report . Canberra: Australian Government Publishing Service.

Katz, R. S., and P. Mair. 1995. ‘Changing Models of Party Organization and Party Democracy’. Party Politics 1(1): 5–28.

Kelley, J. and I. McAllister. 1984. ‘Ballot Paper Cues and the Vote in Australia and Britain - Alphabetic Voting, Sex, and Title’. Public Opinion Quarterly 48(2): 452-466.

Kelly, N. 2004. Vote Weighting in Electoral Systems: The ‘One Vote, One Value’ Debate in Western Australia. Unpublished Honours Thesis, Faculty of Media, Society and Culture. Perth: Curtin University of Technology.

229

Kelly, N. 2006a. ‘The Howard Government’s Electoral Reforms –– a Blight on Democratic Principles’. Paper delivered at the Australasian Political Studies Association conference, Newcastle.

Kelly. 2006b. ‘Western Australian Electoral Reforms: Labor Finally Succeeds’. Australian Journal of Political Science. 41(3): 419–26.

Kelly, N. and M. Sawer. 2008. ‘The state of democracy: The view from the Democratic Audit of Australia’. Revista catalana de dret public . 37 (forthcoming).

Kerr, B. 2005. The State Franchise in Victoria, 1842–2005 . Professional Intern Report. Melbourne: Victorian Electoral Commission. URL: . Retrieved 14 December 2007.

Lee, R. M. 1993. Doing Research on Sensitive Topics . London: Sage Publications.

Liebowitz, S. and S. E. Margolis. 1995. ‘Path Dependence, Lock-in and History’. Journal of Law, Economics, and Organization 11: 205–226.

Lijphart, A. 1994. Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies 1945–1990 . Oxford: Oxford University Press.

Lijphart, A. 1999. Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries . New Haven: Yale University Press.

López-Pintor, R. 2000. Electoral Management Bodies as Institutions of Governance . Bureau for Development Policy, United Nations Development Programme. URL: . Retrieved 3 July 2008.

Louth, J. and L. Hill. 2005. ‘Compulsory voting in Australia: Turnout with and without it’. Australian Review of Public Affairs . 6(1): 25–37.

Macilwain, M. 2007. History of the State Electoral Office 1907–2007. Adelaide: State Electoral Office.

Maley, M. 1980. Probability and Electoral Bias . Department of Political Science, School of General Studies. Canberra: Australian National University.

Mansell, M. 1993. ‘Aborigines and the Constitution’. Human Rights Defender . 2(4): 27–8.

Marsh, I. (ed.). 2006. Political Parties in Transition . Sydney: The Federation Press.

Martin, C. 2004. ‘Electoral Bill (Serial 195) –– 2 nd Reading Debate’. Hansard . Darwin: Legislative Assembly of the Northern Territory. URL: . Retrieved 6 February 2008.

Massicotte, L., A. Blais, and A. Yoshinaka. 2004. Establishing the Rules of the Game: Election Laws in Democracies . Toronto: University of Toronto Press.

McAllister, I., T. Makkai, and C. Patterson. 1992. Informal voting in the 1987 and 1990 Australian federal elections . Canberra: Australian Government Publishing Service.

McGrath, A. 1996. The Frauding of Votes . Kensington: Tower House Publications.

230

McNeill, I. 2004. New Electoral Legislation for the Northern Territory . Darwin: Legislative Assembly of the Northern Territory.

MEC [Minter Ellison Consulting]. 2003. Independent Review of the Northern Territory Electoral System: Final Report and Recommendations . Canberra: Minter Ellison Consulting.

Medew, R. 2003. Informal Vote Survey, House of Representatives, 2001 Election . Research Report Number 1. Canberra: Australian Electoral Commission. URL: . Retrieved 16 April 2007.

Mercurio, B. and G. Williams. 2004. ‘Australian Electoral Law: 'Free and Fair'?’ Federal Law Review 32(3): 365–395.

Millar, R. and N. McKenzie. 2008. ‘New law set to clean up electoral funding rules’. The Age , 20 September.

Minchin, N. 2004. ‘Electoral and Referendum Amendment (Enrolment Integrity and Other Measures) Bill 2004, 2 nd Reading Debate’. Hansard , 24 June, 25136–51.

Minichiello, V., R. Arioni, E. Timewell, and L. Alexander. 1992. In-Depth Interviewing: Researching People . Melbourne: Longman Cheshire.

Miskin, S. and G. Baker. 2006. Political finance disclosure under current and proposed thresholds . Research Note 27 2005–06, Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 14 November 2006.

Moore, N. F. 2000. Electoral Amendment Bill 2000 –– 2 nd Reading Speech . Parliamentary Debates, Hansard , Vol. 363: 8530–1. Perth: Parliament of Western Australia.

Mozaffar, S. and A. Schedler. 2002. ‘The Comparative Study of Electoral Governance –– An Introduction’. International Political Science Review 23(1): 5–27.

Murray, A. 2003. ‘Electoral Amendment (Political Honesty) Bill 2003 –– 2 nd Reading’. Hansard , 27 March, 10323–5.

Nelson, P. 2007. Electoral division rankings: Census 2006 first release . Research Paper 12 2007–08, Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 28 November 2007.

Newman, G. 1999. Senate Group Voting Tickets . Research Note 6 1999–2000, Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 5 December 2007.

Newman, G. 2004. Analysis of Declaration Voting . Research Paper Number 3, June 2004. Canberra: Australian Electoral Commission.

Newton-Farrelly, J. 2008. ‘From Jerry-built to Purpose-built: creating fair and equal electoral districts in South Australia’. Paper delivered at the Australasian Study of Parliament Group conference, Brisbane, July.

Norberry, J. and G. Williams. 2002. Voters and the Franchise: the Federal Story . Research Paper No. 17, 2001–02, Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 14 December 2007.

Norris, P. 2004. Electoral Engineering: Voting Rules and Political Behavior . Cambridge: Cambridge University Press.

231

NSWLC [New South Wales Legislative Council]. 2007. Hansard transcript, 26 September. Sydney: New South Wales Parliament. URL: . Retrieved 5 September 2008.

NTEC [Northern Territory Electoral Commission]. 2007. 2005 Legislative Assembly General Election Report: Part 1 of 2 . Darwin: Northern Territory Electoral Commission.

NTEC. 2008. Candidate’s Handbook: Legislative Assembly Elections . Darwin: Northern Territory Electoral Commission.

NTLA [Northern Territory Legislative Assembly]. 2004. ‘Electoral Bill 2004 –– 2 nd Reading Debate’. Hansard . URL: . Retrieved 13 April 2008.

Offe, C. 1996. ‘Designing Institutions in East European Transitions’, in The Theory of Institutional Design , ed. R. E. Goodin. Cambridge: Cambridge University Press.

Orr, G. 1998. ‘Ballotless and Behind Bars: The Denial of the Franchise to Prisoners’. Federal Law Review 26(1), 56–82.

Orr, G. 2004. Australian Electoral Systems –– How Well Do They Serve Political Equality? Canberra: Democratic Audit of Australia, Australian National University.

Orr, G. 2007. Constitutionalising the franchise and the status quo: The High Court on prisoner voting rights . Discussion Paper 19/07, Democratic Audit, Australian National University. URL: . Retrieved 23 September 2007.

Orr, G. 2008. ‘Citizenship, Interests, Community and Expression: Expatriate Voting Rights in Australian Elections’, in Citizenship in a Post-National World: Australia and Europe compared , eds S. Bronitt and K. Rubinstein. Sydney: The Federation Press.

Orr, G. and G. Williams. 2001. ‘Electoral Challenges: Judicial Review of Parliamentary Elections in Australia’. Sydney Law Review 23(1): 53–94.

Orr, G., B. Mercurio, and G. Williams. 2003. ‘Australian Electoral Law: A Stocktake’. Election Law Journal 2(3): 383–402.

PAEC [Public Accounts and Estimates Committee]. 2006. Report on –– A legislative framework for independent officers of Parliament . Melbourne: Parliament of Victoria.

Pettit, P., T. Keady, and B. Blick. 1998. Review of the Governance of the Australian Capital Territory . Canberra: Chief Minister’s Department.

Phillips, H. 1991. ‘The Modern Parliament, 1965–1989’, in Black (ed.), The House on the Hill: A History of the Parliament of Western Australia 1832–1990 . Perth: Parliament of Western Australia.

Phillips, H. and K. Robinson. 2006. The Quest For ‘One Vote One Value’ In Western Australia’s Political History . Perth: Western Australian Electoral Commission.

Pierson, P. 2000. ‘Increasing returns, path dependence, and the study of politics’. American Political Science Review 94(2): 251–267.

232

Print, M., L. Saha, and K. Edwards. 2004. Youth Electoral Study - Report 1: Enrolment and Voting . Canberra: Australian Electoral Commission. URL: . Retrieved 14 October 2006.

Print, M., L. Saha, and K. Edwards. 2005. Youth Electoral Study - Report 2: Youth, Political Engagement and Voting . Canberra, Australian Electoral Commission. URL: . Retrieved 14 October 2006.

Rae, D. W. 1967. The Political Consequences of Electoral Laws . Clinton: Colonial Press.

Reilly, B. 2002. ‘Political Engineering and Party Politics in Papua New Guinea’. Party Politics 8(6): 701–18.

Reilly, B., P. Nordlund and E. Newman. 2008. Political Parties in Conflict-Prone Societies: Encouraging Inclusive Politics and Democratic Development . Policy Brief No. 2, Tokyo: United Nations University.

Reynolds, A. and B. Reilly. 2002. The International IDEA Handbook of Electoral System Design . Stockholm: International IDEA.

Rokkan, S. 1970. Citizens, Elections, Parties: Approaches to the Comparative Study of the Processes of Development . New York: McKay.

Rose, R. 2000. ‘Administration of Elections’, in International Encyclopedia of Elections , ed. R. Rose. London: Macmillan.

Rossman, G. B., and S. F. Rallis. 1998. Learning in the Field: An introduction to qualitative research . Thousand Oaks: Sage Publications.

Sawer, M. 2001a. ‘Pacemakers for the world?’, in Elections: Full, Free and Fair , ed. M. Sawer. Sydney: The Federation Press.

Sawer, M (ed.). 2001b. Elections: Full, Free and Fair . Sydney: The Federation Press.

Sawer, M. 2006a. The Democratic Audit of Australia: Populism vs Citizen Rights . Democratic Audit of Australia, Australian National University. URL: . Retrieved 18 July 2008.

Sawer, M. 2006b. Damaging Democracy? Early Closure of Electoral Rolls . Democratic Audit of Australia, Australian National University. URL: . Retrieved 16 June 2007.

SCEPPF [Select Committee on Electoral and Political Party Funding]. 2008a. Electoral and Political Party Funding in New South Wales . Report No. 1. Sydney: Legislative Council, New South Wales Parliament.

SCEPPF. 2008b. Hearing Transcript, 3 March. URL: . Retrieved 19 June 2008.

SCETYP [Standing Committee on Education, Training and Young People]. 2007. Inquiry into the Eligible Voting Age . Report No. 5. Canberra: Legislative Assembly for the Australian Capital Territory.

233

Schwartz, H. 2004. ‘Down the Wrong Path: Path Dependence, Increasing Returns, and Historical Institutionalism’. Charlottesville: University of Virginia. URL: . Retrieved 14 March 2006.

SCLA [Standing Committee on Legal Affairs]. 2002. The Appropriateness of the Size of the Legislative Assembly for the ACT and Options for Changing the Number of Members, Electorates and any other related matter . Report No. 4. Canberra: Legislative Assembly for the Australian Capital Territory.

SCRRG [Select Committee on the Report of the Review of Governance]. 1999. Report of the Select Committee on the Report of the Review of Governance . Canberra: Legislative Assembly for the Australian Capital Territory.

Selltiz, C., L. W. Wrightsman, and S. W. Cook. 1976. Research Methods in Social Relations . New York: Holt, Rinehart and Winston.

SEO [State Electoral Office, South Australia]. Election Report for the South Australian Elections: 9 February 2002 . Adelaide: State Electoral Office. URL: . Retrieved 17 July 2007.

SEO. 2004. Report on Optional Preferential Voting 2003 –– 2004. Adelaide: State Electoral Office. URL: . Retrieved 24 February 2007.

SEO. 2007a. State Electoral Office: Annual Report 2006-07 . URL: . Retrieved 3 February 2008.

SEO. 2007b. Election Report: South Australian Election: 18 March 2006 . Adelaide: State Electoral Office. URL: . Retrieved 16 December 2007.

SFPALC [Senate Finance and Public Administration Legislation Committee]. 2002. Charter of Political Honesty Bill 2000 . Canberra: Parliament of Australia. URL: . Retrieved 14 July 2008.

SFPALC. 2006. Provisions of the Electoral and Referendum Amendment (Electoral Integrity and Other Measures) Bill 2005 . Canberra: Senate Printing Unit, Parliament House.

Simms, M. 2006. From the Hustings to Harbour Views . Sydney: UNSW Press.

Slipper, P. 1995. ‘Commonwealth Electoral Amendment Bill 1995 –– Second reading speech’, 9 March, in Hansard . URL: . Retrieved 21 September 2008.

Smith, L. 1996. Research methods: qualitative analyses. Unpublished manuscript. Perth: Curtin University of Technology.

Smith, R. and J. O’Mahony. 2006. ‘The cartel parties model and electoral barriers’, in Political Parties in Transition , ed. I. Marsh. Sydney: The Federation Press.

Snowdon, W. 2004. ‘Senate victory used to silence voters’. Media release, 20 December. URL: . Retrieved 17 November 2007.

Soper, C. S. and J. Rydon. 1958. ‘Under-representation and election prediction’. Australian Journal of Politics and History 4: 94–106.

234

SSCFPA [Senate Standing Committee on Finance and Public Administration]. 2007. Electoral and Referendum Legislation Amendment Bill 2007 . Canberra: Senate Printing Unit, Parliament House.

Stake, R. 1995. The Art of Case Study Research . Thousand Oaks: Sage.

Strangio, P. 2004. ‘Labor and Reform of the Victorian Legislative Council, 1950-2003’. Labour History 86: 33–52.

Tasmanian Parliament. 2007. Parliamentary Reform –– Downsizing Parliament . Information sheet. URL: . Retrieved 18 June 2007.

The Age . 2007. ‘Votes for prisoners?: Should prisoners be allowed to vote?’. Online poll. URL: . Retrieved 23 January 2008.

Tsebelis, G. 1990. Nested Games: Rational Choice in Comparative Politics . Berkeley: University of California Press.

Uhr, J. 2000. Rules for Representation: Parliament and the Design of the Australian Electoral System . Research Paper 29 1999–2000, Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 14 April 2007.

VEC [Victorian Electoral Commission]. 2007. Report to Parliament on the 2006 Victorian State election . Melbourne: Victorian Electoral Commission.

Vromen, A. and N. Turnbull. 2006. ‘The Australian Greens: Challengers to the cartel?’, in Political Parties in Transition , ed. I. Marsh. Sydney: The Federation Press.

Walker, F. 1995. ‘Commonwealth Electoral Amendment Bill 1995 –– Second reading speech’, 9 March, in Hansard , Australian Parliament House web site. URL: . Retrieved 21 September 2008.

Wall, A., A. Ellis, A. Ayoub, C. W. Dundas, J. Rukambe, and S. Staino. 2006. Electoral Management Design: The International IDEA Handbook . Stockholm: International IDEA.

Wanna, J. 2004. ‘Democratic and Electoral Shifts in Queensland: Back to First Past the Post Voting’. Discussion Paper, Democratic Audit of Australia, Australian National University. URL: . Retrieved 24 September 2008.

Ward, I. 2006. ‘Cartel parties and election campaigns in Australia’, in Political Parties in Transition , ed. I. Marsh. Sydney: The Federation Press.

Western Australia Legislative Council. 1983. Statistics relating to The General Election held 19 th February, 1983 . Perth: Government Printer.

Williams, G. 1997. ‘Truth in Political Advertising Legislation in Australia’. Research Paper 13 1996-97, Parliamentary Library, Parliament of Australia, Canberra. URL: . Retrieved 18 November 2007.

Williams, G. 2006. ‘Electoral Changes Going Backwards’, in the Sydney Morning Herald , Sydney, 23 June.

Young, S. 2003. ‘Killing competition: restricting access to political communication channels in Australia’. AQ Journal of Contemporary Analysis 75(3): 9–15.

235

Young, S. and J-C. Tham. 2006. Political finance in Australia: a skewed and secret system . Canberra: Democratic Audit of Australia, Australian National University.

236

237

Websites

Australasian Legal Information Institute (Austlii) http://www.austlii.edu.au

Australian Capital Territory Electoral Commission (ACTEC) http://www.elections.act.gov.au

Australian Electoral Commission (AEC) http://www.aec.gov.au

Australian Government and Politics Database (AGPD) http://elections.uwa.edu.au

Electoral Commission of Queensland (ECQ) http://www.ecq.qld.gov.au

Northern Territory Electoral Office (NTEC) http://www.ntec.nt.gov.au

New South Wales Electoral Commission (NSWEC) http://www.elections.nsw.gov.au

South Australian State Electoral Office (SEO) http://www.seo.sa.gov.au

Tasmanian Electoral Commission (TEC) http://www.tec.tas.gov.au

Victorian Electoral Commission (VEC) http://www.vec.vic.gov.au

Western Australian Electoral Office (WAEC) http://www.waec.wa.gov.au

238

239

Appendices

Appendix A List of Interview Participants

Interviewee Position Date Interviewed Commonwealth Ian Campbell Electoral Commissioner, Australian Electoral 15 January 2007 Commission Gary Nairn Special Minister of State 26 March 2007 Alan Griffin Shadow Special Minister of State 23 May 2007 Andrew Murray Australian Democrats Senator; Member, Joint 29 May 2007 Standing Committee on Electoral Matters Daryl Wight Australian Electoral Officer, Victoria, Australian 27 July 2007 Electoral Commission Colin Hughes Electoral Commissioner, Australian Electoral 25 June 2007 Commission (1984–89) NSW Colin Barry Electoral Commissioner 23 July 2007 Victoria Steve Tully Electoral Commissioner 21 June 2007 Queensland David Kerslake Electoral Commissioner, Electoral Commission of 28 June 2007 Queensland Garry Wiltshire Director, Election Operations, Electoral Commission 28 June 2007 of Queensland Dianne Reilly Chair, Legal, Constitutional and Administrative 27 June 2007 Review Committee Betty Kiernan Member, Legal, Constitutional and Administrative 27 June 2007 Review Committee Lesley Clark Chair, Legal, Constitutional and Administrative 1 March 2007 Review Committee (2004–06); Member, Parliamentary Committee for Electoral and Administrative Review (1990–95; Chair 1992–95) Western Australia Warwick Gately Electoral Commissioner, Western Australian Electoral 16 April 2007 Commission Jim McGinty Attorney-General, Minister for Electoral Affairs 11 April 2007 Dan Sullivan Shadow Minister for Electoral Affairs (2001–05) 18 April 2007 South Australia Kay Mousley Electoral Commissioner, State Electoral Office 13 April 2007 Dave Gully Deputy Electoral Commissioner, State Electoral 13 April 2007 Office Di Walker Senior Electoral Officer (Education), State Electoral 13 April 2007 Office Michael Atkinson Attorney-General 9 August 2007 Isobel Redmond Shadow Attorney-General 12 April 2007 Robert Lawson Member, Statutory Officers Committee 10 April 2007 Tasmania Bruce Taylor Electoral Commissioner 20 June 2007 Andrew Hawkey Manager, Education and Information Systems, 20 June 2007 Tasmanian Electoral Commission Nicole Lugg Senior Policy Analyst, Tasmanian Electoral 20 June 2007 Commission Steve Kons Attorney-General 19 June 2007 Michael Hodgman Shadow Attorney-General 19 June 2007

240

Australian Capital Territory Phillip Green Electoral Commissioner 13 November 2006 Simon Corbell Attorney-General 6 February 2007 Bill Stefaniak Leader of the Opposition, Shadow Attorney-General 28 November 2006 Northern Territory Bill Shepheard Electoral Commissioner 22 March 2007

241

Appendix B JSCER and JSCEM (Commonwealth) Reports

Year Report Name 1983* First report 1984* Second report 1986* Determining the entitlement of Federal Territories and new States to representation in the Commonwealth Parliament 1987* The operation during the 1984 General Election of the 1983-84 amendments to Commonwealth electoral legislation 1988 # One vote, one value: Inquiry into the Constitution Alteration (Democratic Elections) Bill 1987 1988 'Is this where I pay the electricity bill?': Inquiry into report on efficiency scrutiny into regionalisation within Australian Electoral Commission 1989 Inquiry into the conduct of the 1987 Federal election and 1988 referendums 1989 Who pays the piper calls the tune: Minimising the risks of funding political campaigns – – Inquiry into conduct of the 1987 Federal Election and 1988 referendums 1989 Inquiry into the ACT election and electoral system 1990 The 1990 Federal Election 1991 Aboriginal and Islander Electoral Information Service 1992 The conduct of elections: New boundaries for cooperation 1992 Counting the vote on election night: Conduct of the 1990 Federal Election, Part II, and preparations for the next election 1992 Ready or not: Refining the process for election '93 - Conduct of the 1990 Federal Election, Part II, and preparations for the next election 1994 Women, elections and Parliament 1994 Financial reporting by political parties: Interim report of the inquiry into the 1993 election and matters related thereto 1994 Conduct of the 1993 Federal election 1996 Electoral Redistribution: Report on the effectiveness and appropriateness of the redistribution provisions of part iii and iv of the Commonwealth Electoral Act 1918 1997 The 1996 Federal Election: Report of an inquiry into the conduct of the 1996 Federal Election and matters related thereto 1997 Industrial Elections: Report of the inquiry into the role of the Australian Electoral Commission (AEC) in conducting industrial elections 2000 The 1998 Federal election: Inquiry into the conduct of the 1998 Federal election and matters related thereto 2001 User friendly, not abuser friendly: Inquiry into the integrity of the electoral roll 2002 Audit Report No.42 of 2001-2002, Integrity of the Electoral Roll 2003 2001 Federal Election: Report of the inquiry into the 2001 Federal Election and Matters Related Thereto 2003 Inquiry into representation of the Territories in the House of Representatives 2005 The 2004 Federal Election: Report of the Inquiry into the Conduct of the 2004 Federal Election and Matters Related Thereto 2006 Funding and Disclosure: Inquiry into disclosure of donations to political parties and candidates 2007 Civics and Electoral Education 2007 Review of certain aspects of the administration of the Australian Electoral Commission 2008 Advisory report on Schedule 1 of the Tax laws Amendment (2008 Measures No. 1) Bill 2008 2008 Advisory report on the Commonwealth Electoral Amendment (Political Donations and Other Measures) Bill 2008 *JSCER Reports # First JSCEM Report

242

243

Appendix C Gallagher’s Least Squares Index –– Calculations

Gallagher’s Least Squares Index is determined by calculating the difference between the percentage of the vote, and the percentage of the seats won for each party. These values are then squared and added. The total is then halved, from which the square root is calculated, giving the Index value. The mathematical equation for the Index is:

n 1 − 2 ∑ (V i S i ) 2 i = 1

Where possible, election data is disaggregated to show all parties receiving at least one per cent of the vote. Parties receiving less than one per cent, and all Independents, are collectively shown as Independents & Others.

These figures may differ slightly from Michael Gallagher’s own figures for these elections, due to the treatment of minor parties. However differences should be minimal, and will not impact on the comparative usefulness of the values.

Data is collected from the various electoral commissions’ election publications and websites, and the Australian Government and Politics Database.

244

245

Appendix D Gallagher’s Least Squares Index –– Federal Elections

House of Representatives

2007 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 43.38 83 55.33 11.9533 142.88 Liberal - CLP 36.60 55 36.67 0.0667 0.00 The Nationals 5.49 10 6.67 1.1767 1.38 The Greens 7.79 0 0.00 -7.7900 60.68 Family First 1.99 0 0.00 -1.9900 3.96 Independents & Others 4.75 2 1.33 -3.4167 11.67 100.00 150 220.59 halve 110.29 sq. root 10.50 2004 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.64 60 40.00 2.3600 5.57 Liberal/Country Liberal Party 40.81 75 50.00 9.1900 84.46 The Nationals 5.89 12 8.00 2.1100 4.45 The Greens 7.19 0 0.00 -7.1900 51.70 Family First 2.01 0 0.00 -2.0100 4.04 Democrats 1.24 0 0.00 -1.2400 1.54 Independents & Others 5.22 3 2.00 -3.2200 10.37 100.00 150 162.12 halve 81.06 sq. root 9.00 2001 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.84 65 43.33 5.4933 30.18 Liberal/Country Liberal Party 37.40 69 46.00 8.6000 73.96 The Nationals 5.61 13 8.67 3.0567 9.34 The Greens 4.96 0 0.00 -4.9600 24.60 Democrats 5.41 0 0.00 -5.4100 29.27 One Nation 4.34 0 0.00 -4.3400 18.84 Independents & Others 4.44 3 2.00 -2.4400 5.95 100.00 150 192.14 halve 96.07 sq. root 9.80 1998 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 40.10 67 45.27 5.1703 26.73 Liberal/Country Liberal Party 34.21 64 43.24 9.0332 81.60 The Nationals 5.29 16 10.81 5.5208 30.48 The Greens 2.14 0 0.00 -2.1400 4.58 Democrats 5.13 0 0.00 -5.1300 26.32 One Nation 8.43 0 0.00 -8.4300 71.06 Independents & Others 4.70 1 0.68 -4.0243 16.20 100.00 148 256.97 halve 128.48 sq. root 11.34

246

1996 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 38.75 49 33.11 -5.6419 31.83 Liberal/Country Liberal Party 39.04 76 51.35 12.3114 151.57 The Nationals 8.21 18 12.16 3.9522 15.62 The Greens 1.74 0 0.00 -1.7400 3.03 Democrats 6.76 0 0.00 -6.7600 45.70 Independents & Others 5.50 5 3.38 -2.1216 4.50 100.00 148 252.25 halve 126.12 sq. root 11.23 1993 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 44.92 80 54.42 9.5018 90.28 Liberal/Country Liberal Party 37.10 49 33.33 -3.7667 14.19 The Nationals 7.17 16 10.88 3.7144 13.80 Democrats 3.75 0 0.00 -3.7500 14.06 Independents & Others 7.06 2 1.36 -5.6995 32.48 100.00 147 164.81 halve 82.41 sq. root 9.08 1990 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 39.44 78 52.70 13.2627 175.90 Liberal/Country Liberal Party 35.04 55 37.16 2.1222 4.50 The Nationals 8.42 14 9.46 1.0395 1.08 Democrats 11.26 0 0.00 -11.2600 126.79 Independents & Others 5.84 1 0.68 -5.1643 26.67 100.00 148 334.94 halve 167.47 sq. root 12.94 1987 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 45.83 86 58.11 12.2781 150.75 Liberal/Country Liberal Party 34.55 43 29.05 -5.4959 30.21 The Nationals 11.52 19 12.84 1.3178 1.74 Democrats 6.03 0 0.00 -6.0300 36.36 Independents & Others 2.07 0 0.00 -2.0700 4.28 100.00 148 223.34 halve 111.67 sq. root 10.57 1984 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 47.55 82 55.41 7.8554 61.71 Liberal/Country Liberal Party 34.38 45 30.41 -3.9746 15.80 The Nationals 10.63 21 14.19 3.5592 12.67 Democrats 5.45 0 0.00 -5.4500 29.70 Independents & Others 1.99 0 0.00 -1.9900 3.96 100.00 148 123.84 halve 61.92 sq. root 7.87

247

1983 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 49.48 75 60.00 10.5200 110.67 Liberal/Country Liberal Party 34.36 33 26.40 -7.9600 63.36 The Nationals 9.21 17 13.60 4.3900 19.27 Democrats 5.03 0 0.00 -5.0300 25.30 Independents & Others 1.92 0 0.00 -1.9200 3.69 100.00 125 222.29 halve 111.15 sq. root 10.54 1980 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 39.44 78 52.70 13.2627 175.90 Liberal/Country Liberal Party 35.04 55 37.16 2.1222 4.50 The Nationals 8.42 14 9.46 1.0395 1.08 Democrats 11.26 0 0.00 -11.2600 126.79 Independents & Others 5.84 1 0.68 -5.1643 26.67 100.00 148 334.94 halve 167.47 sq. root 12.94 1977 House of Representatives % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 39.65 38 30.65 -9.0048 81.09 Liberal/Country Liberal Party 38.09 67 54.03 15.9423 254.16 The Nationals 10.01 19 15.32 5.3126 28.22 Democrats 9.38 0 0.00 -9.3800 87.98 Independents & Others 2.87 0 0.00 -2.8700 8.24 100.00 124 459.69 halve 229.84 sq. root 15.16 1975 House of Representatives No. % of %seats-

Party % seats seats %votes sqd Australian Labor Party 42.84 36 28.35 -14.4935 210.06 Liberal/Country Liberal Party 41.80 68 53.54 11.7433 137.91 National Country Party 11.25 23 18.11 6.8602 47.06 Democratic Labor Party 1.32 0 0.00 -1.3200 1.74 Independents & Others 2.79 0 0.00 -2.7900 7.78 100.00 127 404.56 halve 202.28 sq. root 14.22 1974 House of Representatives No. % of %seats-

Party % seats seats %votes sqd Australian Labor Party 49.30 66 51.97 2.6685 7.12 Liberal/Country Liberal Party 34.95 40 31.50 -3.4539 11.93 Country Party 9.96 21 16.54 6.5754 43.24 Democratic Labor Party 1.42 0 0.00 -1.4200 2.02 Australia Party 2.33 0 0.00 -2.3300 5.43 Independents & Others 2.04 0 0.00 -2.0400 4.16 100.00 127 73.89 halve 36.95 sq. root 6.08

248

1972 House of Representatives No. % of %seats-

Party % seats seats %votes sqd Australian Labor Party 49.59 67 53.60 4.0100 16.08 Liberal/Country Liberal Party 32.04 38 30.40 -1.6400 2.69 Country Party 9.44 20 16.00 6.5600 43.03 Democratic Labor Party 5.25 0 0.00 -5.2500 27.56 Australia Party 2.42 0 0.00 -2.4200 5.86 Independents & Others 1.26 0 0.00 -1.2600 1.59 100.00 125 96.81 halve 48.40 sq. root 6.96

Senate

2007 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 40.30 18 45.00 4.7000 22.09 Coalition 39.94 18 45.00 5.0600 25.60 Democrats 1.29 0 0.00 -1.2900 1.66 The Greens 9.04 3 7.50 -1.5400 2.37 Family First 1.62 0 0.00 -1.6200 2.62 Pauline's United Australia Party 1.12 0 0.00 -1.1200 1.25 Independents & Others 6.69 1 2.50 -4.1900 17.56 100.00 40 73.16 halve 36.58 sq. root 6.05 2004 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 35.02 16 40.00 4.9800 24.80 Coalition 45.09 21 52.50 7.4100 54.91 Democrats 2.09 0 0.00 -2.0900 4.37 The Greens 7.67 2 5.00 -2.6700 7.13 Family First 1.76 1 2.50 0.7400 0.55 One Nation 1.73 0 0.00 -1.7300 2.99 Independents & Others 6.64 0 0.00 -6.6400 44.09 100.00 40 138.84 halve 69.42 sq. root 8.33 2001 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 34.32 14 35.00 0.6800 0.46 Coalition 41.84 20 50.00 8.1600 66.59 Democrats 7.25 4 10.00 2.7500 7.56 The Greens 4.94 2 5.00 0.0600 0.00 One Nation 5.54 0 0.00 -5.5400 30.69 Independents & Others 6.11 0 0.00 -6.1100 37.33 100.00 40 142.64 halve 71.32 sq. root 8.45

249

1998 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 37.31 17 42.50 5.1900 26.94 Coalition 37.68 17 42.50 4.8200 23.23 Democrats 8.45 4 10.00 1.5500 2.40 The Greens 2.72 0 0.00 -2.7200 7.40 One Nation 8.99 1 2.50 -6.4900 42.12 Christian Democratic Party 1.09 0 0.00 -1.0900 1.19 Independents & Others 3.76 1 2.50 -1.2600 1.59 100.00 40 104.87 halve 52.43 sq. root 7.24

1996 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 36.15 14 35.00 -1.1500 1.32 Coalition 43.97 20 50.00 6.0300 36.36 Democrats 10.82 5 12.50 1.6800 2.82 The Greens 1.66 1 2.50 0.8400 0.71 Call to Australia 1.08 0 0.00 -1.0800 1.17 Australians Against Further Immigration 1.26 0 0.00 -1.2600 1.59 Australian Shooters Party 1.05 0 0.00 -1.0500 1.10 Independents & Others 4.01 0 0.00 -4.0100 16.08 100.00 40 61.15 halve 30.57 sq. root 5.53

1993 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 43.50 17 42.50 -1.0000 1.00 Coalition 43.04 19 47.50 4.4600 19.89 Democrats 5.31 2 5.00 -0.3100 0.10 The Greens 2.96 1 2.50 -0.4600 0.21 Call to Australia 1.02 0 0.00 -1.0200 1.04 Independents & Others 4.17 1 2.50 -1.6700 2.79 100.00 40 25.03 halve 12.51 sq. root 3.54

1990 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 38.41 15 37.50 -0.9100 0.83 Coalition 41.92 19 47.50 5.5800 31.14 Democrats 12.63 5 12.50 -0.1300 0.02 The Greens 2.80 1 2.50 -0.3000 0.09 Independents & Others 4.25 0 0.00 -4.2500 18.06 100.00 40 50.13 halve 25.07 sq. root 5.01

250

1987 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 42.83 32 42.11 -0.7247 0.53 Coalition 42.03 34 44.74 2.7068 7.33 Democrats 8.47 7 9.21 0.7405 0.55 Nuclear Disarmament Party 1.09 1 1.32 0.2258 0.05 Independents & Others 5.58 2 2.63 -2.9484 8.69 100.00 76 17.14 halve 8.57 sq. root 2.93 1984 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 42.17 20 43.48 1.3083 1.71 Coalition 39.55 20 43.48 3.9283 15.43 Democrats 7.62 5 10.87 3.2496 10.56 Nuclear Disarmament Party 7.23 1 2.17 -5.0561 25.56 Independents & Others 3.44 0 0.00 -3.4400 11.83 100.00 46 65.10 halve 32.55 sq. root 5.71 1983 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 45.49 30 46.88 1.3850 1.92 Coalition 39.96 28 43.75 3.7900 14.36 Democrats 9.57 5 7.81 -1.7575 3.09 Independents & Others 4.98 1 1.56 -3.4175 11.68 100.00 64 31.05 halve 15.53 sq. root 3.94 1980 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 42.25 15 44.12 1.8676 3.49 Coalition 43.48 15 44.12 0.6376 0.41 Democrats 9.25 3 8.82 -0.4265 0.18 Independents & Others 5.02 1 2.94 -2.0788 4.32 100.00 34 8.40 halve 4.20 sq. root 2.05 1977 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 36.76 14 41.18 4.4165 19.51 Coalition 45.57 18 52.94 7.3712 54.33 Democrats 11.13 2 5.88 -5.2476 27.54 Independents & Others 6.54 0 0.00 -6.5400 42.77 100.00 34 144.15 halve 72.07 sq. root 8.49

251

1975 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 40.91 27 42.19 1.2775 1.63 Coalition 51.70 35 54.69 2.9875 8.93 Democratic Labor Party 2.67 0 0.00 -2.6700 7.13 Liberal Movement 1.07 1 1.56 0.4925 0.24 Independents & Others 3.65 1 1.56 -2.0875 4.36 100.00 64 22.29 halve 11.14 sq. root 3.34 1974 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 47.29 29 48.33 1.0433 1.09 Coalition 43.89 29 48.33 4.4433 19.74 Democratic Labor Party 3.56 0 0.00 -3.5600 12.67 Australia Party 1.39 0 0.00 -1.3900 1.93 Independents & Others 3.87 2 3.33 -0.5367 0.29 100.00 60 35.73 halve 17.86 sq. root 4.23 1970 Senate % No. % of %seats- Party votes seats seats %votes sqd Australian Labor Party 42.22 14 43.75 1.5300 2.34 Coalition 38.18 13 40.63 2.4450 5.98 Democratic Labor Party 11.11 3 9.38 -1.7350 3.01 Australia Party 2.90 0 0.00 -2.9000 8.41 Independents & Others 5.59 2 6.25 0.6600 0.44 100.00 32 20.17 halve 10.09 sq. root 3.18

252

253

Appendix E Gallagher’s Least Squares Index –– NSW Legislative Council

1984 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 46.88 7 46.67 -0.2133 0.05 Coalition 42.61 7 46.67 4.0567 16.46 Call to Australia 6.09 1 6.67 0.5767 0.33 Democrats 3.15 0 0.00 -3.1500 9.92 Independents & Others 1.27 0 0.00 -1.2700 1.61 100.00 15 28.37 halve 14.19 sq. root 3.77 1988 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.51 6 40.00 2.4900 6.20 Coalition 46.15 7 46.67 0.5167 0.27 Democrats 2.70 1 6.67 3.9667 15.73 Call to Australia 5.71 1 6.67 0.9567 0.92 Independent Enterprise,

Freedom and Family 2.42 0 0.00 -2.4200 5.86 Environment Group 1.60 0 0.00 -1.6000 2.56 Independents & Others 3.91 0 0.00 -3.9100 15.29 100.00 15 46.82 halve 23.41 sq. root 4.84 1991 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.29 6 40.00 2.7100 7.34 Coalition 45.34 7 46.67 1.3267 1.76 Greens 3.32 0 0.00 -3.3200 11.02 Democrats 6.70 1 6.67 -0.0333 0.00 Call to Australia 3.58 1 6.67 3.0867 9.53 Independents & Others 3.77 0 0.00 -3.7700 14.21 100.00 15 43.87 halve 21.93 sq. root 4.68 1995 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 35.25 8 38.10 2.8452 8.10 Coalition 38.49 8 38.10 -0.3948 0.16 Greens 3.72 1 4.76 1.0419 1.09 Democrats 3.21 1 4.76 1.5519 2.41 Call to Australia 3.01 1 4.76 1.7519 3.07 A Better Future For Our

Children 1.28 1 4.76 3.4819 12.12 Shooters 2.84 1 4.76 1.9219 3.69 Independents & Others 12.20 0 0.00 -12.2000 148.84 100.00 21 179.47 halve 89.74 sq. root 9.47

254

1999 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.27 8 38.10 0.8252 0.68 Coalition 27.39 6 17.14 -10.2471 105.00 One Nation 6.34 1 4.76 -1.5781 2.49 Democrats 4.01 1 4.76 0.7519 0.57 Christian Democratic Party 3.17 1 4.76 1.5919 2.53 Greens 2.91 1 4.76 1.8519 3.43 Shooters 1.67 0 0.00 -1.6700 2.79 Progressive Labor Party 1.58 0 0.00 -1.5800 2.50 Marijuana Smokers Rights 1.24 0 0.00 -1.2400 1.54 Reform Legal System 1.00 1 4.76 3.7619 14.15 Independents & Others 13.42 2 9.52 -3.8962 15.18 100.00 21 150.86 halve 75.43 sq. root 8.69 2003 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 43.54 10 47.62 4.0790 16.64 Coalition 33.30 7 33.33 0.0333 0.00 Greens 8.60 2 9.52 0.9238 0.85 Christian Democratic Party 3.03 1 4.76 1.7319 3.00 One Nation 1.49 0 0.00 -1.4900 2.22 Democrats 1.57 0 0.00 -1.5700 2.46 Shooters 2.05 1 4.76 2.7119 7.35 Hanson 1.92 0 0.00 -1.9200 3.69 Unity 1.42 0 0.00 -1.4200 2.02 Fishing-Horse-4WD 1.06 0 0.00 -1.0600 1.12 Independents & Others 2.02 0 0.00 -2.0200 4.08 100.00 21 43.44 halve 21.72 sq. root 4.66 2007 NSW Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 39.14 9 42.86 3.7171 13.82 Coalition 34.22 8 38.10 3.8752 15.02 Greens 9.12 2 9.52 0.4038 0.16 Christian Democratic Party 4.42 1 4.76 0.3419 0.12 Democrats 1.78 0 0.00 -1.7800 3.17 Unity 1.21 0 0.00 -1.2100 1.46 Shooters 2.79 1 4.76 1.9719 3.89 Fishing 1.53 0 0.00 -1.5300 2.34 Australians Against Further

Immigration 1.64 0 0.00 -1.6400 2.69 Independents & Others 4.15 0 0.00 -4.1500 17.22 100.00 21 59.89 halve 29.94 sq. root 5.47

255

Appendix F Gallagher’s Least Squares Index –– Victorian Legislative Council

2006 Victorian Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 41.45 19 47.50 6.0500 36.60 Liberal 34.55 15 37.50 2.9500 8.70 Nationals 4.43 2 5.00 0.5700 0.32 Greens 10.58 3 7.50 -3.0800 9.49 Democratic Labor Party 1.97 1 2.50 0.5300 0.28 Family First 3.85 0 0.00 -3.8500 14.82 People Power 1.02 0 0.00 -1.0200 1.04 Independents & Others 2.15 0 0.00 -2.1500 4.62 100.00 40 75.88 halve 37.94 sq. root 6.16 2002 Victorian Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 47.49 17 77.27 29.7827 887.01 Liberal 34.51 3 13.64 -20.8736 435.71 Nationals 4.37 2 9.09 4.7209 22.29 Greens 10.87 0 0.00 -10.8700 118.16 Democrats 1.79 0 0.00 -1.7900 3.20 Independents & Others 0.97 0 0.00 -0.9700 0.94 100.00 22 1467.31 halve 733.65 sq. root 27.09 1999 Victorian Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 42.19 8 36.36 -5.8264 33.95 Liberal 39.75 11 50.00 10.2500 105.06 Nationals 7.25 3 13.64 6.3864 40.79 Greens 2.23 0 0.00 -2.2300 4.97 Democrats 6.81 0 0.00 -6.8100 46.38 Independents & Others 1.77 0 0.00 -1.7700 3.13 100.00 22 234.28 halve 117.14 sq. root 10.82

256

257

Appendix G Gallagher’s Least Squares Index –– Queensland Legislative Assembly 1983 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 43.98 32 39.02 -4.9556 24.56 National 38.93 41 50.00 11.0700 122.54 Liberal 14.88 8 9.76 -5.1239 26.25 Independents & Others 2.21 1 1.22 -0.9905 0.98 100.00 82 174.34 halve 87.17 sq. root 9.34 1986 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 41.35 30 33.71 -7.6421 58.40 National 39.64 49 55.06 15.4162 237.66 Liberal 16.50 10 11.24 -5.2640 27.71 Independents & Others 2.51 0 0.00 -2.5100 6.30 100.00 89 330.07 halve 165.04 sq. root 12.85 1989 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 50.32 54 60.67 10.3542 107.21 National 24.09 27 30.34 6.2471 39.03 Liberal 21.05 8 8.99 -12.0612 145.47 Independents & Others 4.54 0 0.00 -4.5400 20.61 100.00 89 312.32 halve 156.16 sq. root 12.50 1992 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 48.73 54 60.67 11.9442 142.66 National 23.71 26 29.21 5.5035 30.29 Liberal 20.44 9 10.11 -10.3276 106.66 Independents & Others 7.12 0 0.00 -7.1200 50.69 100.00 89 330.31 halve 165.15 sq. root 12.85 1995 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 42.89 45 50.56 7.6718 58.86 National 26.25 29 32.58 6.3343 40.12 Liberal 22.74 14 15.73 -7.0097 49.14 Greens 2.87 0 0.00 -2.8700 8.24 Independents & Others 5.25 1 1.12 -4.1264 17.03 100.00 89 173.38 halve 86.69 sq. root 9.31

258

1998 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 38.86 44 49.44 10.5782 111.90 National 15.17 23 25.84 10.6727 113.91 Liberal 16.09 9 10.11 -5.9776 35.73 Greens 2.36 0 0.00 -2.3600 5.57 One Nation 22.68 11 12.36 -10.3204 106.51 Independents & Others 4.84 2 2.25 -2.5928 6.72 100.00 89 380.34 halve 190.17 sq. root 13.79 2001 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 48.93 66 74.16 25.2273 636.42 National 14.16 12 13.48 -0.6769 0.46 Liberal 14.32 3 3.37 -10.9492 119.89 Greens 2.51 0 0.00 -2.5100 6.30 One Nation 8.69 3 3.37 -5.3192 28.29 City Country Alliance 2.39 0 0.00 -2.3900 5.71 Independents & Others 9.00 5 5.62 -3.3820 11.44 100.00 89 808.50 halve 404.25 sq. root 20.11 2004 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 47.01 63 70.79 23.7765 565.32 National 16.96 15 16.85 -0.1061 0.01 Liberal 18.50 5 5.62 -12.8820 165.95 Greens 6.76 0 0.00 -6.7600 45.70 One Nation 4.88 1 1.12 -3.7564 14.11 Independents & Others 5.89 5 5.62 -0.2720 0.07 100.00 89 791.16 halve 395.58 sq. root 19.89 2006 Queensland Legislative Assembly % of %seats-

Party % votes No. seats seats %votes sqd Australian Labor Party 46.92 59 66.29 19.3721 375.28 National 17.82 17 19.10 1.2811 1.64 Liberal 20.10 8 8.99 -11.1112 123.46 Greens 7.99 0 0.00 -7.9900 63.84 One Nation 0.60 1 1.12 0.5236 0.27 Family First 1.89 0 0.00 -1.8900 3.57 Independents & Others 4.68 4 4.49 -0.1856 0.03 100.00 89 568.10 halve 284.05 sq. root 16.85

259

Appendix H Gallagher’s Least Squares Index –– Western Australian Parliament

2008 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 35.84 28 47.46 11.6176 134.97 Liberal 38.39 24 40.68 2.2880 5.23 National 4.87 4 6.78 1.9097 3.65 Greens 11.92 0 0.00 -11.9200 142.09 Christian Democratic Party 2.58 0 0.00 -2.5800 6.66 Family First 1.95 0 0.00 -1.9500 3.80 Independents & Others 4.45 3 5.08 0.6347 0.40 100.00 59 296.80 halve 148.40 sq. root 12.18 2008 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 36.13 11 30.56 -5.5744 31.07 Liberal 39.61 16 44.44 4.8344 23.37 National 5.32 5 13.89 8.5689 73.43 Greens 11.08 4 11.11 0.0311 0.00 Christian Democratic Party 2.32 0 0.00 -2.3200 5.38 Family First 2.52 0 0.00 -2.5200 6.35 Independents & Others 3.02 0 0.00 -3.0200 9.12 100.00 36 148.73 halve 74.36 sq. root 8.62 2005 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 41.88 32 56.14 14.2604 203.36 Liberal 35.64 18 31.58 -4.0611 16.49 National 3.69 5 8.77 5.0819 25.83 Greens 7.57 0 0.00 -7.5700 57.30 Christian Democratic Party 2.93 0 0.00 -2.9300 8.58 Family First 2.02 0 0.00 -2.0200 4.08 One Nation 1.64 0 0.00 -1.6400 2.69 Independents & Others 4.63 2 3.51 -1.1212 1.26 100.00 57 319.59 halve 159.80 sq. root 12.64 2005 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 43.35 16 47.06 3.7088 13.76 Liberal 37.12 15 44.12 6.9976 48.97 National 2.19 1 2.94 0.7512 0.56 Greens 7.52 2 5.88 -1.6376 2.68 Christian Democratic Party 2.28 0 0.00 -2.2800 5.20 Family First 2.01 0 0.00 -2.0100 4.04 One Nation 1.59 0 0.00 -1.5900 2.53 Independents & Others 3.94 0 0.00 -3.9400 15.52 100.00 34 93.26 halve 46.63 sq. root 6.83

260

2001 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.24 32 56.14 18.9004 357.22 Liberal 31.16 16 28.07 -3.0898 9.55 National 3.26 5 8.77 5.5119 30.38 Greens 7.27 0 0.00 -7.2700 52.85 Democrats 2.64 0 0.00 -2.6400 6.97 One Nation 9.58 0 0.00 -9.5800 91.78 Independents & Others 8.85 4 7.02 -1.8325 3.36 100.00 57 552.11 halve 276.05 sq. root 16.61

2001 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.94 13 38.24 0.2953 0.09 Liberal 33.96 12 35.29 1.3341 1.78 National 2.40 1 2.94 0.5412 0.29 Greens 8.00 5 14.71 6.7059 44.97 Democrats 3.72 0 0.00 -3.7200 13.84 Christian Democratic Party 1.54 0 0.00 -1.5400 2.37 One Nation 9.88 3 8.82 -1.0565 1.12 Independents & Others 2.56 0 0.00 -2.5600 6.55 100.00 34 71.01 halve 35.50 sq. root 5.96

1996 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 35.82 19 33.33 -2.4867 6.18 Liberal 39.90 29 50.88 10.9772 120.50 National 5.79 6 10.53 4.7363 22.43 Greens 4.73 0 0.00 -4.7300 22.37 Democrats 5.08 0 0.00 -5.0800 25.81 Independents & Others 8.68 3 5.26 -3.4168 11.67 100.00 57 208.97 halve 104.48 sq. root 10.22 1996 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 33.08 12 35.29 2.2141 4.90 Coalition 45.94 17 50.00 4.0600 16.48 Greens 5.55 3 8.82 3.2735 10.72 Democrats 6.58 2 5.88 -0.6976 0.49 Australian Marijuana Party 2.49 0 0.00 -2.4900 6.20 Independents & Others 6.36 0 0.00 -6.3600 40.45 100.00 34 79.24 halve 39.62 sq. root 6.29

261

1993 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 37.08 24 42.11 5.0253 25.25 Liberal 44.15 26 45.61 1.4640 2.14 National 5.31 6 10.53 5.2163 27.21 Greens 4.31 0 0.00 -4.3100 18.58 Democrats 2.32 0 0.00 -2.3200 5.38 Independents & Others 6.83 1 1.75 -5.0756 25.76 100.00 57 104.33 halve 52.16 sq. root 7.22

1993 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 36.82 14 41.18 4.3565 18.98 Liberal 45.60 15 44.12 -1.4824 2.20 National 3.99 3 8.82 4.8335 23.36 Greens 5.16 1 2.94 -2.2188 4.92 Democrats 3.01 0 0.00 -3.0100 9.06 Independents & Others 5.42 1 2.94 -2.4788 6.14 100.00 34 64.67 halve 32.33 sq. root 5.69

1989 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 42.46 31 54.39 11.9260 142.23 Liberal 42.79 20 35.09 -7.7023 59.33 National 4.60 6 10.53 5.9263 35.12 Democrats 1.43 0 0.00 -1.4300 2.04 Greypower 5.16 0 0.00 -5.1600 26.63 Independents & Others 3.56 0 0.00 -3.5600 12.67 100.00 57 278.02 halve 139.01 sq. root 11.79

1989 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 41.33 16 47.06 5.7288 32.82 Liberal 41.05 15 44.12 3.0676 9.41 National 5.00 3 8.82 3.8235 14.62 Greens 1.91 0 0.00 -1.9100 3.65 Democrats 3.32 0 0.00 -3.3200 11.02 Greypower 4.02 0 0.00 -4.0200 16.16 Independents & Others 3.37 0 0.00 -3.3700 11.36 100.00 34 99.04 halve 49.52 sq. root 7.04

262

1986 WA Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 53.00 32 56.14 3.1404 9.86 Liberal 41.32 19 33.33 -7.9867 63.79 National 3.71 6 10.53 6.8163 46.46 Independents & Others 1.97 0 0.00 -1.9700 3.88 100.00 57 123.99 halve 62.00 sq. root 7.87 1986 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 44.60 9 52.94 8.3412 69.58 Liberal 41.90 6 35.29 -6.6059 43.64 National 3.30 2 11.76 8.4647 71.65 Democrats 8.10 0 0.00 -8.1000 65.61 Independents & Others 2.10 0 0.00 -2.1000 4.41 100.00 17 254.88 halve 127.44 sq. root 11.29 1983 WA Legislative Assembly* % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 53.16 32 57.14 3.9829 15.86 Liberal 39.86 20 35.71 -4.1457 17.19 National 1.67 2 3.57 1.9014 3.62 National Country Party 3.44 2 3.57 0.1314 0.02 Independents & Others 1.87 0 0.00 -1.8700 3.50 100.00 56 40.18 halve 20.09 sq. root 4.48 *1 seat, Narrogin, was uncontested and won by the National Country Party. It is

not included in the totals.

1983 WA Legislative Council % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 50.63 7 41.18 -9.4535 89.37 Liberal 41.57 9 52.94 11.3712 129.30 National 3.15 1 5.88 2.7324 7.47 National Country Party 3.18 0 0.00 -3.1800 10.11 Independents & Others 1.47 0 0.00 -1.4700 2.16 100.00 17 238.41 halve 119.21 sq. root 10.92

263

Appendix I Gallagher’s Least Squares Index –– Tasmanian Legislative Assembly 1982 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 36.86 14 40.00 3.1400 9.86 Liberal 48.52 19 54.29 5.7657 33.24 Democrats 5.39 1 2.86 -2.5329 6.42 Independents & Others 9.23 1 2.86 -6.3729 40.61 100.00 35 90.13 halve 45.07 sq. root 6.71 1986 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 35.14 14 40.00 4.8600 23.62 Liberal 54.20 19 54.29 0.0857 0.01 Democrats 2.06 0 0.00 -2.0600 4.24 Independents & Others 8.60 2 5.71 -2.8857 8.33 100.00 35 36.20 halve 18.10 sq. root 4.25 1989 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 34.71 13 37.14 2.4329 5.92 Liberal 46.92 17 48.57 1.6514 2.73 Tasmanian Greens 17.13 5 14.29 -2.8443 8.09 Independents & Others 1.24 0 0.00 -1.2400 1.54 100.00 35 18.27 halve 9.14 sq. root 3.02 1992 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 28.85 11 31.43 2.5786 6.65 Liberal 54.11 19 54.29 0.1757 0.03 Tasmanian Greens 13.23 5 14.29 1.0557 1.11 Advance Tasmania 2.55 0 0.00 -2.5500 6.50 Independents & Others 1.26 0 0.00 -1.2600 1.59 100.00 35 -100 15.88 halve 7.94 sq. root 2.82 1996 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 40.47 14 40.00 -0.4700 0.22 Liberal 41.20 16 45.71 4.5143 20.38 Tasmanian Greens 11.14 4 11.43 0.2886 0.08 Nationals 2.20 0 0.00 -2.2000 4.84 Independents & Others 4.99 1 2.86 -2.1329 4.55 100.00 35 30.07 halve 15.04 sq. root 3.88

264

1998 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 44.79 14 56.00 11.2100 125.66 Liberal 38.06 10 40.00 1.9400 3.76 Tasmanian Greens 10.18 1 4.00 -6.1800 38.19 Tasmanian First 5.10 0 0.00 -5.1000 26.01 Independents & Others 1.87 0 0.00 -1.8700 3.50 100.00 25 197.13 halve 98.56 sq. root 9.93 2002 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 51.88 14 56.00 4.1200 16.97 Liberal 27.38 7 28.00 0.6200 0.38 Tasmanian Greens 18.13 4 16.00 -2.1300 4.54 Independents & Others 2.61 0 0.00 -2.6100 6.81 100.00 25 28.71 halve 14.35 sq. root 3.79 2006 Tasmanian Legislative Assembly % No. % of %seats-

Party votes seats seats %votes sqd Australian Labor Party 49.26 14 56.00 6.7400 45.43 Liberal 31.81 7 28.00 -3.8100 14.52 Tasmanian Greens 16.63 4 16.00 -0.6300 0.40 Independents & Others 2.30 0 0.00 -2.3000 5.29 100.00 25 65.63 halve 32.82 sq. root 5.73

265

Appendix J Northern Territory Turnout and Informality –– 2005 General Election

Electorate Indigenous Turnout Informality Number of Population % Rate % Rate % Candidates Port Darwin 4.7 78.2 4.3 2 Wanguri 6.2 87.7 3.3 2 Nelson 6.3 83.4 2.3 3 Nightcliff 6.3 84.2 2.5 5 Fannie Bay 7.4 85.0 4.7 3 Casuarina 8.6 86.6 2.6 4 Sanderson 8.9 89.4 3.8 4 Goyder 9.5 88.8 3.3 5 Johnston 9.7 86.9 3.1 5 Drysdale 10.4 82.1 4.7 2 Brennan 11.3 84.4 3.2 3 Greatorex 12.4 86.3 2.5 3 Karama 13.3 89.1 5.1 2 Blain 13.4 86.0 3.5 4 Millner 14.0 83.1 2.8 5 Braitling 15.9 84.0 2.7 3 Araluen 16.5 82.1 3.0 3 Katherine 24.0 81.8 3.5 2 Daly 45.4 77.6 5.0 4 Nhulunbuy 51.7 65.1 4.1 2 Barkly 60.3 69.4 3.8 3 MacDonnell 67.7 67.7 7.8 5 Arafura 77.2 66.2 4.9 3 Arnhem 82.8 64.2 6.4 3 Stuart 84.3 59.3 4.6 2 Source: NTEC.