Establishing the Rules of the Game: Election Laws in Democracies 9781442674639

Table of contents :
Contents
List of Tables
Preface
Introduction: Establishing the Rules
1. Who Has the Right to Vote?
2. Who Has the Right to Be a Candidate?
3. Registering Voters
4. Who Conducts the Election?
5. How People Vote
6. Sorting Out Winners and Losers
Conclusion: Patterns and Future Directions
Appendix: Countries, Sources, and Variables Included in the Study
Notes
References
Index

Citation preview

ESTABLISHING THE RULES OF THE GAME: ELECTION LAWS IN DEMOCRACIES

Election laws vary greatly across contemporary democratic societies. In Establishing the Rules of the Game, Louis Massicotte, Andre Blais, and Antoine Yoshinaka provide the first thorough examination of these laws. The study incorporates original data collected from more than sixty democracies around the world. The authors focus on six dimensions of election laws: the right to vote, the right to be a candidate, the electoral register, the agency in charge of the election, the procedure for casting votes, and the procedure to sort out the winners and losers. Massicotte, Blais, and Yoshinaka uncover underlying patterns in election laws, explaining why certain types of countries tend to adopt particular sets of rules. For instance, the study demonstrates that former colonies generally adopt laws that are similar to those of their former mother country, and that established democracies tend to be more inclusive than non-established ones. The authors point out demographic patterns and review normative and practical arguments for and against each set of rules, providing invaluable insights into the electoral process and democratic theory. L O U I S MASSICOTTE is an associate professor in the Department of Political Science at 1'Universite de Montreal. A N D R E B L A I S is a professor in the Department of Political Science at 1'Universite de Montreal. A N T O I N E Y O S H I N A K A is a PhD candidate in the Department of Political Science at the University of Rochester.

This page intentionally left blank

Establishing the Rules of the Game Election Laws in Democracies

LOUIS MASSICOTTE ANDRE BLAIS ANTOINE YOSHINAKA

U N I V E R S I T Y OF TORONTO PRESS Toronto Buffalo London

www.utppublishing.com University of Toronto Press Incorporated 2004 Toronto Buffalo London Printed in Canada ISBN 0-8020-8769-8 (cloth) ISBN 0-8020-8564-4 (paper)

Printed on acid-free paper

National Library of Canada Cataloguing in Publication Massicotte, Louis Establishing the rule of the game : election laws in democracies / Louis Massicotte, Andre Blais, Antoine Yoshinaka. Includes bibliographical references and index. ISBN 0-8020-8769-8 (bound). ISBN 0-8020-8564-4 (pbk.) 1. Election law. I. Blais, Andre, 1947III. Title.

. II. Yoshinaka, Antoine

K3293.M38 2003 342'.07 C2003-902728-7

This book has been published with the help of a grant from the Canadian Federation for the Humanities and Social Sciences, through the Aid to Scholarly Publications Programme, using funds provided by the Social Sciences and Humanities Research Council of Canada. University of Toronto Press acknowledges the financial assistance to its publishing program of the Canada Council for the Arts and the Ontario Arts Council. University of Toronto Press acknowledges the financial support for its publishing activities of the Government of Canada through the Book Publishing Industry Development Program.

Contents

List of Tables vii Preface ix

Introduction: Establishing the Rules 3 1 Who Has the Right to Vote? 15 2 Who Has the Right to Be a Candidate?

40

3 Registering Voters 66 4 Who Conducts the Election? 83 5 How People Vote 102 6 Sorting Out Winners and Losers 142 Conclusion: Patterns and Future Directions 158 Appendix: Countries, Sources, and Variables Included in the Study Notes 167 References 177 Index

183

163

This page intentionally left blank

Tables

1.1 Restrictions on the right to vote 18 1.2 British colonialism, political rights, and restrictions on the right to vote 26 1.3 Countries with compulsory voting 36 1.4 British colonialism, political rights, and compulsory voting 38 2.1 Candidacy requirements in legislative elections 42 2.2 Candidacy requirements in presidential elections 50 2.3 British colonialism, political rights, and candidacy requirements (legislative elections) 54 3.1 Registration procedures in 63 democracies 68 3.2 British colonialism, political rights, and registration 75 4.1 Central electoral administration 84 4.2 British colonialism, political rights, and the presence of an electoral commission 96 5.1 Polling day 103 5.2 British colonialism, political rights, and polling day 119 5.3 Ballot papers in legislative elections 124 5.4 Ballot papers in presidential elections 126 5.5 British colonialism, political rights, and ballot papers in legislative elections 130 5.6 Special voting procedures 134 5.7 British colonialism, political rights, and special voting procedures 137 5.8 Political rights and the index of voting procedures 140 6.1 Counting the votes 144 6.2 British colonialism, political rights, and counting procedures 148 6.3 Recounts and petitions 152 6.4 British colonialism, political rights, and election petitions 156

This page intentionally left blank

Preface

The idea behind this book is quite simple. Many studies have examined electoral systems, their nature, causes, and consequences. There are also a good number of studies that analyse the regulation of money in election campaigns. But there is very little research, and no systematic analysis, of other basic dimensions of election laws, such as those that pertain to who has the right to vote and to be a candidate, and who conducts the election. There is a huge gap in the literature, and that gap ought to be filled. Thus this book. As with most research projects, the idea was simple but doing the required research was much more difficult than anticipated. Election laws are there to be studied, but we have discovered that they are not quite there all the time, at least not in the form we would like them to be. Election laws change and it is sometimes a challenge to find the most recent version. Some provisions of these laws are quite ambiguous. Some aspects of election legislation may be incorporated in other laws. And so on ... Despite all these difficulties, we have completed our journey. We had three objectives. The first was to provide an accurate description of the wide array of election laws in contemporary democracies. The second goal was to review the debate about the pros and cons of the various rules that can be adopted. The third was to uncover broad sociological patterns, that is, to determine what kinds of countries tend to adopt what kinds of rules. We have made every effort to cross-check the accuracy of the information presented in the text and in the tables about the content of election laws. We are painfully aware that there may still be some errors. All we can say is that we have tried very hard to verify the 'facts.' Readers must be reminded that the data describe the situation as we could ascertain it as of December 1999.

x

Preface

We also review the arguments that have been put forward in favour of, or in opposition to, certain types of rules. Our objective is not to convince readers that certain laws are 'better' than others, but rather to help them better understand the merits and limits of various options and to form a more informed opinion. Unfortunately, the review of the debate is sometimes brief, for the simple reason that the literature on these aspects of election laws is rather thin. It is our hope that researchers will use our presentation as a starting point for a deeper examination of the issues. Finally, we establish a number of patterns in the kinds of election laws that are adopted in contemporary democracies. We show, in particular, that former colonies tend to enact the same rules as their mother countries, and that there is a tendency for established democracies to be more inclusive than non-established ones. We have only scratched the surface here. We believe that our 'macroscopic' approach is useful because it provides a broad picture of the similarities and differences among groups of countries. But there is clearly a need for more focused comparative case studies in order to provide a better understanding of what makes governments decide to adopt certain types of rules. Again, our hope is that our study will trigger many more. This study would not have been possible without the support of many persons and institutions. We benefited from the generous financial support of the Social Sciences and Humanities Research Council of Canada. We acknowledge the competent and rigorous research assistance of Daniel Charron, Agnieszka Dobrzynska, and Jean-Francois Mayer. We thank Kees Aarts, Kenneth Benoit, Paul Harris, and Alain Pelletier for their help in clarifying specific provisions of election laws. We are grateful for the commitment of the University of Toronto Press and their excellent staff, especially Chris Bucci, Virgil Duff, Lauren Freeman, Douglas Hildebrand, and Anne Laughlin. We wish to thank the anonymous reviewers whose comments on our manuscript greatly improved the final product. Lastly, we thank our families for their indefectible support.

ESTABLISHING THE RULES OF THE GAME: ELECTION LAWS IN DEMOCRACIES

This page intentionally left blank

Introduction

Establishing the Rules

An election is a game in which parties or candidates attempt to convince electors to vote for them (Laponce 1969). The objective is for a candidate to garner enough support to get elected and, in a parliamentary system, for a party to have enough of its candidates elected to form the government or to be part of the ruling coalition. It is well known that winning in a game hinges to a large extent on the rules of the game and on how these rules are interpreted. Elections are no exception. It is impossible to make sense of the meaning of an election without taking into account who has (and who does not have) the right to vote, who has (and who does not have) the right to be a candidate, how the electoral register is established, who conducts the election, how people express their preferences, and how the winners and losers are sorted out. Each chapter of this book examines how a wide variety of democratic societies answers each of these questions. This is a study of election laws in democratic societies. Election laws define the rules of the game and constitute, as such, an essential ingredient of the electoral process. The analysis of election laws is a largely unexplored territory. This may be surprising, given that the study of voting, elections, and parties is one of the most advanced fields of political science. Previous comparable studies, such as the one by Mackenzie (1958), lack the diversity (it focused on Commonwealth countries) and have not been replicated in more recent times. There may be numerous reasons for this benign neglect. One reason is that this is a field where quantitative data are readily available; it attracts analysts with expertise in sophisticated statistical techniques more than researchers with a strong background in law or philosophy. This study attempts to redress some of the imbalance. We take election laws seriously, and we believe that they ought to be taken seriously.

4 Establishing the Rules of the Game

Election laws in general have been neglected, but not all election laws. There is a vast literature on electoral systems - that is, electoral formulas, district magnitude, and structure of ballots. It is one of the most developed subfields of the discipline. We have ourselves proposed conceptual and empirical surveys of types of electoral systems (Blais 1988; Blais and Massicotte 2002; Massicotte and Blais 1999). Seminal studies have looked at their political consequences (Cox 1997; Katz 1997; Lijphart 1994; Rae 1967; Taagepera and Shugart 1989). And, more recently, there has been innovative work on the factors leading to their adoption (Benoit and Schiemann 2001; Boix 1999; De-Mesquita 2000; Kaminski 2002). There have also been quite interesting analyses of the drawing of district boundaries (Butler and Cain 1992; Courtney 2001; McLean and Butler 1996). Electoral systems are undoubtedly a major component of election laws, but they are only one component. We believe that it is necessary to adopt a wider perspective and to achieve a broader understanding of the whole gamut of electoral rules and of their implications in a democratic society. This study is designed to fill a huge gap in our understanding of the rules under which democratic elections are conducted. In this introduction we look at why electoral laws matter, at the method of this study, and at the contents of this book. WHY ELECTORAL LAWS MATTER

The 1990s witnessed the largest expansion of democratic rights seen since the Second World War. Prior to the fall of the Berlin Wall, the universe of democratic countries that political scientists used to include in their comparative studies never exceeded thirty members, mostly located in western Europe and in parts of other continents settled by Europeans. Outside this fortunate circle, Western-style democracy was often perceived as fragile where it existed and an irrelevant dream where it did not - which was the case for most inhabitants of this planet. That democratic countries were among the most economically successful and technologically advanced hinted that they might be the wave of the future. But few observers anticipated the extent to which they came to be emulated by countries that seemed so little auspicious for democracy. The data compiled by Freedom House since 1972 illustrate the amplitude of the changes wrought by the fall of Communism (for a summary of the techniques used by this U.S.-based organization, see below under

Introduction: Establishing the Rules 5

'Method'). Most previously Communist countries have moved in the direction of free elections, as did some right-wing regimes previously supported by the West in the name of fighting Communism, as well as regimes that played on the East-West divide in order to secure tolerance or even support for their authoritarian rule. Looking at the number of countries rated 1 or 2 by that organization on a scale ranging from 1 to 7, we find 55 in 1982, 78 in 1992 and 88 in 2001. There is nothing irreversible in these changes. Witness the persistence of authoritarian regimes as well as recent turnarounds in that direction. The democratic waves of 1848 and 1918 were followed by sharp reverses and times when the very survival of democracy was in doubt. Yet the trend on our scale is visibly upwards. This advance of the democratic paradigm has coincided with renewed interest in the mechanics of democracy. Scores of elections and referendums have been held in central and eastern European and developing countries throughout the 1990s, generating requests for the presence of international observers. These monitors in turn had to gauge the impact of electoral arrangements that went far beyond electoral formulae - traditionally the favourite hunting ground of political scientists. This led to renewed emphasis on issues taken for granted in established democracies. Politicians, officials, and citizens in these countries are keener on election laws because they suspect, or know by experience, that there are many ways to manipulate the verdict of the voters. That the electoral machinery should not be solely in the hands of supporters of the government of the day, for example, is taken for granted in the West. Yet in many emerging democracies, this remains a sharply disputed issue. Governments prefer having the electoral machinery staffed entirely by their supporters in order to avoid surprises (or for darker purposes), while opponents complain that there will be widespread fraud if there is no independent or multi-party electoral commission. It is assumed in the West that once the outcome of the election in a polling district or a constituency is known, it will be duly reported, yet there are countries where the single official document confirming the result is transmitted to higher authorities through policemen and possibly tampered with along the way. The decline of trust in government, parliamentary institutions, and politicians in established democracies throughout the 1990s has also helped to reignite debates on electoral reform, especially as it is now more difficult to contrast the weaknesses of democracy with the vices of totalitarian regimes. One possible answer to this crisis of confidence

6 Establishing the Rules of the Game

is reform of electoral rules. Many observers believe that new electoral systems, the introduction and/or more frequent use of direct-democracy devices such as referendums and the popular initiative, and/or more stringent election finance laws would help to reduce the widening gap between rulers and public. They point to disturbing trends such as the decline in turnout at elections in most democracies throughout the 1990s and at times argue that new voting procedures, such as mail-in ballots, would drive more citizens to the polls. Whether institutional reforms alone can solve this problem is matter for debate (Massicotte 1999,178-^81; Nevitte 2002), but there is no doubt that they rank among the palliatives that deserve consideration. Democracy raises major issues that go beyond electoral arrangements. For example: Is democracy the unimpeded, unlimited rule of leaders supported by an electoral majority or the rule of the majority tempered by respect for minority rights under the supervision of a powerful judiciary? Should elected representatives be given a blank cheque, or should the people be empowered to review some of their decisions at referendums, put their own priorities forward through popular initiatives, or even recall them before the end of their term if they are unfit? Yet few people would challenge the idea that democracy requires free elections that produce effective decision-makers. The ideal is that of citizens who after free and full debate decide who their rulers will be. The reality, however, is that there are many ways to reach this goal. The tremendous proliferation of election laws in some parts of the world in the 1990s justifies by itself a comparative examination. The decade saw, for example, major alterations to election legislation in the United Kingdom and Canada resulting in the former in the creation of an electoral commission and in the latter in the introduction of a register of electors. Brazil has implemented the use of electronic voting machines across all its territory. Election laws may also have a highly symbolic significance. Whether or not the ballots cast by prison inmates can alter the final outcome, their very right to vote is an issue whose relevance exceeds the electoral arena. Whether non-citizens or citizens abroad are enfranchised raises a debate on the very boundaries of the political community. Excluding the mentally deficient from the franchise irrespective of the severity of their handicap sends also a powerful message to some of the individuals concerned. But there is more: election laws are not neu-

Introduction: Establishing the Rules 7

tral, their design increases or decreases the chances of the various parties and candidates. At the 2000 U.S. presidential election, Al Gore outpolled George W. Bush by half a million votes yet lost the election. This anomaly occurred because the U.S. president is elected by the people through a 538-member electoral college. Although the existence of the college is usually defended as a logical requirement of federalism, the United States remains the only federation with a presidential system where the president is elected other than by direct popular vote. The common view until recently was that this hardly mattered, because at every election held after 1888 the candidate who led in the popular vote also won an absolute majority in the college and hence the election. Yet in 2000, it mattered immensely, for Gore lost despite his lead in the popular vote nationwide. This paradoxical result occurred because the apportionment of seats among states in the college provides a slight bonus to the smaller (i.e., less populated) states. Throughout the 1990s, the ten smallest units of the union (i.e., the nine smallest states and the District of Columbia), with 2.8 per cent of the U.S. population, had 5.9 per cent of the seats in the college, while the two largest states (California and New York) had 19.2 per cent of the population but only 16.2 per cent of the electors. In a close contest, this could be enough to make the difference.1 The 21 units carried by Gore had a total population of 128.6 million (52 per cent of the total), while the 30 that went to Bush had an aggregate of 120.1 million. By carrying most smaller states, Bush was able to overcome his - admittedly slight - deficit in the popular vote. Had seats in the college been distributed strictly according to population, actual voting figures suggest that the outcome of the election would have gone the other way. At the 1993 election in New Zealand, Prime Minister Bolger's National Party won a one-seat absolute majority when overseas voters - a category still disfranchised in many democracies - swung one constituency to National. For decades prior to the adoption of the U.S. federal voting rights acts, various devices - poll taxes, means test, and restrictive registration laws - in southern states, severely diluted the impact of the votes of African Americans. The historical importance of the right to vote is widely acknowledged. Without the enfranchisement of the poor, women, and Blacks, the political dynamics of the twentieth century would have been quite different. Enfranchisement gave disadvantaged groups a stake in the

8 Establishing the Rules of the Game

political community and provided them with an opportunity to influence the political process to their advantage. Most citizens have the right to vote. In some countries, a sizeable number of people are barred from voting because they are not citizens and cannot acquire citizenship under the terms of their entry. They are expected to pay taxes and to obey the laws, yet they have no say in either matter. In countries where the very presence of foreigners with different racial backgrounds is a substantial public issue, it is easier for extremists to indulge in immigrant-bashing, for the latter are unable to retaliate by supporting other parties. The exclusion of non-citizens from the franchise seems natural, especially in countries where foreigners are few, yet in a world where territorial boundaries may become less meaningful a growing number of countries have started to grant foreigners the right to vote in local elections. Excluding prison inmates from the franchise can affect elections. Some two million Americans are currently imprisoned and may not vote. About one in ten African Americans aged 25 to 29 is jailed. A large proportion of prison inmates are African Americans, usually a solidly Democratic constituency. Election laws matter not only for academics, but also for voters. A faulty registration procedure that results in thousands of electors being left out may be traumatic for the individuals concerned, and not only when the outcome of the election is close, as exemplified by the high number of complaints filed with election officials following each election. In an era of increasing disaffection for politics and politicians, voting remains one of the precious few opportunities for citizens to make some difference and to legitimize the electoral process by their involvement. Citizens who go to the polls only to find that they cannot vote because their name has been omitted from the list, or because they have been impersonated by somebody else, are understandably angry. Election officials are also concerned about the details of the electoral legislation that they have been appointed to implement. There is much evidence that they have become more proactive over the years in exploring new ways and means to facilitate voting, to ensure the fairness and credibility of election, or to reduce costs. For example, Elections Canada was active throughout the 1980s and 1990s in identifying possible improvements to the Canada Elections Act, and successive chief electoral officers have issued numerous reports advocating possible reforms. The performance of that duty requires reliable information on practices in other jurisdictions, widely used procedures, and prob-

Introduction: Establishing the Rules 9

lems associated with each option. This is even truer of election officials in countries undergoing democratic development. Election-law issues are of interest also to politicians, because they understand that legislative change may well affect their electoral prospects. Post-election weeks are usually peppered with complaints from politicians, not always those who lost, about some aspects of the election process, either following complaints from their workers or supporters or because they are outraged by the result. This brings us to another reason for considering election legislation carefully. Taken in isolation, each aspect of the electoral law may look innocuous. Yet the various dimensions are segments of a long chain that is no more solid than any of its individual links. The debate that followed the U.S. presidential election of 2000 focused on issues that sounded arcane to many observers concerned with 'the big picture': whether voting cards in Florida had been suitably punched, whether a recount in that state could be done within the prescribed deadline, whether some voters had been omitted from the list. Yet these issues mattered tremendously not only because they possibly affected the outcome, but also because they cast doubt on the very legitimacy of the new president. It is doubtful that many citizens found time to get acquainted with the nuts and bolts of the issues at stake, but what ultimately counted was the impression that was left in many voters' minds. The contest was in effect decided by a few thousand votes cast in a single state and counted under the supervision of higher officials formerly identified with one party. These officials were ultimately accountable to the governor, who happened to be the brother of one of the candidates. The controversy was refereed by two courts, one with a majority openly identified with one party, and the other staffed by a majority of judges appointed by presidents from the other party. In other words, an apparently minor issue had the potential to lead to a major constitutional crisis. The goal of the electoral process is not simply to produce winners, but also to ensure that the victory of the winners be seen as legitimate even to those who did not support them (Nadeau and Blais 1993). Politicians are intense individuals for whom defeat is a harsh blow, especially when the end result is a photo-finish. In such instances, the temptation is great, for them or for their supporters, to point to any irregularity or instance of fraud to question the winner's legitimacy. Dissatisfaction with the outcome can easily lead to blaming the rules. The analogy between election laws and 'the rules of the game' is sound but we should not push it too far. Sporting matches are not mat-

10 Establishing the Rules of the Game

ters of life and death (although the amounts of money involved may be quite astounding), and the rules regulating them are usually very old and remain a stable quantity. In contrast, politics is no mere 'game/ and election laws are rules that are often questionable, that are debated, and altered routinely, sometimes against the wishes of one side. In this context, politicians may question the legal framework that resulted in their defeat - a reaction that may, however, reduce the legitimacy of the elected leaders and of the polity as a whole. In the Philippines, for example, blatant manipulation of the counting process by President Ferdinand Marcos in 1986 led to widespread public protest and ultimately to his downfall. If the electoral machinery did not matter, there would be no need to send international election observers into developing democracies. Human rights organizations in these countries would not devote immense efforts to training legions of local observers for the same purpose. If foreign and local observers have become a feature of elections in developing countries, it is because the international community has been keen to ascertain on the spot whether the election was conducted in a free and fair manner. It is not enough to reach judgment on whether the electoral formula has produced an outcome that is proportional to the nominal strength of parties; it is also indispensable to make sure that the process was fair in its actual implementation. METHOD Our research is based on first-hand sources: constitutions and election laws. These legal documents do not necessarily tell the whole story, but they remain the most accurate statement of the rules of the game.2 We also sent a questionnaire to election officials in the countries selected. Their answers proved very useful in clarifying ambiguous provisions or explaining supplementary details not specified in election statutes. Interviews with experts in some of the countries were also extremely helpful. The objective of this study is to provide a comprehensive examination of legislative and presidential election laws in existing democracies. In order to determine whether a country belongs to that group, we rely on the annual reports of Freedom House on the state of democratic rights and civil liberties in all countries of the world. On these two criteria, each country receives a rating between 1 and 7, with higher numbers indicating poorer showing. We consider 'democratic'

Introduction: Establishing the Rules 11

all countries that had a political-rights score of 1 or 2 in the 1996-7 survey, which is consistent with Freedom House's approach of considering countries with scores of 1 and 2 as 'free' countries (Gastil 1979, 24).3 Freedom House's ratings provide the most systematic evaluation of the degree of democracy, and it has been judged quite accurate (Bollen 1993). Its index of democracy is now widely used in cross-national research (Burkhart and Lewis-Beck 1994; Helliwell 1994; Cox 1997). We excluded countries with a population of less than 100,000 from the list. We had to drop two federal countries, Switzerland and the United States. This had nothing to do with any unfavourable assessment of their political institutions, and we acknowledge that both have made wonderful contributions to the development of election law. However, unlike most federations, both rely largely, even for federal elections, on the election laws adopted by each state or canton, which accordingly vary from one jurisdiction to another. As our study necessitates a single 'answer' for each country, we found it impossible to compress the variety of solutions found in the subnational jurisdictions of either country into a single 'answer,' to be compared with those for other countries. Selecting the largest subnational unit in each case as 'typical' was not acceptable, as there are 50 U.S. states and 26 Swiss cantons, none of which includes more than 17 per cent of the total population. This makes problematic any generalization based on either California or Zurich, for example. The total number of these subnational jurisdictions in the two countries exceeds the number of sovereign countries covered in the book. Furthermore, nine Swiss cantons have a population smaller than the benchmark that we applied to sovereign countries - namely, 100,000. This left us with a group of 72 countries, of which we had to delete nine for lack of information. Our study thus includes 63 countries from all continents: 18 from North and South America, 8 from Asia, 6 from Oceania, 8 from Africa, and 23 from Europe (10 of which are former communist countries).4 We examine six dimensions of election laws: the right to vote, the right to be a candidate, the electoral register, the agency in charge of the election, the procedure for casting votes, and the procedure to sort out winners and losers. All these dimensions belong to what one could call the conduct of elections. We have assembled considerable information on election laws in democratic societies. In each chapter, we examine the kinds of rules that have been established in our sample, we determine the relative

12 Establishing the Rules of the Game

frequency of various rules, we point out sociological patterns - which kinds of countries tend to adopt which kinds of legislation - and we review normative and practical arguments for and against each set of rules. Previous research has shown that former British colonies have been prone to adopt the electoral system (single-member plurality) of the mother country (Blais and Massicotte 1997), and it remains to be seen whether the same colonial influence affects other aspects of their election legislation. We also wish to determine whether there are systematic differences between long-standing and new democracies. For instance, with respect to the right to vote, are democracies where political rights in general are usually perceived to be better protected more inclusive? Freedom House considers countries that had a politicalrights rating of 1 uninterruptedly from 1987 to 1996 'established' democracies, while other democracies are 'non-established.' Because they have a major administrative component, these issues may appear 'technical'; they are, in some sense. Deciding whether to create a new electoral register for each election or to set up a permanent electoral list depends substantially on such technical considerations as the reliability of census data. But as numerous authors in public administration have stressed, there is no clear-cut separation between administration and politics. We show below that decisions on how to establish an electoral register are not 'neutral,' that they involve value judgments between contradictory objectives, and that they may have notable political consequences. We do not cover all aspects of election laws. We leave aside the analysis of electoral systems - the object of numerous studies. We also ignore three other dimensions: the regulation of campaign finances, the delimitation of electoral districts, and the regulation of political communication. These domains are very complex, and each deserves a separate study. As our comparison is between countries, we do not deal with state and local electoral rules. CONTENTS OF THE BOOK Chapter 1 addresses the franchise. Most definitions of democracy require some sort of citizen participation in the decision-making process. Universal suffrage is usually construed as essential to a fully democratic constitution. However, all societies restrict the right to vote, on account of, for example, age, citizenship, and country or electoral district of residence. Chapter 1 focuses on seven such restrictions,

Introduction: Establishing the Rules 13

including disfranchisement of prison inmates, of mentally deficient persons, and of citizens residing abroad. Chapter 2 flips over the coin and looks at candidates: who has the right to be a candidate in legislative and presidential elections? Once again, it describes and analyses restrictions to this fundamental right. Coupled with the previous chapter, it should give the reader an accurate account of the inclusiveness - or lack thereof - of various countries in deciding who has the right to participate in the democratic process. The compilation and updating of electoral registers form the topic of chapter 3. In all of our countries but one, electoral registers include the names of those who have the right to vote and the electoral district where they may exercise this right. The registration process can be very intricate; for this reason, we resort to case studies that show the breadth of procedures in force. Chapter 4 lists the various types of central electoral administration that exist and describes their appointment process. Since the details of the appointment process defy clear and simple characterization, we resort once again to case studies to show how the administration actually functions. Chapter 5 deals with the conduct of elections on polling day. To start with, we define polling day in democratic countries. We then review what electors must do once they arrive at the polling station, as well as the types of ballot papers and special procedures involved in voting. Finally, chapter 6 discusses the ultimate part of the game: determining the results. Elections produce winners and losers, and this chapter provides an overview of how ballot papers are counted, how parties and candidates may ask for a recount, and who has the authority to try election petitions and invalidate the results. Our objective in this volume is to open up a field of research - the comparative study of the conduct of elections - that has been largely neglected. It is time to redress the imbalance created by the literature's almost entire focus on electoral systems. Katz (2000) is right. It is time to look at the 'other' aspects of election laws. Because we are exploring largely unknown territory, we have taken great care to assemble systematic comparative information in election laws of our 63 countries.5 We hope that other scholars and researchers will use that information for their own specific purposes. We show evidence of an astonishing variety of election laws across contemporary democratic societies. We have also tried to uncover

14 Establishing the Rules of the Game

underlying patterns in the kinds of countries that have tended to adopt a particular set of rules. The various chapters present and describe these patterns. We focus on two dimensions: whether former British colonies have been prone to follow the example of the mother country and whether non-established democracies have adopted a set of rules distinct from those of established democracies.

Chapter One

Who Has the Right to Vote?

Universal suffrage is usually a basic criterion for an election to be deemed democratic. According to Dahl (1989, 233), one of the seven conditions for the existence of what he calls a polyarchy is that 'practically all adults have the right to vote.' Similarly, universal suffrage is the very first criterion listed by Butler, Penniman, and Ranney (1981, 3). Yet, as Katz (1997, 216) reminds us, 'no country allows all adults to vote ... Although the basic trend over the last 200 years has been to remove one barrier after another, many restrictions remain.' Throughout the nineteenth and early twentieth centuries, the franchise was a lively issue. Whether women and less affluent citizens should be enfranchised was a hotly debated topic. In contrast, contemporary disqualifications affect smaller groups such as prison inmates, non-citizens, and mentally deficient persons. Still the issue of who should and should not have the right to vote is deeply perplexing. Katz (1997, 216) identifies three major types of restrictions on the right to vote: 'those based on community membership and having a personal stake in the election, those based on competence, and those based on autonomy'. The distinction between competence and autonomy is fuzzy, as Katz (1997, 232) himself acknowledges.1 But the distinction between restrictions based on community and those based on competence is an important one. Katz's first heading raises the question of whether only those who have formal citizenship should have the right to vote and whether they should have resided and should still be residing in the community in order to be qualified to vote. In the same vein, we would locate an issue not discussed by Katz - whether prisoners should have the right to vote - which raises the question of whether 'imprisonment could

16 Establishing the Rules of the Game

still be looked on as the temporary exclusion of the individual from the community' (Report of the Royal Commission on the Electoral System 1986, 236). Katz's second set of restrictions has to do with competence. This raises two sets of issues. First, at what age are people mature enough to be able to vote in a reasonable manner? Second, should people with mental disabilities have the right to vote, and/or should that right depend on the severity of the disability? This chapter considers how democratic countries limit voting rights, and it examines compulsory voting. We look first at restrictions on the right to vote: those related to voting age, mental disabilities, citizenship, residence (district, country, abroad), and imprisonment. The chapter provides an overview of laws in place and of the debate about who should and who should not have the right to vote. We examine the general pattern of election laws. We describe the frequency and severity of the various restrictions. We look at whether a consensus exists among democratic countries about legitimate restrictions. We also identify particularly innovative kinds of legislation. We consider, second, whether existing democracies perceive voting as a right or a duty. More precisely, we look at countries where voting is compulsory and sanctions for non-compliance. Tables 1.1 and 1.2 present summary information on restrictions on the right to vote in the 63 countries. Tables 1.3 and 1.4 provide information on compulsory voting. RESTRICTIONS ON THE RIGHT TO VOTE We first discuss competence-based restrictions on suffrage (age and mental disability), followed by restrictions based on community (citizenship, residence, and criminality). Minimum Voting Age The most common restriction based on competence is of course the voting age. Excluding non-adults from the franchise is commonly justified on the ground that only mature people can make reasoned choices and that, in view of the difficulty of measuring personal maturity, it is more useful to rely on age. Children and adolescents may lack the requisite knowledge and understanding and may be influenced by their parents. Supporters of a higher voting age believe that maturity increases over

Who Has the Right to Vote? 17

time; therefore the older, the better. Those who want to lower voting age argue that adolescents are better informed and more independentminded than they used to be. They also point out that in other areas of life, teenagers are treated as adults: they may hold a job, they pay taxes, and, in some instances, they may be prosecuted in a court of law. They should therefore have the same rights as adults. Yet there is a near-consensus within the democratic arena regarding the minimum voting age.2 Fifty-nine of our countries (94 per cent) set the voting age at 18. Only Brazil has a lower threshold, at 16.3 Japan and Taiwan set the voting age at 20. The highest minimum voting age that we find is 21, in Samoa. Every country has a single minimum voting age for all electors. This has not always been the case. Before 1995, the constitution of Bolivia made a distinction based on voters' marital status. Married citizens got the vote at 18, and all others at 21.4 Since 1995, the minimum voting age in Bolivia has been 18 for all citizens.5 Two of the three countries with voting age higher than 18 are in Asia. The voting age is higher in a few other Asian countries not included in our analysis: South Korea (20), Malaysia (21), Maldives (21), Pakistan (21), and Singapore (21). Many Asian societies have a different conception of age - as testified by the greater prestige they bestow on the older generation - which leads them to believe that people achieve wisdom later in life and that there should be no hurry in allowing a person to vote. Whether in family life or in business, elders are revered and take more responsibility than their younger counterparts (Pye 1985). Looking at Table 1.2, we find that established democracies are more likely to set the minimum voting age at 18. Indeed, while all established democracies set the voting age at 18, only 90 per cent of nonestablished ones do so. Mental Disabilities Depriving mentally deficient people of the right to vote is seemingly a self-evident solution. How can people make a reasonable choice when their very own sanity is in serious doubt? This lack of capacity and the impracticability of allowing patients in mental hospitals to vote are used to justify their disfranchisement (Robertson 1994, 289-92). Yet criteria for mental illness have varied across time and space, and while serious illness may warrant disqualification, lighter and occasional mental problems should not. However, it is nearly impossible to draw

Table 1.1 Restrictions on the right to vote

Citizenship requirement

Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Yes

Yes

NA

No

No

Yes (sentence not available)

18

Yes

Yes or citizenship of another Commonwealth country before 1 984

1 month

No

No, for 6 years (with the intention of returning to the country)

Yes, sentence of 5 years or more

Bahamas

18

Yes

Yes

3 months

No

Yes

Yes, any sentence

Bangladesh

18

Yes

Yes

No

No

Yes

No

Barbados

18

Yes

Yes or citizenship of another Commonwealth

3 months (Commonwealth citizens only)

3 years (Commonwealth citizens only)

No, for 5 years

Yes, any sentence

Belgium

18

Yes

Yes

No

No

No

Yes, sentence exceeding 4 months

Belize

18

Yes

Yes or citizenship of another Commonwealth country

2 months

1 year (Commonwealth citizens only)

Yes

Yes, sentence exceeding 1 year

Benin

18

Yes

Yes

No

No

No

Yes, sentence of 3 months or more

Country

Voting age

Disfranchisement of mentally deficient persons

Argentina

18

Australia

Disfranchisement of prison inmates

Table 1.1

(Continued)

Citizenship requirement

Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Yes

Yes

No

No

Yes (No for presidential elections)

No

16

Yes

Yes

No

No

Yes (No for presidential elections)

Yes, any sentence

Bulgaria

18

Yes

Yes

No

No

No

Yes, any sentence

Canada

18

No

Yes

No

No

No, for 5 years (with the intention of returning to the country)

Yes, sentence exceeding 2 years

Cape Verde

18

Yes

Yes

No

No

No

Yes, any sentence

Chile

18

Yes

Yes or residence for 5 years

No

5 years (noncitizens only)

Yes

Yes, any sentence

Costa Rica

18

Yes

Yes

No

1 year (naturalized citizens only)

No

NA

Cyprus

18

Yes

Yes

4 months

6 months

Yes

Yes, any sentence

Czech Republic 18

Yes

Yes

No

No

No

No

Denmark

18

Yes

Yes

No

No

No, for 12 years

No

Ecuador

18

Yes

Yes

No

No

Yes

Yes (sentence not available)

Country

Voting age

Disfranchisement of mentally deficient persons

Bolivia

18

Brazil

Disfranchisement of prison inmates

Table 1.1 (Continued)

Citizenship requirement

Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Disfranchisement of prison inmates

Yes

Yes

No

No

No

Yes, any sentence

18

Yes

Yes

6 months

No

No

Yes (certain offences only)

Germany

18

Yes

Yes

No

3 months

No (in European countries) No for 10 years (in other countries)

No

Guyana

18

Yes

Yes or citizenship of another Commonwealth country

No

1 year (Commonwealth citizens only)

Yes

No

Hungary

18

Yes

Yes

No

No

Yes

Yes, any sentence

India

18

Yes

Yes

No

No

Yes

Yes, any sentence

Ireland

18

No

Yes or British citizenship

NA

NA

NA

No

Israel

18

NA

Yes

No

No

Yes

NA

Italy

18

No

Yes

No

No

No

NA

Country

Voting age

Disfranchisement of mentally deficient persons

Estonia

18

France

Table 1.1(Condinuced)

Country

Voting age

Disfranchisement of mentally deficient persons

Jamaica

18

Yes

Yes or citizenship of another Commonwealth country

Japan

20

Yes

Latvia

18

Lithuania

Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

No

1 year (Commonwealth citizens only)

Yes

Yes, sentence exceeding 6 momths

Yes

3 months

No

NA

Yes (certain offences only)

Yes

Yes

No

No

No

Yes, any sentence

18

Yes

Yes

No

No

No

No

Luxembourg

18

Yes

Yes

No

No

No

Yes, any sentence

Madagascar

18

Yes

Yes

No

No

Yes

Yes, any sentence

Malawi

18

Yes

Yes or residence for 7 years

No

7 years (non-citizens Yes only

No

Mali

18

Yes

Yes

6 months

No

No

Yes, sentence exceeding 1 month

Malta

18

Yes

Yes

No

6 months (in the preceding 18 months)

No (as long as not more than 12 months outside the country in the preceding 18 months)

Yes, sentence exceeding 1 year

Citizenship requirement

Disfranchisement of prison inmates

Table 1.1

(Continued)

Citizenship requirement

Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Disfranchisement of prison inmates

Yes

Yes

3 months

9 months

No

Yes, any sentence

18

Yes

Yes

No

No

Yes

Yes, any sentence

Namibia

18

Yes

Yes

No

No

NA

No

Netherlands

18

Yes

Yes

No

No

No

Yes, sentence of 1 year or more (certain offences only)

New Zealand

18

Yes

Yes or permanent residence

1 month

1 year

No, for 1 year (permanent residents) No, for 3 years (citizens)

Yes, sentence of 3 years or more

Panama

18

Yes

Yes

6 months

No

No

Yes, any sentence

Papua New Guinea

18

Yes

Yes

6 months

No

Yes

Yes, sentence exceeding 9 months

Philippines

18

Yes

Yes

6 months

1 year

No (with the intention of returning to the country)

Yes, sentence of 1 year or more

Poland

18

Yes

Yes

No

No

No

No

Country

Voting age

Disfranchisement of mentally deficient persons

Micronesia

18

Mongolia

Table 1 1

(Continued) Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Country

Voting age

Disfranchisement of mentally deficient persons

Portugal

18

Yes

Yes, or citizenship of EU member state, or Brazilian citizenship with equal rights status

No

No

No (legislative elections only)

Yes, any sentence

Romania

18

Yes

Yes

No

No

No

No

St Lucia

18

Yes

Yes or citizenship of another Commonwealth country

2 months

7 years (Commonwealth citizens only)

No, for 3 years

Yes, any sentence

St Vincent and 1 8 the Grenadines

Yes

Yes or citizenship of another Commonwealth country

3 months

1 year (Commonwealth citizens only)

No, for 5 years

Yes, any sentence

Samoa

21

Yes

Yes

No

1 year (only for those not citizens by birth or whose parents are not citizens)

No

Yes, any sentence

Sao Tome and Principe

18

Yes

Yes

No

No

No

Yes (certain offences only)

Citizenship requirement

Disfranchisement of prison inmates

Table 1.1

(Continued)

Citizenship requirement

Minimum period of residence in the electoral district required

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Yes

Yes

No

No

No

Yes (sentence not available)

18

Yes

Yes

No

No

No

No

South Africa

18

Yes

Yes

No

No

NA

No

Spain

18

Yes

Yes

No

No

No

Yes (certain offences only)

Sweden

18

No

Yes

No

No

No

No

Taiwan

20

Yes

Yes

4 months

No

No

NA

Trinidad and Tobago

18

Yes

Yes or citizenship of another Commonwealth country

2 months

1 year (Commonwealth citizens only)

Yes

Yes, sentence exceeding 1 year

United Kingdom 18

Yes*

Yes or citizenship of another Commonwealth country or of Ireland

No

No

No, for 20 yearst

Yes, any sentence

Uruguay

Yes

Yes or residence for 15 years

No

3 years (naturalized Yes citizens) or 15 years (non-citizens)

Country

Voting age

Disfranchisement of mentally deficient persons

Slovakia

18

Slovenia

18

Disfranchisement of prison inmates

Yes, any sentence

Table 1.1

(Concluded)

Country

Voting age

Disfranchisement of mentally deficient persons

Citizenship requirement

Vanuatu

18

Yes

Yes

Venezuela

18

Yes

Yes

Minimum period of residence in the electoral district required 3 months (only for those voting in an electoral district other than the one in which they were born)

No

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

Disfranchisement of prison inmates

No

No

No

No

No

Yes, any sentence

NA = Not available * Under new legislation passed in 2000, patients resident in a mental hospital, whether voluntarily or detained as a result of mental health legislation, may now register to vote. T Effective 2002, the qualifying period for overseas voters was reduced to 15 years.

Table 1.2 British colonialism, political rights, and restrictions on right to vote

British colonialism Former British

Other countries

Political rights Established democracies Other countries

Total

Voting age of 18

Disfranchisememt of mentally deficient persons

Right to vote restricted to citizens only

Minimum period of residence in the electoral district required

96% (n = 23)

91% (n = 22)

52%*

(n =; 23)

92% (n = 39)

95% (n = 39)

100%f (n = 22)

82%

90% (n = 41)

100%

94% (63 countries)

Minimum period of residence in the country required

Disfranchisement of citizens residing abroad

50%

55%*

55%

(n = 22)

(n = 22)

(n = 20)

68% (n = 22)

92% (n = 39)

18% (n = 38)

15% (n = 39)

21% (n = 38)

75% (n = 36)

59%f (n = 22)

38% (n = 21)

43% (n = 21)

14%*

(n = 21)

80% (n = 20)

(n = 40)

85% (n = 41)

25% (n = 40)

22% (n = 41)

42% (n = 38)

69% (n = 39)

94% (62 countries)

76% (63 countries)

30% (61 countries)

29% (62 countries)

32% (59 countries)

73% (59 countries)

f

(n = 22)

* Difference significant at 0.01 level (two-tailed f-test). T Difference significant at 0.05 level (two-tailed f-test).

f

f

Disfranchisement of prison inmates

Who Has the Right to Vote? 27

a line that is not arbitrary between 'serious' and 'non-serious' disabilities (Denoncourt 1991,120-1). Only four countries - Canada, Ireland, Italy, and Sweden - do not restrict in any way the right to vote for mentally challenged persons. In Canada, this has been the case since a judicial decision rendered in 1988. Previously, mentally disabled persons did not have the vote. The remaining 58 countries all have some kind of restriction, usually on people adjudged incompetent or of unsound mind by a court of law.6 In some cases, the simple fact of being a patient in a mental hospital is enough to warrant disfranchisement. Twelve countries, including Bulgaria, Chile, Estonia, Guyana, Jamaica, and the Netherlands, have disfranchisement for mental reasons entrenched in the constitution. Table 1.2 shows a significant difference between more- and lessestablished democracies on this dimension. It is not that the former automatically allow all mentally deficient people to vote - very few countries do so. Only where democracy is strongly established does it seem to be possible to envisage ignoring serious mental deficiency as a cause for denying people the vote. Citizenship Requirement Requiring citizenship for voting purposes arguably helps to preserve the cohesiveness and boundaries of the national community. Before receiving the vote, one should be fully integrated in one's society. Recently arrived immigrants may be less familiar with issues and more vulnerable to manipulation. Some observers find it shocking that recent immigrants might in close contests prevent the majority of citizens of longer standing from getting what they want. Those who favour relaxation of the citizenship requirement for voting argue that it amounts to political discrimination, since it treats non-citizens as mere 'subjects' of the state (Beaud 1992, 413)7 Immigrants who pay taxes and obey the laws, should have a say on these taxes and laws. In John Stuart Mill's words (1972, 279), 'it is a personal injustice to withhold from any one, unless for the prevention of greater evils, the ordinary privilege of having his voice reckoned in the disposal of affairs in which he has the same interest as other people.' We find that 48 countries (76 per cent) restrict the right to vote to cit-

28 Establishing the Rules of the Game

izens. At the other end of the spectrum, four countries do not require voters to be citizens, but this does not automatically mean enfranchisement for everyone. Non-citizens must have been residing in Chile for at least five years to be granted the vote there. Malawi has a seven-year residence requirement. Uruguay requires fifteen years, coupled with 'good conduct.' In New Zealand, non-citizens must be permanent residents to vote. The third possibility is to enfranchise non-citizens coming from specific countries only. We find eleven countries with such an arrangement. All but one (Portugal) share a common bond: they are former British colonies (plus the United Kingdom itself) and give the right to vote to current residents who are citizens of a Commonwealth country. 'Commonwealth clauses' are found in Australia, Barbados, Belize, Guyana, Jamaica, St Lucia, St Vincent and the Grenadines, Trinidad and Tobago, and the United Kingdom8; Ireland extends the vote to British citizens only. Portugal grants the right to vote to citizens of the European Union ordinarily residing in the country and to Brazilian nationals, who have special 'equal rights.'9 While nearly all other countries limit the vote to citizens only (92 per cent), only half of former British colonies do so (52 per cent) (see Table 1.2). However, the 'Commonwealth clause' enfranchises a relatively small proportion of the electorate. If we exclude this exemption, the difference is no longer significant (only four countries let non-citizens vote, two of them former British colonies). We find a statistically significant difference between established and non-established democracies (see Table 1.2). The former are less likely to restrict the right to vote to citizens. However, analogous to the difference between former British colonies and other countries, most of the erstwhile colonies are also established democracies. Once we factor this in, established democracies are no less likely than non-established to require citizenship. In every country but two - Ireland and Taiwan - the citizenship requirement is the same for legislative and presidential elections. In Ireland, British citizens may vote only in legislative elections. For presidential elections, the electorate is composed of Irish citizens only. In Taiwan, to vote in presidential contests non-citizens need to have resided in the country for just four months, whereas only citizens may vote in legislative elections. In both countries, qualifications are less restrictive with respect to the more important national election.

Who Has the Right to Vote? 29 Residence Requirements ELECTORAL DISTRICT

The requirement that electors have been residing in their electoral district or constituency for some period prior to the election is defended on the ground that integration in a local community is essential for helping determine its future. Opponents claim that this can result in widespread disfranchisement of people who move frequently and that elections are national contests, not purely local ones. Eighteen of our countries require a minimum period of residence in an electoral district. The duration varies from one month in Australia and New Zealand to six months in France, Mali, Panama, Papua New Guinea, and the Philippines. The median requirement in those countries is three months.10 Although Britain has no such requirement, in order to vote for Westminster in Northern Ireland a person must have been residing there during the full three months prior to the qualifying date. Vanuatu is a unique case, with a three-month residence requirement for electors who wish to vote in an electoral district other than the one in which they were born. Barbados requires three months of residence in the electoral district only for those enfranchised under the 'Commonwealth clause.' Table 1.2 shows the significant difference between former British colonies and other countries concerning residence requirement vis-a-vis electoral district. This might seem surprising, as the former colonies do not follow the example of the mother country. However, their electoral system may explain their higher propensity to require a minimum period of residence. Indeed, such residence requirements are rarer among countries using a system of proportional representation , or PR (7 per cent), than in other countries (48 per cent). And previous research has shown that former British colonies tend to adopt a firstpast-the-post electoral system (Blais and Massicotte 1997). Furthermore, the average residence requirement is shorter (2.7 months) there than in other countries with such a requirement (4.9 months). Established and non-established democracies display no statistically significant difference in this matter. However, the former require on average shorter residence in the electoral district (2.6 months) than do non-established ones (4.3 months). COUNTRY RESIDENCE

Requiring electors to have been residing in the country for some

30 Establishing the Rules of the Game

period prior to the election tends to be based on the belief that recent arrivals are not informed enough to cast a meaningful vote. Some observers counter that there is no reason to disfranchise people just because they happen to have been away in the recent past. These people have the same obligations; they should have the same rights. How many of our democracies require a period of residence in the country? Eighteen do so (29 per cent).11 Minimum duration varies from three months in Germany.to seven years in Malawi and St Lucia. The median requirement is 12 months.12 In 12 countries, the requirement is not the same for all electors. For instance, in Belize, St Lucia, St Vincent and the Grenadines, and Trinidad and Tobago, the requirement applies only to 'Commonwealth clause' electors. Costa Rica requires residence of one year for naturalized citizens; citizens by birth are exempted from this requirement. Chile imposes a five-year residence condition on non-citizens. In Samoa, there is a 12-month residence requirement only for those who are not citizens by birth or whose parents are not citizens. Uruguay stipulates that non-citizens must have resided in the country for at least 15 years, naturalized citizens for at least three years, and citizens by birth not at all.13 Table 1.2 shows that more former British colonies (55 per cent) than other countries (15 per cent) require a minimum period of residence in the country. However, in most former colonies, the minimum period of residence applies solely to Commonwealth citizens. Only 14 per cent the same proportion as in other countries - require a minimum period of residence for all citizens. The difference between established and non-established democracies is not significant (p < 0.12). However, more established democracies (43 per cent) than non-established ones (22 per cent) require a minimum period of residence. CITIZENS RESIDING ABROAD

Demanding actual residence in the country for voting purposes was for long a standard and firm requirement of election laws. Canada relaxed this rule for soldiers, who had to serve abroad. It was felt unfair that military service should deprive them of a voice in the running of the country. This rationale led later to the extension of the same privilege to diplomats and other civil servants abroad. Many commentators argued that preserving the right to vote of civil and military servants abroad while disfranchising other citizens who happened to be abroad for study or travel, providing international assistance, and so

Who Has the Right to Vote? 31

on amounted to discrimination, especially where people are more likely to spend long periods outside the country, which reasoning led Parliament in 1993 to provide voting mechanisms for these people as well. In countries with high rates of emigration, maintaining the right to vote of expatriates may send a message to them that they are still part of the national community and will be welcome if they come back. Some observers insist that expatriates have less interest in the running of their country of origin, especially if they do not pay taxes there, or they point out the costs of such voting and the danger of fraud. Many feel that expatriates' right to vote in two different countries amounts to unacceptable privilege. Do citizens residing abroad retain their right to vote in democratic countries? They do in a majority (40) of our countries. In 30, such as France, Mali, the Philippines, and Venezuela, they retain the right to vote indefinitely. In the remaining 10, they retain the right to vote for a period ranging from three to twenty years.14 In New Zealand, citizens lose their right to vote three years after leaving the country, and permanent residents do so after only one year abroad. In Germany, citizens who live in a member state of the Council of Europe keep their vote indefinitely, while Germans living in any other country keep it for 10 years. In three countries, Australia, Canada, and the Philippines, citizens residing abroad must, on application for registration, state their intention of returning to the country in order not to be disfranchised. The latter condition was also in force in the United Kingdom from 1985 to 1989. Three countries established different rules for different types of elections. In Bolivia and Brazil, citizens living abroad keep their right to vote in presidential elections only. In Portugal, they do so for legislative elections only. The Netherlands constitutes an interesting case. Dutch citizens residing abroad have the right to vote, except for those residing in the Dutch Antilles or in Aruba and who have not been resident in the Netherlands for at least 10 years. In some cases, however, the right to vote is more symbolic than real. We find ten countries where citizens residing abroad must return to the country in order to cast their vote on election day. This arrangement is found notably in Barbados, the Czech Republic, Italy, Malta, St Lucia, and Slovakia. In Italy, while electors need to go back to the country to vote, the state tries to facilitate the exercise: state employees working abroad have up to three days to vote and have all their travel costs

32 Establishing the Rules of the Game

reimbursed by the state, while ordinary citizens residing abroad because of their work receive only train fare. More than half of Britain's former colonies disfranchise citizens residing abroad; 21 per cent of other countries do so. As with residence requirements, former British colonies are not in sync with the United Kingdom. However, it is only since 1985 that Britain has allowed its citizens residing abroad to vote. And many former British colonies are not known for massive emigration. More established democracies are less inclined to disfranchise citizens residing abroad. The pattern is the same as with respect to mentally deficient persons. However, while there seems to be a clear norm among established democracies that citizens residing abroad should keep the right to vote, very few grant the same right to all mentally deficient persons. Prison Inmates The disfranchisement of prisoners stems from a belief that society is based on an implicit contract that obliges everyone to comply with the law. Those who break the law violate their pledge and are hence unworthy of participating in the democratic process: 'only citizens have the right to vote, and it would not be reasonable to consider criminals as citizens' (Planinc 1987, 154). The possibility - however remote - that in a tight race the vote of people incarcerated could tip the balance is abhorrent to some. Allowing prison inmates to vote also raises practical problems, such as how and where they might vote. Enfranchising inmates is argued on various grounds. Some commentators believe that the contemporary penal regime aims ultimately at rehabilitation rather than at punishment and that preserving the voting rights of inmates facilitates their social reintegration. Some propose a distinction between those who serve short sentences and those incarcerated for serious crimes, and that only the latter should lose voting rights. Are prisoners disfranchised in democratic countries for any prison sentence whatsoever? They are in 23 countries, such as Brazil, India, Portugal, the United Kingdom, and Venezuela. Yet 16 countries, such as Germany, Namibia, and Sweden, let all prisoners keep their voting rights.15 Others remove the right to vote for sentences of specified length or longer, ranging from one month in Mali to five years in Australia. In France, Japan, Sao Tome and Principe, and Spain, only some prisoners lose their vote, depending on the nature of the crime.

Who Has the Right to Vote? 33

In Belgium and the Philippines, disfranchisement extends beyond the prison sentence. Belgium disfranchises for life those imprisoned for five years or more, for 12 years those imprisoned for between three and five years, and for 6 years those imprisoned for between four months and three years. In the Philippines, disfranchisement of those imprisoned for a year or more expires five years after the completion of the sentence.16 COMPULSORY VOTING The problem of the twentieth century/ a British observer wrote in the 1920s, 'has shown itself to be that of persuading the people to make use of the right for which they clamoured: to get them not only to vote in a responsible manner, but to get them to vote at all' (Robson 1923, 569). This issue became more salient during the 1990s, which saw electoral turnout declining in most democracies (Blais 2000). One radical solution is for the law to oblige electors to vote and even to make nonvoters liable to some kind of sanction. Voting is then called 'compulsory/ whether there are sanctions or not, and normally occurs at levels substantially higher than those in other countries (Blais and Dobrzynska 1998; Franklin 1999; 2002). The origins of compulsory voting have been traced back to the Swiss canton of St. Gallen in 1835 (Gosnell 1935). Belgium adopted that rule in 1893, and Australia in 1924. Some countries that formerly had compulsory voting, such as the Netherlands, Spain, and Czechoslovakia, dropped it later. Compulsory voting was advocated more frequently before the Second World War than it is today. Possibly the experience of the interwar period, when high-turnout Weimar Germany went totalitarian while lower-turnout Britain and the United States stayed democratic, led legislators to conclude that a high electoral turnout, while desirable in theory, was no guarantee for the survival of democracy. This section looks at the cases for and against compulsory voting, at how widespread the practice is, and at what explains its existence. The Cases for and against Compulsory Voting Compulsory voting is premised on a view of voting as a civic duty, to be enforced by the state. Every member of the community should participate in the deliberations of the body politic. Low turnouts, it is alleged, hamper the legitimacy of elected officials by throwing doubt

34 Establishing the Rules of the Game

as to what the result would have been if every voter had gone to the polls. It is argued that compulsory voting yields collective, and ultimately individual, goods because it 'permits citizens to be "true citizens" in the Greenian sense (by enabling them to participate in the business of governing) but it also enables the state to be a "true" one, that is, a state that actively works to prevent exclusion' (Hill 2002, 96). Less disinterested politicians may advocate compulsory voting as a way to improve the performance of parties that do well among sections of the electorate with traditionally low turnouts or more generally, as a way to spare all political parties from spending resources in order to bring electors to the polls (Jaensch 1995,19-21). Opponents of compulsory voting retort that voting is not a duty, but a right, and that people should be free to exercise or not to exercise that right. One scholar went as far as suggesting that compulsory voting is incompatible in principle with free elections, as 'one of the democratic liberties is to decline to vote for any of the alternatives offered' (Mackenzie 1958, 131). In an era when politicians get little respect from the public (see Nevitte 1996), non-voting may be a legitimate way to express disillusionment with, or alienation from, standard party politics - a convenient warning signal to an out-of-touch political class. Some observers also point out the difficulty of enforcing such a rule in the face of massive non-compliance or the danger that electors dragged to the polls reluctantly may spoil their ballot or vote for extremist or marginal candidates. Finally, many feel, as a politician once disdainfully put it, that 'the opinions of the negligent and apathetic section of the electors are not worth having' (quoted in Hughes 1966, 83). Compulsory voting works to the advantage of parties supported by categories of people who would not vote otherwise. In Australia, there is evidence that Labor and Democrat supporters are overrepresented among those who would not turn out if voting was voluntary, while the Coalition parties (Liberal and National parties) have a huge lead among those who would have voted irrespective of compulsion (Jackman 1999).17 How Widespread Is Compulsory Voting? On compulsory voting, countries fall into three groups: those where voting is compulsory and non-voting is sanctioned by the law; those where voting is compulsory, but with no sanctions against non-voters; and those where voting is not compulsory - or 'voluntary voting.'

Who Has the Right to Vote? 35

The third group happens to include a strong majority of democracies: in 45 of our countries (71 per cent), voting is not compulsory, and neither the constitution nor election law provides that it is one's 'duty' to cast a ballot on election day. In the other 18 countries, voting is compulsory, with sanctions (11 countries) or without (7 countries). Table 1.3 lists these countries. Twelve of the 18 have written the obligation to vote in their constitution.18 The most stringent rule - compulsory voting backed by sanctions is found in 11 countries (17 per cent of democracies): Argentina, Australia, Belgium, Bolivia, Brazil, Chile, Cyprus, Ecuador, Luxembourg, the Philippines, and Uruguay. The nature of the penalties is specified by the law (they are listed by country in Table 1.3). In most cases, nonvoting is sanctioned by a fine, which may be higher for repeat offenders. In some of these countries, non-voters may lose certain rights. For instance, in Argentina non-voters may not work in public administration for three years following the offence. In Brazil, those who do not vote in three consecutive elections are struck off the electoral register. In the Philippines, non-voters may not be candidates at the next election and may not be appointed to a position in public administration within the next year. In a few of these countries, some groups of electors are excused if they do not vote. In Argentina, this is the case for electors aged 70 or more. In Brazil, voting is optional for electors aged between 16 and 18 years old or 70 and over. In Cyprus, those aged 70 and over are excused only if they reside at least 50 miles from the polling station (Table 1.3 lists the categories of electors exempted from compulsory voting). Seven countries (Cape Verde, Costa Rica, Italy, Panama, Portugal, Sao Tome, and Venezuela) make voting compulsory but impose no sanctions for non-voters. Hence legislators rely on voluntary adherence to the law to reach the desired outcome. For instance, in Italy, the notion of voting as a 'civic duty' is entrenched in the constitution (section 48). Non-voters do not face any penalty per se. However, failure to vote is recorded on one's ID card for five years, and it is alleged that this may hurt chances of employment. What Explains Compulsory Voting? When grouping countries of both categories of compulsory voting together, we find that there are no major differences between established and non-established democracies (see Table 1.4). However, one finds significant differences between former British colonies, which

36 Establishing the Rules of the Game Table 1.3 Countries with compulsory voting Countries

Applicable sanctions

Categories of electors exempted

Argentina

Fine No possibility of working in public service for three years following offence

Electors 70 and over Judges and their substitutes Electors living more than 500 km from the polling station Electors residing abroad Handicapped electors who are unable to vote and those attending to handicapped electors Certain categories of public servants

Australia

Fine Reprimand

None

Belgium

Fine Electors in custody Repeat offenders may be excluded from the electoral register for a period of 10 years, during which they may not receivre any nomination, promotion or r distinction from a public authority

Bolivia

Fine

None

Brazil

Fine Repeat offenders excluded from electoral register Non-voters may not work in public service or receive passport

Electors 1 6-1 8 or 70 years and over Illiterate electors

Cape Verde

None (civic duty)

None

Chile

Fine

Electors unable to vote because of illness Electors abroad or located more than 200 km from the polling station

Costa Rica

None (civic duty)

None

Cyprus

Fine

Electors 70 and over and residing 50 miles or more from polling station

Who Has the Right to Vote? 37 Table 1.3

(Concluded)

Countries

Applicable sanctions

Categories of electors exempted

Ecaudor

Fine

Electors over 75 Illiterate electors

Italy

None (civic duty)

None

Luxembourgj

Fine

Electors residing in another commune Electors over 70 Electors with 'reasonable motive'

Panama

None (civic duty)

None

Philippines

Fine Non-voter may not be candidate in following election Non- voter may not be appointed to public administration in following year

None

Portugal

Non (civic duty)

None

Sao Tome and Principej

None (civic duty)

None

Uruguay

None Fine Repeat offenders lose voting right s

Venezuela

None (civic duty)

None

tend not to oblige electors to vote, and other countries, 41 per cent of which have compulsory voting. The list in Table 1.3 presents interesting patterns. Many countries with a Catholic majority require electors to vote.19 Perhaps it is natural for Catholics to believe that citizens have moral obligations not only towards God and the church, but also towards the state. As a result, the act of voting may be portrayed as a criterion of good citizenship and communal life. A less disingenuous explanation is that historically the church has feared that political apathy might deny victory to churchsupported parties and accordingly pressed for compulsory voting.

38 Establishing the Rules of the Game Table 1.4 British colonialism, political rights, and compulsory voting Compulsory voting (irrespective of sanctions) British colonialism Former British colonies Other countries

9%* (n = 23) 41%

(n = 39)

Political rights Established democracies Other countries Total

31% (n = 22) 27% (n = 41) 29% (63 countries)

* Difference significant at 0.01 level (two-tailed f-test).

Every former Spanish and former Portuguese colony has some form of compulsory voting. Finding such similarity within a group of former colonies necessitates a look at the former colonial power. Indeed, Portugal currently has compulsory voting without sanctions. Two of its three former colonies in our sample - Cape Verde and Sao Tome - also do not penalize non-voters. However, Brazil does. Voting was once compulsory in Spain. The electoral law of 1907 provided for compulsory voting and sanctions for non-voters, although enforcement was lax (Gosnell 1935). Furthermore, in continents such as North America, Africa, and Asia, where compulsory voting is the exception rather than the rule, former Spanish and Portuguese colonies are nevertheless inclined towards compulsory voting.20 In the 1920s, an observer noted that compulsory voting had been introduced in 'countries of a great dissimilar character' (Robson 1923, 575). This still holds true today. Yet the chief finding of this chapter is that culture matters: while Anglo-Protestant countries, with the shining exception of Australia, do not have compulsory voting, most Hispanic Catholic countries do, even if, as in the Philippines, their democratic training took place under American rather than under Spanish colonial rule.

Who Has the Right to Vote? 39 CONCLUSION Of the seven potential restrictions to the right to vote examined in this chapter, only two have given rise to a near-consensus among democracies. An overwhelming majority have decided that the voting age should be 18 and that people with severe mental deficiency should not vote. Outside former British colonies, few countries grant the vote to non-citizens, and PR countries have no required minimum period of residence in the electoral district. There is wide divergence about disfranchisement of prison inmates, a country-residence requirement, and enfranchisement of citizens residing abroad. This chapter on the franchise suggests that the most important issue concerning the right to vote in a democracy pertains to the residence requirement. The issues, especially of whether to allow citizens who reside abroad to vote, under what conditions, and for how long raise deep questions about the meaning of democracy in a world in which people are becoming increasingly mobile. One of our objectives was to establish whether former British colonies have emulated the right-to-vote laws of the mother country, as they have done with respect to electoral systems. There is a clear colonial heritage: about half of them have granted citizens of Commonwealth countries the right to vote. Partly because most of these erstwhile colonies have single-member constituencies, they are more likely to have electoral-district requirements. Most others do not allow citizens residing abroad to vote. In doing so, they follow what was British practice until 1985. How long will they wait before emulating the recent British law? We have also looked at whether countries with secure political rights have fewer restrictions on the right to vote. The short answer is yes: they tend to be somewhat more inclusive, although the difference is relatively small. On compulsory voting, most onetime British, Spanish, and Portuguese colonies follow their former rulers' lead. Even in continents with little compulsory voting, former Spanish and Portuguese colonies have adopted it. Does such a requirement enhance or hamper democracy? It is hard to say. Compulsory voting is rather uncommon in democracies, even well-established ones. Only one country (the Netherlands, in 1970) has changed its legislation (abolishing compulsory voting) in the recent past. Perhaps this is an issue that has remained dormant over the last few decades, but that may come back on the political agenda.21

Chapter Two

Who Has the Right to Be a Candidate?

Historically countries have tended to impose age, gender, religious, and property qualifications on the right to be a candidate - as on the right to vote - but these have become less stringent or have simply withered along with voting disqualifications based on similar grounds. Among the countries covered by this study, we found no restriction on candidacy based on property, gender, or religion, although such restrictions existed in the past.1 This chapter, however, deals with other types of restrictions under seven headings - age, citizenship, residence (in district and/or in country), incompatibilities, monetary deposits, supporting signatures, and other restrictions. Restrictions on candidacies take two forms: the law may make a person ineligible by not allowing him or her to stand as a candidate - if, for example, he or she does not fulfill requirements about age, citizenship, or residence. In addition, it may declare membership of parliament/ legislature or the presidency incompatible with various offices, which caveat obliges candidates, once elected, to relinquish the other offices. The fourth section of this chapter deals with incompatibilities. The distinction between these two types of restrictions has very practical consequences. For example, in a federal country where membership in a state or provincial legislature is incompatible with membership in the national legislature, a state legislator may run for national office without resigning his or her present position, thus enjoying all the advantages of incumbency, including indemnity, throughout the election campaign. If state legislators are ineligible for the national parliament, they must resign prior to running for office and risk having no seat at all in the end. As our analysis focuses on restrictions on candidacy whatever their legal standing, we do not distinguish ineligibilities and incompatibilities in this chapter.

Who Has the Right to Be a Candidate? 41

Some restrictions on eligibility are fairly universal and uncontroversial, although detailed arrangements may vary from one country to another: this is the case with age and citizenship. Others are meant to prevent the multiplication of farcical candidates by imposing procedural requirements that oblige would-be candidates either to pay some amount of money and to risk the loss thereof (deposit) or to prove that they have the support of a sizeable segment of the electorate (confirmed by signatures on the nomination paper). Conditions for eligibility are supplemented by incompatibilities, which purport to prevent parliamentarians or presidents from holding offices or positions deemed to interfere with the proper fulfilment of their duties. We have focused our research on incompatibilities between national and regional or local elective offices and between elective office and public service which may lead to conflicts of interest. Incompatibilities may also be established between membership in the upper and lower legislative chambers and between executive and legislative positions. In addition, certain business activities may be incompatible with elective office. Rules on eligibility and compatibility are typically located in the national constitution, the election law, and the various laws governing the national parliament, the judiciary, and the public service. Such documents require close reading, as ineligibilities are sometimes implicit. For example, a proviso indicating that one must have the right to vote in order to be eligible implies that all voting disqualifications provided for elsewhere in the legislation also apply to candidates. If prison inmates are, for example, disqualified from voting, they are disqualified from standing as candidates as well. We summarize most common candidacy requirements for legislative elections in Table 2.1 and for presidential elections in Table 2.2. AGE

All countries covered in this survey impose some minimum age for being elected to either the legislature or the presidency. This suggests the existence of a widespread conviction that some degree of knowledge, life experience, and maturity is indispensable for filling a public office and that age provides a useful benchmark. As we saw in chapter 1, the same rationale has been accepted with regard to the franchise. Many people feel that stricter age restrictions are particularly relevant for higher positions such as the presidency. In this regard, the U.S. constitution sets a pattern by establishing 25, 30, and 35 as the minimum

Table 2.1 Candidacy requirements in legislative elections Ineligibility or incompatibility with other elective mandates or public service*

Country

Age

Citizenship and residence

Monetary deposit

Argentina

25

Citizenship for 4 years Born in electoral district or resident for 2 years

No

Regional elective mandate Local elective mandate Public service

Australia

18

Citizenship

Candidate deposit, reimbursed if candidate obtains at least 4 per cent of first preferences

Regional elective mandate Public service

Bahamas

21

Citizenship Residence in country for 1 year

Candidate deposit, reimbursed if candidate obtains more than 1/6 of valid votes

Public service

Bangladesh

25

Citizenship

Candidate deposit, reimbursed if candidate obtains at least 1/8 of valid votes

Public service

Barbados

21

Citizenship

Candidate deposit, reimbursed if candidate obtains more than 1/6 of valid votes

Public service

Belgium

21

Citizenship

No

Regional elective mandate Public service

Belize

18

Citizenship Residence in country for 1 year

Candidate deposit, reimbursed if candidate obtains more than 1 0 per cent of valid votes

NA

Benin

25

Citizenship Residence in country for 1 year (natural-born citizens) or for 1 0 years (naturalized citizens)

No

Public service

Table 2.1

(Continued)

Country

Age

Citizenship and residence

Monetary deposit

Ineligibility or incompatibility with other elective mandates or public service*

Bolivia

25

Citizenship by birth

No

Public service

Brazil

21

Citizenship by birth Residence in electoral district

No

Regional elective mandate Local elective mandate Public service

Bulgaria

21

Citizenship

No

Public service

Canada

18

Citizenship

Candidate deposit, reimbursed in half if candidate obtains at least 15 per cent of valid votes and the other half when candidate transmits return of election expenses

Regional elective mandate Local elective mandate (in three provinces) Public service

Cape Verde

18

Citizenship

No

Judges, members of armed forces, diplomats, members of election commission

Chile

21

Citizenship for 5 years No Residence in electoral district for 2 years

Regional elective mandate Local elective mandate Public service

Costa Rica

21

Citizenship for 10 years Residence in country for 10 years (naturalized citizens only)

No

Regional elective mandate Local elective mandate

Cyprus

25

Citizenship

Candidate deposit, reimbursed if candidate obtains at least 4 per cent of electoral quota

Regional elective mandate Local elective mandate Public service

Table 2.1

Country

(Continued)

Age

Ineligibility or incompatibility with other elective mandates or public service*

Citizenship and residence

Monetary deposit

Czech Republic 21

Citizenship

List deposit, reimbursed if list obtains at least 5 per cent of valid votes (7 per cent for coalition of two parties, 9 per cent for coalition of three parties, 1 1 per cent for coalition of four parties or more)

NA

Denmark

18

Citizenship

No

NA

Ecuador

25

Citizenship by birth Born in province or resident therein for 3 years (provincial deputies only)

No

Regional elective mandate Local elective mandate Public service

Estonia

21

Citizenship

Candidate deposit, reimbursed if candidate obtains 1/2 of electoral quota or if list of candidates obtains at least 5 per cent of valid votes

Public service

France

23

Citizenship

No

Public service

Germany

18

Citizenship for 1 year

No

NA

Guyana

18

Citizenship

No

Regional elective mandate Local elective mandate Public service

Hungary

18

Citizenship

No

Public service

India

25

Citizenship

Candidate deposit, reimbursed if candidate obtains more than 1/6 of valid votes

Regional elective mandate Public service

Table 2.1

(Continued) Ineligibility or incompatibility with other elective mandates or public service*

Country

Age

Citizenship and residence

Monetary deposit

Ireland

21

Citizenship

Candidate deposit, reimbursed if, at any time during count, candidate obtains more than 25 per cent of electoral quota

Public service

Israel

21

Citizenship

List deposit, reimbursed if list obtains at least 1 .5 per cent of valid votes

High public officials

Italy

25

Citizenship

No

Regional elective mandate Public service

Jamaica

21

Citizenship or Commonwealth citizenship Residence in country for 1 year

Candidate deposit, reimbursed if candidate obtains at least 1/8 of valid votes

Public service

Japan

25

Citizenship

NA

Certain specified public service personnel and others connected with elections

Latvia

21

Citizenship

List deposit, reimbursed if at least one candidate from list elected

Local elective mandate Members of armed forces, judges, prosecutors, members of National Control or Audit Department, state controller

Lithuania

25

Citizenship

Candidate deposit (in single-member district), reimbursed if candidate is elected List deposit (in multi-member district), reimbursed if list obtains at least 5 per cent of valid votes and if party transmits return of election expenses

Local elective mandate Public service

Table 2.1

(Continued)

Country

Age

Citizenship and residence

Monetary deposit

Ineligibility or incompatibility with other elective mandates or public service*

Luxembourg

21

Citizenship

No

Public service

Madagascar

21

Citizenship for 10 years

No

Regional elective mandate Local elective mandate Public service

Malawi

21

Citizenship

Candidate deposit, reimbursed if candidate obtains at least 5 per cent of valid votes

Public service

Mali

21

NA

Candidate and List deposit, not refundable

NA

Malta

18

Citizenship Residence in country for 6 month!5

Candidate deposit, reimbursed if candidate obtains at least 1 0 per cent of electoral quota

Public service

Micronesia

30

Citizenship for 15 years Residence in country for 5 years

Candidate deposit, not refundable

Regional elective mandate Local elective mandate Public service

Mongolia

25

Citizenship

Candidate deposit, reimbursed in its entirety if candidate is elected; reimbursed in half if candidate obtains a number of votes equal to or higher than average obtained by all candidates

NA

Namibia

21

Citizenship

No

Regional elective mandate Local elective mandate Public service

Table 2.1

(Continued) Ineligibility or incompatibility with other elective mandates or public service*

Country

Age

Citizenship and residence

Monetary deposit

Netherlands

18

Citizenship

List deposit (only for parties that did not gain any seat at previous election), reimbursed if list obtains at least 75 per cent of electoral quota

Public service

New Zealand

18

Citizenship Residence in country for 1 year

Candidate deposit, reimbursed if candidate obtains at least 5 per cent of valid votes

Public service

Panama

21

Citizenship for 15 years Residence in electoral district for 1 year

No

Public service

Papua New Guinea

25

Citizenship Born in electoral district or resident for 2 or 5 years1

Candidate deposit, reimbursed if candidate obtains at least 10 per cent of votes obtained by elected candidate

NA

Philippines

25

Citizenship by birth Residence in electoral district for 1 year

No

Regional elective mandate Local elective mandate Public service

Poland

21

Citizenship Residence in country for 5 years

No

Public service

Portugal

18

Citizenship

No

Regional elective mandate Local elective mandate Public service

Romania

23

Citizenship

No

Certain high officials

Table 2.1

(Continued) Ineligibility or incompatibility with other elective mandates or public service*

Country

Age

Citizenship and residence

Monetary deposit

St Lucia

21

Citizenship for 1 year Residence in country for 1 year (naturalized citizens only)

Candidate deposit, reimbursed if candidate obtains at least 1/8 of valid votes

Public service

St Vincent and the Grenadines

21

Citizenship or Commonwealth Citizenship Residence in country for 1 year

Candidate deposit, reimbursed if candidate obtains more than 1 5 per cent of valid votes

Public service

Samoa

21

Citizenship

Candidate deposit, reimbursed if candidate obtains at least 25 per cent of votes obtained by elected candidate

Public service

Sao Tome and Principe

18

Citizenship

No

NA

Slovakia

21

Citizenship

No

Judges, prosecutors, police officers, prison guards, members of armed forces

Slovenia

18

Citizenship

No

NA

South Africa

18

Citizenship

List deposit, reimbursed if at least one candidate from list is elected

Regional elective mandate Local elective mandate Public service

Spain

18

Citizenship

No

Regional elective mandate Public service

Table 2.1

(Concluded)

Country

Age

Citizenship and residence

Monetary deposit

Ineligibility or incompatibility with other elective mandates or public service*

Sweden

18

Citizenship

No

NA

Taiwan

23

Citizenship for 1 0 years Residence in electoral district for 4 months

Candidate deposit, reimbursed if candidate obtains at least 10 per cent of valid votes or, in case of overseas candidates and candidates in nationwide electoral district, reimbursed if candidate elected

Public service

Trinidad and Tobago

18

Citizenship Residence in country for 2 years Residence in electoral district for 2 months

Candidate deposit, reimbursed if candidate obtains at least 1/8 of valid votes

Public service

United Kingdom 21

Citizenship

Candidate deposit, reimbursed if candidate obtains more than 5 per cent of valid votes

Public service

Uruguay

25

Citizenship for 5 years

No

Regional elective mandate Local elective mandate Public service

Vanuatu

25

Citizenship

Candidate deposit, not refundable

Public service

Venezuela

21

Citizenship by birth

No

Regional elective mandate Local elective mandate Public service

NA = Not available. * Offices listed in this column are those specified in the electoral law. The absence of a specific office from the list means only that no specific ineligibility or incompatibility has been found by the authors. f Either two years immediately preceding the nomination or five years at any time.

Table 2.2 Candidacy requirements in presidential elections Country

Age

Citizenship and residence

Monetary deposit

Signature requirements

Argentina

30

Citizenship by birth or by descent Residence in country for 6 years

No

No

Benin

40 Citizenship for 10 years (max: 70)

Yes, reimbursed if candidate obtains at least 10 per cent of valid votes in first round

No

Bolivia

35

Citizenship by birth

No

No

Brazil

35

Citizenship by birth

No

No

Bulgaria

40

Citizenship by birth Residence in country for 5 years

Yes, reimbursed if candidate obtains at least 1 per cent of valid votes in first round

15,000 electors (independent candidates only)

Cape Verde

35

Citizenship by birth Residence in country for 3 years

No

1 ,000 electors

Chile

40

Citizenship by birth

No

0.5 per cent of electorate (independent candidates only)

Costa Rica

30

Citizenship by birth

No

No

Cyprus

35

Citizenship

Yes

NA

Ecuador

35

Citizenship by birth

No

No

France

23

Citizenship

No

No

Guyana

18

Citizenship

No

300 electors

Table 2.2

(Continued)

Country

Age

Citizenship and residence

Monetary deposit

Signature requirements

Ireland

35

Citizenship

No

No

Lithuania

40

Citizenship by birth Residence in country for 3 years

Yes, reimbursed if candidate collects 20,000 signatures

20,000 electors

Madagascar

40

Citizenship by birth

Yes, reimbursed if candidate obtains at least 1 0 per cent of valid votes in first round

No

Malawi

35

Citizenship by birth or by descent

NA

NA

Mali

35

Citizenship by birth Residence in country for 1 year

Yes, reimbursed if candidate obtains at least 5 per cent of valid votes in first round

No

Mongolia

45

Citizenship by birth Residence in country for 5 years

No

No

Namibia

35

Citizenship

Yes, reimbursed if candidate obtains at least 10 per cent of valid votes in first round or if no candidate obtains majority of valid votes

300 electors in at least 1 0 regions (independent candidates only)

Panama

35

Citizenship by birth

No

No

Philippines

40

Citizenship by birth Residence in country for 1 0 years

No

No

Poland

35

Citizenship

No

100,000 electors

Portugal

35

Citizenship by birth

No

7,500 electors

Romania

35

Citizenship

No

1 00,000 electors

Table 2.2

(Concluded)

Country

Age

Citizenship and residence

Monetary deposit

Signature requirements

Sao Tome and Principe

35

Citizenship by birth

No

250 electors

Slovakia

35

Citizenship

No

No

Slovenia

18

Citizenship

NA

NA

Taiwan

40

Citizenship by birth Residence in country for 4 months

Yes, reimbursed if independent candidate collects required number of signatures and then, if candidate obtains numbers of votes equal to or higher than 5 per cent of electorate (see text)

1 .5 per cent of electorate (independent candidates only)

Uruguay

35

Citizenship by birth

No

No

Venezuela

30

Citizenship by birth

No

No

NA = Not available.

Who Has the Right to Be a Candidate? 53

age for being elected to the House of Representatives, the Senate, and the Presidency respectively. Many countries also impose a higher minimum age for membership in the second (upper) chamber than for sitting in the first chamber, which squares well with the view that the youthful excesses of the latter must be curbed by the wisdom and experience of the former. The other theoretically possible age disqualification, preventing some people from standing for an office because they are too old, is largely absent, at least for direct elections.2 Statespersons such as Charles de Gaulle, who became president of France at the age of 67, and Konrad Adenauer, elected chancellor of West Germany at 74 (and in office until his late 80s), show that old age is not necessarily an obstacle for an outstanding political career. Yet recent revelations concerning the health problems of some leaders have fed doubts about the wisdom of some of their decisions (Accoce and Rentchnick 1996). On average, the age of eligibility for candidates in legislative elections is 21 years.3 The lowest age of eligibility, found in 18 countries, is 18. Twenty-six countries set the age of eligibility at 21, three at 23, and fifteen at 25. The highest threshold (30 years) is in Micronesia, where the Senate - the single house of the parliament of this federal country includes some features of an upper house, as one member is elected for a longer term in each state, irrespective of its population. As Table 2.3 shows, on average, being a former British colony does not significantly affect the age of eligibility. The same cannot be said about the level of democracy. Established democracies tend to allow candidates to run at a lower age than non-established ones. The difference is small but significant. Indeed, while only 6 of 41 (15 per cent) non-established democracies allow 18 year-olds to stand as candidates, more than half of established democracies (55 per cent) do so. It thus seems that more established democracies tend to adopt age requirements similar to those pertaining to the right to vote. Latin American countries tend to select either 21 (as do Brazil and Chile) or 25 (as Argentina). Asian countries, in keeping with their tendency to grant the right to vote at a later age, typically set the age of eligibility at 25 (found notably in India, Japan, and the Philippines). In 44 countries (70 per cent), the age of eligibility is higher than the voting age; never is it lower. This suggests that most countries demand more maturity and wisdom from their legislators than from those who elect them. For presidential elections, the average age of eligibility, at 34 years, is appreciably higher than for legislative elections. The required age

Table 2.3 British colonialism, political rights, and candidacy requirements (legislative elections) Average age of eligibility

Stringent citizenship Requirements of residence Country-residence requirements requirements in electoral district

20.8 years (n = 23)

9%* (n = 23)

9% (n = 23)

43% (n = 21)

91%* (n = 23)

2 1.6 years (n = 39)

37% (n = 38)

18% (n = 38)

29% (n = 38)

24% (n = 38)

19.8 years* (n = 22)

14%f (n = 22)

5%f (n = 22)

29% (n = 21)

55% (n = 22)

Other countries

22.2 years (n = 41)

33% (n = 40)

20% (n = 40)

36% (n = 39)

48% (n = 40)

Total

21 .3 years (63 countries)

26% (62 countries)

15% (62 countries)

33% (60 countries)

50% (62 countries)

British colonialism Former British colonies Other countries Political rights Established democracies

* Differences significant at 01 level (two-tailed Mest). f Differences significant at 10 level (two-tailed Mest).

Monetary deposits

Who Has the Right to Be a Candidate? 55

ranges from 18 (Guyana and Slovenia) to 45 (Mongolia). The two most frequent figures are 35 (found in 16 countries) and 40 (7). Here again, the (few) Asian countries that provide for direct election of the president impose a higher age (40 in Taiwan and the Philippines, 45 in Mongolia). We find only one country, Benin, where there is a maximum age of eligibility, set at 70 years for presidential candidates. In most countries, the age requirement for presidential candidates is more stringent than in the case of candidates to the legislature. All but three countries with direct presidential elections have a higher age of eligibility for presidential candidates. The average difference in age requirements for legislative candidates in presidential systems is 13 years. CITIZENSHIP REQUIREMENTS Citizenship is also a widely accepted criterion for being a candidate, the rationale being that one must be a bona fide member of the community in order to lead it or to represent its citizens. Being ruled by people who owe allegiance to another country historically ranks as a powerful stimulant for nationalist passions, as suggested by the dismantling of colonial empires. However, the increasing number of countries that provide for dual citizenship, and the increasing number of individuals taking advantage of that possibility, mean that citizenship per se may not entirely remove any doubt as to one's allegiance.4 All countries demand citizenship for presidential elections, and almost all do so for legislative elections. Only in Jamaica and St Vincent does one not need to be a citizen in order to be elected. These countries also extend the 'Commonwealth clause' described in chapter 1 to candidates, which makes them the most inclusive in terms of citizenship requirements. A more interesting contrast emerges between those countries where citizenship, whether obtained by birth or naturalization, is a necessary qualification for public office and those where only citizens by birth are eligible. This latter option has the effect of excluding all first-generation immigrants. For legislative elections, only five countries (Bolivia, Brazil, Ecuador, the Philippines, and Venezuela) require members of parliament to be natural-born citizens. Four of these five are in South America, and the fifth was formerly under Spanish rule. In all other countries, simple citizenship is enough, but some are less restrictive than others. For instance, 44 countries specify formal citizenship as the sole requirement.5 In 10 nations, citizenship must have been held for a

56 Establishing the Rules of the Game

minimum period, ranging from 12 months in Germany and St Lucia to 15 years in Micronesia and Panama. Thus, while citizenship is usually necessary for election to parliament, the exact requirements show some variation. To illustrate this point, we created a variable measuring those citizenship requirements. We coded countries (46) that require only formal citizenship or where the Commonwealth clause is applicable as 'lax/ Others (16), including those requiring citizenship by birth or descent or for at least a minimum period, we call 'stringent.' Unsurprisingly, we find that a significantly lower proportion of former British colonies and established democracies have stringent citizenship requirements (see Table 2.3). In contrast, requiring citizenship by birth is a frequent requirement for presidential candidates. Eighteen countries (60 per cent of those that hold direct presidential elections) do so. Elsewhere, a presidential candidate need simply be a citizen, except in Argentina, Malawi, and Namibia, where one must be citizen either by birth or by descent, and in Benin, where one must have been a citizen for at least 10 years. In the United States, the presidency is the sole office for which the constitution specifically requires incumbents to be natural-born citizens. The requirement that candidates be born in the country has very practical consequences in some countries. It can be used to disqualify opponents of the government who happen not to satisfy the condition. For example, opponents in the referendum on a new constitution for the Ivory Coast held in 2000 argued that the proviso requiring that the president be a bona fide Ivorian born of two Ivorian parents was specifically aimed at disqualifying the main opposition candidate. Once again, some patterns emerge. For instance, whereas only a quarter of our sample have 'stringent' requirements for legislative elections, that ratio reverses for presidential elections (73 per cent). Also, there is not a single country where citizenship requirements for candidates to the national legislature are more stringent than for presidential candidates. In other words, if one is eligible for the presidency, one is also eligible for the national legislature. This denotes the higher status often bestowed on the head of the executive branch of government. RESIDENCE REQUIREMENT Electoral District

Residence requirements for electoral districts find their source in a view of parliamentarians as at least partly delegates from a specific

Who Has the Right to Be a Candidate? 57

portion of the national territory. It is assumed that electors are better represented by a resident of their district. Requirements of district residence prevent the domination of peripheral districts by politicians from core areas. Having a locality rule makes it harder for party leaders to drop so-called carpetbaggers, whose credentials may be based on their good relations with the party brass rather than on close constituency ties. Against this view stands the doctrine, prevalent among European constitutional lawyers, that parliamentarians, once elected, represent not a particular district, but the nation as a whole. Germany, Italy, and Portugal6 have even written this theory into the national constitution. From this perspective, there should be no district-residence requirement for prospective candidates. Indeed, legislators are expected to transcend 'parochial' interests and to examine issues and problems from a national perspective. It is also argued that it is up to the voters to decide whether they wish to be represented by someone residing in their district. The absence of residence requirements may facilitate adjustment of electoral boundaries in countries with single-member districts, as no member will be legally obliged either to change place of residence or to switch to another district following an alteration of electoral boundaries. The vast majority of democracies do not impose any such requirement. Some go as far as to allow candidates to stand in more than one electoral district. Of course, if elected in two districts, they may keep only one seat, and a by-election has to be held for the other.7 This is the case in Bangladesh and India.8 In only six countries - Brazil, Chile, Panama, the Philippines, Taiwan, and Trinidad and Tobago - must candidates at legislative elections reside in their electoral district to be eligible. In Brazil, candidates must simply be residing in their electoral district, whereas the other five require a minimum period of residence in the district, ranging from two months in Trinidad and Tobago to two years in Chile. Three other countries - Argentina, Ecuador, and Papua New Guinea - require a period of residence, but only for those who were not born in the electoral district - a formula that incidentally provides a bit of protection against carpetbaggers, as prospective candidates may have moved from the place of their birth at an early age. The minimum duration of residence is two years in Argentina and three years in Ecuador (for provincial deputies only). In Papua New Guinea, candidates not born in the electoral district must have been living there for two years immediately preceding their nomination or for five years at any time.

58 Establishing the Rules of the Game

All in all, only nine countries require some kind of condition of residence in the electoral district. A higher proportion of non-established democracies have such a requirement. As they do with age and citizenship, strong democracies tend to show more inclusiveness; only one requires residence in the electoral district (Trinidad and Tobago). Also, a smaller proportion of former British colonies has electoral-district requirements - although this difference is not significant - despite their greater propensity to require electoral-district residence for the right to vote (see chapter 1). Country Residence This requirement, especially if it applies for long periods of time, has the same objective and is based on the same belief as the citizenship requirement - candidates should be thoroughly grounded in the national community. It may be criticized for creating an obstacle against recently arrived immigrants. In non-democratic countries moving towards free and fair elections, several potential candidates, having been in exile for many years, may be, intentionally or not, excluded by that requirement. No minimum period of residence in the country is required for legislative candidates in 45 democracies, although the requirement of residence in an electoral district found in five of these (Brazil, Chile, Panama, the Philippines, and Taiwan) logically implies residence in the country as well. In ten countries, all candidates at legislative elections must have resided in the country, for a duration ranging from six months in Malta to five years in Micronesia and Poland. In Benin, natural-born citizens must have been residing in the country for one year, and naturalized citizens for ten years. Further, in Costa Rica and St Lucia, naturalized citizens must have been residing in the country for 10 years and one year, respectively. In Argentina, Ecuador, and Papua New Guinea, the electoral-district-residence requirements described above also apply for the country. Only eight countries that have presidential elections demand that candidates be residents of the country for at least a minimum period, ranging from four months in Taiwan to 10 years in the Philippines. INCOMPATIBILITIES9 Incompatibilities stem from a belief that legislators or presidents should devote all their energy to the office to which they have been

Who Has the Right to Be a Candidate? 59

elected. They can also be defended as a way to ensure that regional or local offices will not be monopolized by national politicians who use them as mere weapons to comfort their national position, as is commonly alleged to be the case in some countries (Dion 1986). In federal countries, prohibiting dual mandates has historically been a way to protect the autonomy of member states or provinces against the central government. The desire to avoid conflicts of interest and corruption is also a major cause for legislated incompatibilities. The rationale against such incompatibilities is that they deprive parliaments of local talents and the familiarity with grassroots politics that is entailed by the fulfilment of a local or regional office. In federal countries, dual membership of central and state legislatures may reduce friction between both levels of government. A less frequently voiced argument is that local or regional offices, especially mayoral duties, held concurrently with a national legislative office, reinforce the stature of members of parliament and provide them with useful resources for fulfilling their national mandate. Excluding civil servants from legislatures stems from the doctrine of separation of powers. This practice dates back to eighteenth-century Britain and was aimed at protecting the independence of Parliament from the corrupting power of the crown. In days when ministries owed office to the favour of the sovereign and not just to the support of Parliament, members of Parliament holding an office of profit under the crown were perceived by some of their colleagues as a kind of 'fifth column' that helped to support whatever ministry the crown might be pleased to appoint. Still today, many people see serving in the legislature while belonging to the civil service as a basic conflict of interest, because a member of the legislature is expected to control the bureaucracy or at least to hold it accountable. There is also a fear that allowing civil servants to run for, or to fill, elective offices will politicize an institution that is expected to remain above the political fray.10 This view is countered by the assertion that civil servants bring an intimate and practical knowledge of government that may enlighten debate in the legislature. Some commentators feel that banning civil servants from assemblies while allowing legislators to work in the private sector amounts to discrimination. One can bridge the gap between these competing views by invoking the rule that no civil servant may sit in parliament but granting that civil servants may receive a leave of absence without pay from the civil service for the duration of the campaign or of their term if they are elected. Following their political career, they may have to resume their

60 Establishing the Rules of the Game

duties under the authority of their former political opponents - a situation that is hardly optimal in terms of the loyalty that ministers normally expect from their subordinates. Granting this right to civil servants provides them with a decisive advantage over would-be candidates working in the private sector, who may enjoy no such longterm privilege. We find 21 countries, including Australia, Canada, and India, where elective mandates at the regional level are incompatible with mandates at the national level. This incompatibility is found notably in all federal countries (the data for Germany were not available) and in the Spanish Estado autonomico. On the opposite end, five countries (Bolivia, France, the Netherlands, Papua New Guinea, and Taiwan) do not forbid such a combination.11 Slightly less common are incompatibilities between elective mandates at the local and national levels. Indeed, 17 countries, such as Cyprus, the Philippines, and Portugal, do not allow elective offices to be held at both levels simultaneously. In contrast, dual mandates are allowed in 10 countries, such as France, the Netherlands, and Spain. In Canada, three provinces (Alberta, Nova Scotia, and Quebec) explicitly forbid local officials from being at the same time federal legislators. For the remaining countries, we found no explicit provision for or against dual mandates. More widespread are incompatibilities with membership in the public service. As many as 46 countries do not allow public servants to serve simultaneously an elective mandate at the national level. Two countries, the Czech Republic and Denmark, do not impose any restriction. Six countries, including Israel and Slovakia, forbid only some categories of public officials (senior bureaucrats, soldiers, and judges, for example) from holding an elective office in the national parliament. For the remaining countries, we could not confirm the (in)compatibility of legislative office and membership in the public service. In some countries where public servants may not exercise an elective mandate, they are entitled to a leave of absence without pay from the public service for the duration of the mandate. Eight countries, such as Bulgaria and Brazil, have such an arrangement. In Israel, only some high-ranking public officials may obtain a leave of absence. As such rules may be included in the laws dealing with the public service or in collective agreements rather than in election legislation, or may not have been reported to us by the election officials whom we contacted, these figures are suggestive only.

Who Has the Right to Be a Candidate? 61 DEPOSIT

Requiring a deposit, especially a non-refundable one, presumably screens out farcical would-be candidates - at least those who cannot afford to lose too much money. As well, candidates compel the state to spend additional public money and should therefore perhaps pay for part of the administrative costs that they cause.12 For example, in countries following the French model of voting by ballot and envelope (see chapter 5), separate ballot papers for each presidential candidate must be printed, irrespective of the number of votes that each candidate is expected to receive. Deposits act like any surety demanded in private transactions: as they may be otherwise confiscated, they help to ensure that candidates will comply with their obligations. Opponents object that deposits, amounting to discrimination in favor of wealthy candidates, act as a camouflaged property requirement and bar entry by new forces into the political system. Many people, especially those who resent the influence of money over the political process, find truly abhorrent the principle that one has to pay to run for election. Democracies divide equally between those where a deposit is (31) and is not (31) required. Patterns are quite striking. Every former British colony but two (Guyana and Namibia) follows Britain's example and requires a deposit from legislative candidates. Seventy-six per cent of other countries do not require a monetary deposit, including France, Portugal, Spain, all South American countries, and all former Spanish and Portuguese colonies. Every country with plurality or majority electoral formulas and single-member districts (but France) requires deposits, while PR countries tend not to. Nationwide contests may be more likely to attract frivolous candidates, and so the costs engendered by each candidate at that level are likely to be much higher. As a consequence, one might expect deposits to be more common for direct presidential elections. In practice, we found just the opposite. For presidential contests, 20 countries do not require a monetary deposit, while eight countries require one, which is refundable in seven of them.13 Here again, not a single South American country requires a deposit, while former British colonies - a group notoriously less inclined to hold direct presidential elections - offer no clear pattern: Cyprus and Namibia require a deposit, Ireland and Guyana do not, while information was not available for Malawi. Ireland and Mongolia do not require a deposit by presidential candidates but

62 Establishing the Rules of the Game

do so for legislative candidates. These two exceptions do not seem to be related to the relative strength of the presidency vis-a-vis the legislature, since the Irish parliament has more clout than its president, while the reverse is true in Mongolia. Deposits for legislative candidates may be paid in several ways. Twenty-four countries require deposits to be made on behalf of each candidate, and five insist on a deposit for each list.of candidates. Two others call for a deposit for each candidate and for each list. In every country but one where deposits are mandatory, every candidate and/ or list of candidates must pay the required sum. The Netherlands, where deposits are demanded only from parties that did not win a seat at the previous election, stands as the exception. One way to deal with the argument that deposits hamper less wealthy candidates is to provide for their reimbursement should the candidate get at least a minimum level of support, thus proving his or her seriousness. In almost all countries (28 out of 31) that require deposits, reimbursement follows if candidates/party lists meet certain requirements specified by law.14 The most common requirement is obtaining at least a minimum percentage of all valid votes cast in the electoral district. In 15 countries, this is the only condition. Thresholds range from 1.5 per cent (Israel) to 16.7 per cent of valid votes (Bahamas, Barbados and India), with 10 per cent being the average. In four countries with PR systems, a party list must obtain at least a specified percentage of the electoral quota. The threshold ranges from 10 per cent of the quota for individual candidates in Malta to 75 per cent of the quota for party lists in the Netherlands. In Latvia and South Africa, the party deposit is lost unless at least one candidate from the list is elected. A similar criterion for reimbursement, found in Papua New Guinea and Samoa, is for the candidate to obtain a specified percentage or more of the number of votes received by the winner in the district (10 per cent in the former and 25 per cent in the latter). Five countries impose more than a single condition for refunding deposits. In Canada, a candidate's deposit is divided into two halves, each being subject to a different criterion for refunding: one-half of the deposit is reimbursed to candidates who were elected or received at least 15 per cent of the valid votes in their electoral district, while the other half is refunded when candidates (irrespective of their performance at the polls) file their return of election expenses with the chief electoral officer.15 Estonia reimburses individual candidates if they obtain at least one-half of the electoral quota or if the list on which they

Who Has the Right to Be a Candidate? 63

are nominated obtains at least 5 per cent of the vote. In Mongolia, elected candidates are fully reimbursed, while candidates who obtain at least the average vote received by all candidates in the electoral district are reimbursed half the deposit. More complex arrangements exist in two countries that have mixed electoral systems involving singlemember districts as well as multi-member districts. In Lithuania, candidates in single-member districts are reimbursed only if elected, while list deposits are reimbursed to parties that obtain at least 5 per cent of the votes and transmit their return of election expenses. Furthermore, the amount of the party deposit is doubled at the following election for parties that have failed to submit relevant documents. In Taiwan, while overseas candidates and party-list candidates must be elected in order to be reimbursed, the condition in a single-member district is to obtain 10 per cent or more of the vote. For presidential elections, the most common condition for refunding is a minimum percentage of valid votes cast in the country, which ranges from 1 per cent in Bulgaria to 10 per cent in Benin, Madagascar, and Namibia. In the latter countries, the deposit is also reimbursed if no candidate obtains a majority of votes. In Lithuania, the deposit is refunded to candidates who can muster 20,000 signatures. In Taiwan, independent candidates must first pay a deposit that is returned if they collect the required number of signatures (1.5 per cent of the electorate); then all candidates must pay another deposit, which is refunded only to those who obtain a number of votes equal to at least 5 per cent of the electorate. SUPPORTING SIGNATURES Some commentators advocate as more consonant with democratic ideals requiring that a candidacy be sponsored by at least a minimum number of electors to discourage frivolous candidates. However, nothing guarantees that those who sign the nomination papers really support the candidate - some candidates actually receive fewer votes than supporting signatures.16 Also, this requirement may amount to a (nonrefundable) deposit, since candidates must pay workers to collect signatures. Requiring the support of local elected officials, as is done for French presidential elections, may also act as a screening device that assists existing parties. Presidential candidates must be supported by a specified number of signatures of electors in 11 countries. In seven of them, every candidate

64 Establishing the Rules of the Game

must be supported by a number of electors, ranging from 250 in Sao Tome to 100,000 in Poland and Romania. In four countries - Bulgaria, Chile, Namibia, and Taiwan - only independent candidates have to collect signatures. In Bulgaria, candidates must be supported by 15,000 electors unless they are nominated by a registered political party. In Chile, independent candidates must be supported by 0.5 per cent of the electorate.17 In Namibia, independent candidacies must be nominated by at least 300 electors in at least 10 regions. In Taiwan, the requirement is 1.5 per cent of the total number of electors in the most recent election of national popular representatives. No signatures of electors are necessary in the other 16 countries with direct presidential elections for which we could gather the relevant information. However, in two of them, presidential candidates must have support from at least a specific number of representatives. In France, nominations must be supported by at least 500 elected officials - France has many local and regional offices, including over 36,000 mayors. In Ireland, support must come from 20 members of Parliament, or four county councils, or a former president. South American countries are less inclined to require signatures (only two out of our nine do). All European countries with direct presidential elections insist on signatures, except France and Ireland, which demand the support of elected officials. Africa is split equally between those that require signatures (Cape Verde, Sao Tome, and Namibia) and those three former French colonies (Benin, Madagascar, and Mali) that do not. OTHER RESTRICTIONS This chapter has focused on major and widespread constraints on candidacies. Our survey would be incomplete, however, without mention of some peculiar ineligibilities imposed by a few democracies. Some requirements seemingly aim at ensuring that legislators or presidents possess at least minimum abilities for their duties. In Chile, one must have completed secondary education or the equivalent in order to stand as a candidate in legislative elections.18 In Guyana, Malawi, St. Lucia, and St. Vincent, national legislators must be able to read and speak the English language with a degree of proficiency sufficient to participate fully in the work of parliament. There is also a language-proficiency requirement in Estonia, Latvia, and the Philippines.19 Fear of student radicalism may also explain why, in Taiwan,

Who Has the Right to Be a Candidate? 65

students currently attending school may not be candidates at an election. In Venezuela, members of the clergy are ineligible as presidential candidates - a prohibition inspired perhaps by a fear that clergymen use their influence over parishioners to bolster their political ambitions. A similar disqualification exists against members of the clergy (but only if they belong to the Anglican, Presbyterian, and Roman Catholic churches) for legislative elections in Britain; peers are also ineligible, unless they disclaim their peerage. In France, members who failed to submit a statement of their personal financial situation (declaration de patrimoine) or their election-expenses report are ineligible. The Mongolian constitution stipulates that presidential candidates must be nominated by a political party represented in the legislature. It winnows down the number of potential candidates and acts as a filter, excluding independent candidacies. Finally, in Luxembourg, a country with fewer than 400,000 people, members of parliament may not be 'parents ou allies jusqu 'au deuxieme degre' (in case voters decide otherwise, the youngest must forfeit the mandate). Perhaps is it believed that kinship and politics do not make a happy marriage. CONCLUSION Typically, candidates at legislative elections must be at least 21 years old, while the average minimum for presidential contests is 35. Citizenship is almost always required, but seldom is residence in the district or country, except for presidential candidates. Incompatibilities between national and regional or local mandates and with membership in the public service seem to prevail in a majority of countries. As for monetary deposits and supporting signatures in the case of presidential elections, there appears to be an almost equal split between those countries that require and those that do not require them. Established democracies generally impose fewer eligibility requirements. Former British colonies almost always demand monetary deposits, but partly because they have single-member districts - and all countries but one with single-member districts require deposits. Deposits are strikingly absent in all South American countries and in Spain, Portugal, and their other former colonies.

Chapter Three

Registering Voters

At the core of the electoral process lies the inclusion of electors on an electoral register.1 Although this has not always been the procedure, electors are nowadays listed on an electoral register in every democracy but one - Latvia.2 Using an electoral register diminishes the likelihood of electoral fraud. Political parties may also use the register when canvassing potential supporters. However, registers must be compiled and updated, and this inevitably entails costs for the state, which must make the burden for electors as small as possible. Once government decides to use an electoral register, it must make it as accurate as possible. Two pitfalls await the legislator in the search for accuracy: there may be too many people registered, or too few. The latter happens when omission from registration prevents otherwise qualified electors from casting a vote. To counter this possibility, legislators should aim for inclusiveness by offering many possibilities for electors to get on the register. But election authorities must also ensure that all individuals registered are truly qualified electors and that each elector is listed only once. They can do this by making sure that the register does not include people who do not have the right to vote. They must therefore strike the right balance between inclusiveness and strictness when devising the registration process. Fundamental differences arise among democracies with respect to the registration of electors. This chapter focuses on four such differences. First, does the state initiate the first registration of an elector, or is it up to the potential elector to seek registration? Second, is the first registration done on a continuous basis, or are there specific registration periods? Third, how does one ensure accurate registers (i.e., inclusiveness)? Fourth, are political parties involved at any stage? The processes are very complex, and we cannot summarize all of them, but

Registering Voters 67

we examine in detail the processes in four countries - Germany, Canada, Spain, and Australia - to highlight the interconnection between the various stages of the procedure. INDIVIDUAL OR STATE INITIATIVES The literature on democratic elections groups registration processes into three types: automatic registration, voluntary self-registration, and compulsory self-registration (Choe 1997, 136; Katz 1997, 115-16). We can distinguish these types on the basis of two criteria: first, whether it is up to qualified electors to take steps to be registered or whether the state itself takes steps to register electors, and, second, whether registration is voluntary or compulsory. We examine the latter issue below. Each method of registration has its advantages and drawbacks. The elector-initiated approach is cheaper, since the state minimizes its input in the process. It may also instill a sense of duty among electors, who are told that if they value their right to vote they must act to ensure that right. However, many otherwise qualified electors may be unable or unwilling to register or may simply forget to do so. The result may be a high proportion of unregistered electors. This shortcoming is especially worrying when under-registration occurs predominantly among the most disadvantaged groups of society; not only is the register incomplete, but it is disproportionately skewed in favour of the better-off. The registration process in the United States is regularly cited as a potent example of the defects of such a system (for example, Cunningham 1991; Powell 1986). State-initiated registration requires less effort from electors. It probably omits fewer electors, and the register tends to be more complete. It can be an arduous task, however, especially with a highly mobile population, and very expensive (Mackenzie 1958,117). The process for registering newly qualified electors is the most accurate indicator of the state's involvement. Does the state have the initiative (state-initiated registration) at the time of the first registration of electors, or does it not (registration on application)? Table 3.1 summarizes the procedure in each country. We find 20 countries, such as Australia, France, the Philippines, and Uruguay, where qualified electors must take the initiative. This does not mean that the state does nothing. For example, in Australia electoral authorities make alterations to the register every month when they receive notices of deaths from the registrar-general and lists of convictions from the controller-general of prisons. In France, a prefect or sub-prefect can ask for the deletion from

Table 3.1 Registration procedures in 63 democracies

Country

Initiative by state during first registration

First registration

Compulsory registration

House-to-house or other inquiry

Date of closing of register

Political parties involved

Argentina

Yes

Continuous

No

No

1 65 days before election

No

Australia

No

Continuous

Yes

Yes

26-51 days before election

No

Bahamas

No

Continuous

No

Yes

22-27 days before election

No

Bangladesh

Yes

NA

No

Yes

NA

No

Barbados

No

Continuous

No

No

19 days after writ is issued

No

Belgium

Yes

Continuous

No

No

1 2 days before election

No

Belize

Yes

Continuous

No

Yes

Between issue of writ and nomination day

Yes

Benin

No

Continuous

Yes

No

Election day

Yes

Bolivia

No

Continuous

Yes

No

45 days before election

Yes

Brazil

No

Continuous

Yes

No

1 00 days before election

Yes

Bulgaria

Yes

Continuous

No

No

30 days before election

No

Canada

Yes

Continuous

No

Yes

Election day

Yes

Cape Verde

Yes

Annual

Yes

No

30 days before election

Yes

Chile

No

Monthly

No

No

20 days before election

No

Table 3.1

(Continued)

Country

Initiative by state during first registration

First registration

Compulsory registration

House-to-house or other inquiry

Date of closing of register

Political parties involved

Costa Rica

Yes

Continuous

Yes

No

1 20 days before election

Yes

Cyprus

Yes

NA

Yes

No

Every 3 months*

No

Czech Republic

NA

NA

No

NA

NA

NA

Denmark

Yes

Continuous

No

No

1 8 days before election

No

Ecuador

Yes

Continuous

No

No

30 days before election

No

Estonia

NA

NA

No

NA

NA

NA

France

No

Continuous

Yes

No

1 0 days before election

No

Germany

Yes

Elections

No

No

1-3 days before election

No

Guyana

Yes

NA

Yes

Yes

NA

Yes

Hungary

Yes

Continuous

No

No

Election day

No

India

Yes

NA

No

Yes

See notet

No

Ireland

Yes

Annual

No

Yes

1 2 days before election

No

Israel

Yes

NA

No

No

NA

NA

Italy

NA

NA

No

No

NA

NA

Jamaica

Yes

Annual

No

Yes

4-1 1 days before election

Yes

Table 3.1 (Continued)

Country

Initiative by state during first registration

First registration

Compulsory registration

House-to-house or other inquiry

Date of closing of register

Political parties involved

Japan

Yes

Annual

No

No

NA

NA

Latvia

No register

No register

No register

No register

No register

No register

Lithuania

Yes

NA

No

No

Election day

Yes

Luxembourg

Yes

Annual

No

No

NA

No

Madagascar

Yes

Annual

No

No

5 days before election

No

Malawi

No

Elections

Yes

No

1 4 days before election

Yes

Mali

Yes

Annual

No

No

December 31 1

Yes

Malta

Yes

Continuous

No

No

Twice a year5

Yes

Micronesia

No

Continuous

No

No

30 days before election

No

Mongolia

NA

NA

No

NA

15 days before election

NA

Namibia

No

See note"

No

Yes

20-30 days before election

No

Netherlands

Yes

Continuous

No

No

43 days before election

No

New Zealand

No

Continuous

Yes

Yes

1 day before election*

No

Panama

Yes

Continuous

Yes

NA

1 65 days before election

No

Papua New Guinea

No

Continuous

Yes

No

70-84 days before election

No

Table 3.1

(Continued)

Country

Initiative by state during first registration

First registration

Compulsory registration

House-to-house or other inquiry

Date of closing of register

Political parties involved

Philippines

No

Elections

Yes

Yes

37 days before election

Yes

Poland

Yes

Continuous

No

No

Election day

No

Portugal

No

Monthly

Yes

No

NA

Yes

Romania

Yes

Annual

Yes

No

45-75 days before election

No

St Lucia

No

Continuous

No

Yes

At least every 6 months**

No

St Vincent and the Grenadines

No

Continuous

No

No

Every 4 monthstt

No

Samoa

No

Continuous

No

Yes

See notesii

No

Sao Tome and Principe

NA

NA

Yes

NA

NA

NA

Slovakia

Yes

Continuous

No

No

Election day

No

Slovenia

NA

NA

NA

NA

NA

NA

South Africa

No

NA

No

No

Prescribed by regulations

NA

Spain

Yes

Monthly

Yes

No

40-46 days before election

No

Sweden

Yes

Elections

No

No

NA

No

Taiwan

NA

NA

NA

NA

NA

NA

Table 3.1

(Concluded)

Country

Initiative by state during first registration

First registration

Compulsory registration

House-to-house or other inquiry

Date of closing of register

Political parties involved

Trinidad and Tobago

Yes

Continuous

No

Yes

22-25 days before election

Yes

United Kingdom

Yes

Annual

No

Yes"

13 days before election

No

Uruguay

No

Continuous

Yes

No

Election day

Yes

Vanuatu

NA

NA

No

No

15 June""

No

Venezuela

Yes

Monthly

Yes

No

90 days before election

Yes

NA = Not available. * The register is published every 3 months (in January, April, July, and October). It closes 10 days after each publication and remains in force until the next publication. f Last day fixed for nominating a candidate (7 days after issue of writ). * Each year, on 31 December. The register remains in force for one year. However, the president of the Independent National Election Commission may order a special revision at any time. Also, electors who move because of a job transfer may register and vote on polling day. 5 Published twice a year, in April and October. Claims may be made in the 21 days following publication. The register remains in force until the next publication. " Updated at intervals of not more than 10 years or from time to time as determined by the president. * Those who will become 18 on or before polling day may register and vote on election day. ** Published at intervals of not more than 6 months. Once published, it remains in force until the next publication. n The 'quarterly list' is published every 4 months. Once published, it remains in force until the publication of the following 'quarterly list.' ** The register is closed before election day, on a date fixed by the chief electoral officer. §§ Under the Representation of the People Act 2000, qualified electors may now contact their local electoral registration officers at any time of the year in order to have their name entered or deleted, effective from the next month. 11 " The provisional register is published every year for a period of at least 14 days, ending not later than 15 June. During that period, electors may claim their inclusion in the register. The corrected register remains in force for one year.

Registering Voters 73

the list of an unduly registered person, and the National Institute of Statistics and Economic Studies forwards to the mayors of communes the names of electors who have lost their right to vote. In the Philippines, at least every two years, the names of registered voters are checked by mail or house-to-house canvass. In those countries, however, the main initiative lies with individual citizens. In 34 countries, such as Argentina, Canada, Germany, and Israel, the state takes the initiative for registering new electors. It may do so in several ways. There may be a house-to-house canvassing of electors. State institutions may be obliged to transmit data on qualified electors. Nevertheless, electors may check whether they are duly registered and, if they are not, may request registration. Every European country but two (France and Portugal) has state-initiated registration. Asian countries are also more inclined to that system, whereas none of the countries located in Oceania does so. TIMING OF FIRST REGISTRATION A second issue, after individual versus state initiatives, concerns the period when the first registration of a qualified elector takes place. There are three possibilities for registration: continuous, at regular intervals, or before each election.3 In a majority of countries, such as Belgium, New Zealand, and Uruguay, registration is continuous. In 14 countries, such as Ireland, Jamaica, and Japan, it takes place at regular intervals. Only four countries - Germany, Malawi, the Philippines, and Sweden - conduct a new registration before each election. Arguments for and against each of these systems abound. Continuous registration does not restrict the registration period to a few days or weeks, thus allowing for constant updates and changes to the register. However, keeping the register up-to-date invariably drives up costs. Also, since the first registration does not occur in an election campaign, people may feel less concerned about the importance of getting registered. This last point is often mentioned as a strength of before-the-election registration. Electors who are aware of the upcoming election are more likely to get on the register. INCLUSIVENESS The most common concern vis-a-vis the registration process has to do with the completeness of the electoral register, particularly where reg-

74 Establishing the Rules of the Game

istration is not state-initiated. How to avoid high rates of unregistered electors? Legislators have come up with a number of solutions. The first is to make registration compulsory. We find this in 20 countries, including Australia, Brazil, and France. Only four countries in Europe fall into this category, as that continent favours state-initiated systems, which are less likely to be compulsory. A majority (55 per cent) of countries where electors must take the initiative for being registered have compulsory registration. As with compulsory voting, non-compliance with the law may be punished by a fine. But in some instances, as in France, registration is only a 'civic duty.' A relatively high proportion of countries (40 per cent) with compulsory registration do not have compulsory voting. Forcing electors to vote is probably seen as too drastic a requirement, whereas obliging them to register does not infringe on their decision to exercise or not their right to vote. The intended effect is indirect: compulsory registration arguably fosters higher political awareness, which in turn increases turnout. This combination is found in Benin, France, Guyana, Malawi, New Zealand, Papua New Guinea, Romania, and Spain. A second remedy to low rates of registration is for electoral authorities to make a door-to-door canvass or other similar inquiry, either for registering new electors or for verifying information on registered electors. In 16 countries, the law permits or requires such an inquiry that may take two forms: door-to-door canvassing or letters sent to electors. Here, the highly visible pattern is that of former British colonies following Britain's example, with 61 per cent of them having a door-todoor inquiry (see Table 3.2). A third solution is to allow qualified electors to register on election day.4 This means in effect that the electoral register 'closes' at the end of polling day. Opponents of this method argue that it increases the possibility of election fraud, as not much time is available to check the credentials of late applicants. This solution is available to all electors in seven countries. In Mali, only people who have moved because of a job transfer can take advantage of such an arrangement. In New Zealand, only those who turn 18 on or close to election day may register on polling day. In countries that do not permit registration on election day, the register closes at some earlier time. As to the specific closing date, some are more permissive than others. For instance, in New Zealand electors may register until the day before polling day. At the other extreme, the

Table 3.2 British colonialism, political rights, and registration

British colonialism Former British colonies

Other countries

Political rights Established democracies

Other countries

Total

State-initiated first registration

Continuous registration

Compulsory registration

House-to-house or other inquiry

Political parties involved

50% (n = 22)

75%

26% (n = 23)

61%* (n = 23)

33% (n = 21)

71%

57%

(n = 31)

(n = 30)

39% (n = 36)

3% (n = 32)

37% (n = 30)

71% (n = 21)

65%

32% (n = 22)

36% (n = 22)

29% (n = 21)

58% (n = 33)

59%

34% (n = 38)

24% (n = 34)

39% (n = 31)

63% (54 countries)

62%

33% (60 countries)

29% (56 countries)

35% (52 countries)

* Difference significant at .01 level (two-tailed f-test).

(n = 16)

(n = 20)

(n = 27)

(47 countries)

76 Establishing the Rules of the Game

electoral register in Argentina and Panama closes 165 days (i.e., 51/2 months) before polling day. Usually, countries where the legislative assembly may be dissolved tend to close the register at some point between the call of the election and polling day. This allows electors to register once they are aware that an election is imminent. In countries where the legislative assembly may not be dissolved (as is the case in most presidential systems), the date of the election is known well in advance. Authorities in those systems tend to close the register farther from election day, which explains the length of the interval in Argentina and Panama. In seven countries, the register is closed on a fixed date, independently of the election cycle. A fourth - but rarely employed - solution is to allow premature registration, which registers under-aged but otherwise qualified electors as if they were of voting age. This occurs in Australia, Guyana, Micronesia, New Zealand, and Spain. In these countries, under-aged electors may seek their inscription on the register.5 This arrangement allows additional opportunity for registration. THE ROLE OF POLITICAL PARTIES Do political parties play a role in the registration process? In a majority of countries (34), they do not. This is the best way of ensuring that the registration process is free of partisan meddling, provided of course that the civil service is not politicized. In 18 countries, political parties are involved; in five of them, party representatives run the registration process. The other approach, used in 16 countries, is for parties to nominate or appoint observers to the process.6 This may be seen as a compromise: registration is non-partisan, but parties may check the conduct of the process. FOUR CASE STUDIES We now briefly summarize the registration processes in four democracies: Germany, Canada, Spain, and Australia. These short descriptions highlight the great variety of registration procedures in use today. Germany The state initiates the first registration of German electors. The local authorities of each commune compile the electoral register before each

Registering Voters 77

election, using information included in each municipal register. When a polling district includes more than one commune or parts of various communes, the local authority of each commune compiles the electoral register for its own portion of the polling district. On the 35th day before the election, the local authority of the commune enters the names of qualified electors who reside in the commune, who are on a vessel at sea, or who are incarcerated in the commune. All other electors need to apply in order to be registered. Registration is not compulsory. Applications may be made until the 21st day before the election. Collective applications are authorized. Disabled electors may to this end rely on another person's assistance. Electors residing abroad may apply to the diplomatic and consular missions of Germany, to the federal returning officer, or to the constituency returning officer. In order to foster the awareness of electors residing abroad, public notices of registration procedures are advertised in German in both a daily and a weekly national newspaper, as well as in a regional newspaper. Electors who move to another commune between the 35th day and the 21st day before the election must apply to the commune to which they have moved in order to be registered. Those who move within the same commune between the 35th day and 21st day before the election remain on the register of the polling district in which they were entered on the qualifying date. Not later than the 21st day before the election, the local authority of the commune notifies every elector listed on the register by sending a postcard indicating the location of the polling station and the polling hours. From the 20th day to the 16th day before polling day, the public may examine the register. Anyone who deems it incorrect or incomplete may file an objection. Should it be obviously incorrect or incomplete, the local authority may also revise it on its own initiative. The register is closed sometime between one and three days before the election. Electors who wish to vote at another polling station or by mail need to apply to the local authority for a polling card no later than the second day before the election. Furthermore, no later than the 8th day before the election, the local authority requests from the management of institutions that will have special polling stations (hospitals, nursing homes, penal institutions) a register of electors who are inmates of the institution or work there. It sends polling cards to electors there so that they may cast a vote. Political parties play no direct role in the registration process.

78 Establishing the Rules of the Game Canada

In 1996, Canada changed its registration process from a prior-to-theelection enumeration of electors to continuous registration. As in Germany, the first registration is state-initiated. The chief electoral officer (instead of local authorities) establishes and maintains a register of Canadians who are qualified to vote. Unlike in Germany, the register i not prepared at election time but is rather updated continuously on the basis of information sent by various federal agencies (such as the Canada Customs and Revenue Agency) and by electors.7 There are agreements with provincial agencies for the transmission of data. For instance, changes in the address of electors who renew their driver's license (a provincial responsibility) are forwarded to the chief electoral officer. It is also possible to update the register by examining provincial electoral registers. Before entering new electors on the register, the chief electoral officer asks them if they wish to be registered.8 No such permission is required in Germany. Unless they reply by the affirmative, new electors will not be registered. At any time, any person may apply for registration by the chief electoral officer. Canadian citizens residing temporarily abroad may apply for registration. Their application must be received by the chief electoral officer no later than the 6th day before the election. They must vote, by mail, in the electoral district of their last place of residence in the country. The same deadline applies to residents of Canada who choose to vote by mail. Forthwith after the issue of the writ for election, the chief electoral officer prepares a preliminary list of electors for each electoral district and transmits it to the returning officer of each district. As soon as possible after the issue of the writ, the chief electoral officer determines the commencement date for the revision of the preliminary list. The revision ends on the 6th day before the election. For the purpose of revising the preliminary list, the returning officer in each electoral district appoints revising officers from a list of names submitted by the registered political parties whose candidates came first and second in the latest election in the district. Party representatives thus participate directly in the registration process. During the revision period, electors who are disqualified are removed from the register. Unregistered electors may apply for registration. Application for registration can be made by proxy. Electors who move during the

Registering Voters 79

revision period may notify the revising officers and vote in their new district. No later than the 24th day before the election, each returning officer sends a notice of confirmation of registration to every elector, except members of the armed forces, electors residing abroad, prison inmates, and electors who wish to vote by mail. Not later than the 5th day before the election, each returning officer sends a similar notice to electors who registered during the revision period. The notice of confirmation of registration indicates the elector's polling station, the location of the polling station and of advance polling stations, and the date and hours when they will be open. Finally, unlike in Germany, an elector who has not registered may do so on polling day and may then cast a ballot. Spain As in both Germany and Canada, so in Spain the state initiates the first registration of electors. The Electoral Register Office, within the National Statistical Institute, is in charge of compiling the electoral register under the direction and supervision of the Central Electoral Commission. The Electoral Register Office has delegations in each of the 50 provinces, and local councils and consulates must co-operate in the process of registration. Unlike Germany and Canada, Spain provides for compulsory registration. There are, however, no fines for non-registration. Local councils deal with the registration of persons residing within their boundaries. Spaniards residing abroad may register at diplomatic missions in their country of residence. The registration process is not continuous as in Canada but takes place monthly. Every month, local councils forward to the provincial delegations of the Electoral Register Office a list of new registrations, deregistrations, and changes of address within their area of jurisdiction. Also monthly, the Public Record Office and the Record Office for Persons Detained in Penal Institutions and Persons at Large notify the provincial delegations of the Electoral Register Office of any civil or penal circumstance that may affect registration in the electoral register. The Electoral Register Office then carries out a monthly update of the register on the basis of the data received before the first day of every month. Every year, in the first update of the register, the names of those who will reach voting age between 1 January and 31 December are for-

80 Establishing the Rules of the Game

warded by the local councils to the provincial delegations of the Electoral Register Office. For the purpose of an election, the electoral register in force is the one closed on the first day of the month preceding the date on which notice for the election is given. Local councils and consulates then open the registers for public inspection in their respective areas of jurisdiction for a period of eight days, starting the 6th day following the notice of election. During this period, any person may file a claim concerning his or her registration data, either to the provincial delegation of the Electoral Register Office or to the local council. Claims may not be made after that eight-day period. As there are between 54 and 60 days between the notice for election and polling day, it is impossible to register in the last six weeks before the election. This contrasts sharply with Canadian legislation, which allows registration up to and including election day, and with Germany, where electors may register up to (but excluding) election day. After the 17th day following the notice of election, the Electoral Register Office forwards to all electors a registration card including the updated data on their registration, as well as the location of the polling place where they should vote. As in Germany and contrary to Canada, local authorities participate closely in the process and parties are not directly involved. Australia Unlike the three countries described above, Australia does not have a state-initiated first registration. Electors must apply to be entered on the electoral register. Registration is, however, compulsory; people who do not register within 21 days after becoming qualified electors may be fined up to $50 (about U.S. $27). Applications are received on a continuous basis. The registrar-general also forwards to each divisional returning officer, every month, a list of names of all persons aged 17 or older whose deaths have been registered during the previous month. The controller-general of prisons does the same for people serving a sentence of five years or more. Names of disqualified electors are struck off the register. The Australian Electoral Commission also conducts a review of the register by sending targeted mailings to rural residential addresses where no electors are currently registered. In urban areas, officers of the commission conduct a door-to-door inquiry. This review, usually

Registering Voters 81

conducted every two years, adds a small but significant number of revisions to the roll. Persons aged 17 may apply for registration, as long as they otherwise qualify to be registered. But unlike in Spain, the process is not state-initiated. Australians living abroad must apply to the divisional returning officer if they wish to be treated as eligible overseas electors. They must do so within 3 months before their departure or within 1 year afterwards. Their vote is counted in the division (electoral district) in which they were registered before leaving the country. An elector who does not have a permanent address anywhere in the country may apply to be treated as an itinerant elector. These people are usually registered for the last subdivision in which they were registered. Changes to the electoral register may be made until the 7th day following the issue of the writ. The register is closed no less than 26 days and no more than 51 days before polling day. Australia has the earliest closing date for the electoral register among the four countries examined here. But as in Spain, registration there is compulsory. Perhaps there is no urge to allow registration near polling day, as electors are required by law to register earlier. As in Canada, local authorities are not involved, and, as in Germany and Spain, parties are excluded from the process. CONCLUSION In most countries, the state takes the initiative in constituting a register listing the names and particulars of all those who have the right to vote; in a significant minority (20), citizens apply to have their name on the electoral list. In most countries, the electoral register is permanent; only four conduct a new registration for each election. In most countries, registration is voluntary, although 20 have compulsory registration. Most countries revise the electoral register during the election campaign; only in seven does the law allow for registration up to and including election day. Finally, a clear majority completely exclude political parties from the compiling and updating of the register or involve them only as observers. We found few significant cross-national patterns. Former British colonies are more prone to rely on door-to-door inquiries. But we noticed little systematic difference between established and non-established democracies, which suggests that there may be little consensus on con-

82 Establishing the Rules of the Game

struction of electoral registers. Most notably, the seven countries that allow registration up to and including election day are a very diverse lot: Benin, Canada, Hungary, Lithuania, Poland, Slovakia, and Uruguay. Established democracies do not seem to make it easier for qualified electors to have their names on the electoral register.

Chapter Four

Who Conducts the Election?

Free and fair elections cannot take place without a legitimate and transparent electoral administration. In this chapter, we examine the central bodies for election administration in democracies. More precisely, we look at how the election laws establish their appointment by government, legislature, judiciary, or some combination of these at their composition - commission, public official, or minister - and at their term of office (Table 4.1). We also describe briefly the central administration in two countries, Australia and Canada. Three Models for Authority There are three basic approaches in constituting central election authorities. One is to appoint a commission that has a final say over administrative matters pertaining to elections. Such a commission involves more than one person in the decision-making process. And having these persons nominated by different entities ensures wider representation. The effectiveness of this formula depends on how individuals representing different and sometimes conflicting interests work together. The second approach is to appoint a single public official to supervise the electoral process. This method probably increases efficiency, preventing wrangling among different people. However, it is not easy to find a person whom all parties will trust, and partisanship may taint the nomination process and the official's conduct. The third approach is for a government minister to be in charge of the conduct of elections. This method makes the person in charge directly responsible to the legislature and, ultimately, to the electorate. However, the chief official who supervises the political process also happens usually to be one of the contenders.

Table 4.1 Central electoral administration Country

Nature of electoral authority

Method of appointment

Term of office

Argentina

4-member commission (National Electoral Chamber)

NA

NA

Australia

3-member commission (Australian Electoral Commission)

Government and judiciary: Chair appointed by governor general from a list of three eligible judges submitted by chief judge of Federal Court. Electoral commissioner and other member appointed by governor general.

7 years

Bahamas

Public official (parliamentary commissioner)

Other: Appointed by governor general, acting on advice of Public Service Commission, which consists of three to five members appointed by governor general acting on recommendation of the prime minister after consultation with leader of opposition.

NA

Bangladesh

Commission (Election Commission)

Government: Members appointed by president, acting on advice of prime minister.

5 years

Barbados

5-member commission (Electoral and Boundaries Commission for Barbados)

Government and opposition: Chair and two other members appointed by governor general, acting on recommendation of prime minister after consultation with leader of opposition. Deputy chair and one other member appointed by governor general, acting on recommendation of leader of opposition after consultation with prime minister.

5 years

Belgium

Minister of interior

Government

Legislative term

Belize

5-member commission (Election and Boundaries Commission)

Government and opposition: Chair and two members appointed by governor general, acting on advice of prime minister given after consultation with leader of opposition. Remaining two members appointed by governor general, acting on advice of prime minister given after concurrence of leader of opposition.

5 years

Table 4.1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Benin

23-member commission (National Autonomous Electoral Commission)

Other: Fifteen members chosen by National Assembly; three chosen by government; four chosen by the College des Magistrals du siege; one chosen by Commission of Human Rights.

NA

Bolivia

5-member commission (National Electoral Court)

Government and legislature: One member appointed by president; four members appointed by the National Congress.

4 years (president)

Brazil

7-member commission (Superior Electoral Court)

Government and judiciary: Two members appointed by president from among names submitted by Supreme Federal Court; three members chosen by Supreme Federal Court; two members chosen by Superior Justice Court.

2 years

Bulgaria

24-member commission (Central Election Commission)

Government and parties: Members appointed by president. However, nominations must be countersigned by prime minister, after having consulted ALL opposition parties. Commission consists of chair, deputy chair, secretary, and members, including representatives of political parties as well as people with no party affiliation. Majority must be lawyers by education.

Legislative term

Canada

Public official (chief electoral officer)

Legislature: Appointed by House of Commons.

Until 65 years old

Cape Verde

5-member commission (National Election Commission)

Legislature: Members appointed by National Assembly (two-thirds majority required).

6 years

Chile

Public official (director of Electoral Service)

Government and senate: Appointed by president with the support of a majority of senators.

NA

Table 4.1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Costa Rica

3-member commission (Supreme Tribunal of Elections)

Judiciary: Members appointed by Supreme Court.

Legislative term

Cyprus

Public official (general returning officer)

Government: Appointed by minister of interior.

Appointed at each election

Czech Republic

Commission (Central Election Commission)

Government and parties: Each party or coalition that submits a list of candidates has right to be represented. Minister of interior appoints members of commission from among names submitted by parties and coalitions.

Legislative term

Denmark

Minister of the interior

Government

Legislative term

Ecuador

7-member commission (Supreme Electoral Tribunal)

Legislature: Members appointed by National Congress.

4 years

Estonia

7-member commission (National Election Committee)

Other: One member is a judge of a court of first instance, appointed by chief justice of Supreme Court. One member is a judge of a court of appeal, appointed by chief justice of Supreme Court. One member is adviser to legal chancellor, appointed by legal chancellor. One member is official of State Audit Office, appointed by auditor general. One member is deputy public prosecutor, appointed by public prosecutor. One member is official of Office of the Riigikogu, appointed by director of the Office of Riigikogu. One member is official of State Chancellery, appointed by director general of State Chancellery.

4 years

France

Minister of the interior

Government

Legislative term

Germany

9-member commission and public official (federal returning officer)

Other (Commission): Eight members are appointed by federal returning officer. Normally chosen from among names proposed by parties. Federal returning officer chairs commission.

Commission: legislative term Public official: indefinitely

Table 4.1

( Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Guyana

7-member commission (Election Commission)

Government and opposition: Chair a person who holds or has held office as judge of court, appointed by the president from list of six persons submitted by the minority leader after consultation with political parties represented in National Assembly, other than party to which president belongs. Three members appointed by president alone; three members appointed by president, acting on advice of minority leader after consulting political parties represented in National Assembly.

5 years

Hungary

5-member (minimum) commission (National Electoral Committee)

Legislature: Parliament elects five members and necessary number of additional members; minister of home affairs makes proposals taking into consideration proposals of parties.

NA

India

3-member commission (Election Commission)

Government: Members appointed by president acting on advice of Council of Ministers.

6 years or until 65 years of age, whichever is earlier

Ireland

Minister of the environment

Government

Legislative term

Israel

30-member commission (Central Elections Committee)

Legislature: Each party group with four or more members in Knesset has one seat for every four members. Remaining seats distributed among party groups that took part in that distribution in order of size of member surpluses. Every party group that has not secured representation under this has one member on the committee.

Legislative term

Italy

5-member commission (National Central Electoral Office)

Government: Members appointed by prime minister.

NA

Jamaica

Public official (chief electoral officer)

Government: Appointed by governor general.

NA

Table 4.1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Japan

NA

NA

NA

Latvia

9-member commission (Central Electoral Commission)

Legislature and judiciary: Chair and seven members elected by Saeima; one member elected by the Supreme Court from among its judges.

Legislative term

Lithuania

12-member (minimum) commission (Central Electoral Committee)

Other: Chair appointed by Seimas (Parliament). Three members, having higher legal education, chosen by lot from six candidates proposed by Lawyers' Society. Other members proposed by parties and political organizations that have at least one Seimas member in the multi-candidate electoral area. These latter members must come from lists of candidates presented at previous election. Minister of justice and the Lithuanian Lawyers' Society may propose more candidates for commission. In all cases, members proposed by minister of justice and Lawyers' Society must make up more than 1/2 of committee. Otherwise, committee is enlarged by drawing lots equally from candidates proposed by minister of justice and Lawyers' Society, whose member must be twice as many as necessary for additional appointment of members to committee.

4 years

Luxembourg

Minister of the interior

Government

Legislative term

Madagascar

7-member commission (National Electoral Commission)

Other: One member appointed by president. Two members appointed by prime minister. One member appointed by president (speaker) of National Assembly. One member appointed by president of Senate. One member appointed by president of the bar. One member appointed by president of journalists' order (guild).

NA

Table 4 1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Malawi

7-member (minimum) commission (Electoral Commission)

Government and parties: Members appointed by president. However, for first general election (1994), members of commission were required to be appointed by president on recommendation of National Consultative Council. Council consisted of representatives of all political parties registered under Political Parties Act 1993 by 30 November 1993. Each registered political party was entitled to designate not more than seven representatives to council. Also, constitution states chair is judge nominated by Judicial Service Commission.

NA

Mali

30-member commission (Independent National Electoral Commission)

Other: Eight members appointed by government. Seven members appointed by political parties of parliamentary majority. Seven members appointed by political parties of parliamentary opposition. One member appointed by the office of the 'AMUPI.' One member appointed by Catholic church. One member appointed by the 'AGEMPEM' (Protestant group). One member appointed by bar. One member appointed by Mali's Human Rights Association. One member appointed by Mali's delegation to the African League of Human Rights. One member appointed by Autonomous Union of the Magistracy. One member appointed by the CAFO (women's group).

Appointed at each election

Malta

5-member (minimum) commission

Government and opposition: Members appointed by president, acting on advice of prime minister, given after consultation with leader of opposition.

3 years

Micronesia

Public official (national election director)

Legislature: Appointed by president, with the advice and consent of Congress.

NA

Mongolia

1 5-member commission (General Election Committee)

Legislature: Members appointed by Parliament.

5 years

Table 4.1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Namibia

5-member commission (Electoral Commission)

Legislature: President, with approval, by resolution (majority of votes cast), of National Assembly, appoints judge or former judge of Supreme Court or High Court to chair commission and four other members.

5 years

Netherlands

7-member commission (Electoral Council)

Government: Members appointed, by royal decree, by government.

4 years

New Zealand

Public official (chief electoral officer)

Other: Appointment by secretary of justice.

NA

Panama

3-member commission (Electoral Tribunal)

Government, legislature, and judiciary: Executive, legislative, and judicial (Supreme Court) branches of government appoint one member each.

10 years

Papua New Guinea

Public official (Electoral Commissioner)

Other: Appointed by Electoral Commission Appointments Committee. Committee consists of prime minister, or a minister nominated by him or her, who is chair; chair and deputy chair of appropriate Permanent Parliamentary Committee; and chair of Public Services Commission.

6 years

Philippines

7-member commission (Commission on Elections)

Government and legislature: Members appointed by president, with consent of Commission on Appointments. Commission on Appointments consists of president of Senate, 1 2 senators, and 12 members of House of Representatives, elected by each house on basis of proportional representation.

7 years

Table 4.1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Poland

9-member commission (National Electoral Commission)

Judiciary: Three members judges of Constitutional Tribunal, designated by president of Constitutional Tribunal. Three members judges of Supreme Court, designated by first president of Supreme Court. Three members judges of the Supreme Administrative Court, designated by president of Supreme Administrative Court. All nine members officially appointed by president of republic, which is merely a formality.

Judges' term

Portugal

9-member commission (National Elections Commission)

Government, legislature, and judiciary: Chair a judge appointed by superior magistracy council. Five members appointed by Assembly of the Republic, each one from lists submitted by 5 parties with most seats. Three members (specialists) appointed by minister of interior, minister of foreign affairs, and minister of communications.

Legislative term

Romania

23-member commission (Central Electoral Bureau)

Judiciary and parties: Seven members judges of Supreme Court appointed by lot by president of Supreme Court. Sixteen members party representatives (procedure NA) .

Appointed at each election

St Lucia

3-member commission (Electoral Commission)

Government and opposition: Chair appointed by governor general, acting on his or her own deliberate judgment. One member appointed by governor general, acting on advice of prime minister. One member appointed by governor general, acting on advice of leader of opposition.

Legislative term

St Vincent and the Grenadines

Public official (supervisor of elections)

Other: Appointed by Public Service Commission after consultation with prime minister.

Legislative term

Samoa

Public official

Other: Appointed by Public Service Commission.

NA

Table 4.1

(Continued)

Country

Nature of electoral authority

Method of appointment

Term of office

Sao Tome and Principe

Commission (National Electoral Commission)

Legislature: Chair appointed by National Assembly. Each party may appoint representative (procedure NA).

NA

Slovakia

Commission (Slovakian Electoral Commission)

Parties: Each party that submitted list of candidates in at least one region of country can delegate two members to commission.

Appointed at each election

Slovenia

6-member commission (State Electoral Board)

Legislature: President of board is judge of the Supreme Court, appointed-by National Assembly. Two members of legal profession appointed by National Assembly. Three members appointed on proposal of delegate clubs of three parties with highest number of seats in assembly.

4 years

South Africa

5-member commission (Electoral Commission)

Other: A panel - of president of Constitutional Court, a representative of Human Rights Commission, a representative on Human Rights Commission, a representative of Commission on Gender Equality, and public prosecutor - submits list of eight persons to National Assembly. A National Assembly committee, composed proportionally of members of all parties represented in Assembly, nominates five persons from list. Assembly recommends appointments by resolution adopted by majority of its members/President appoints members of commission on recommendation of National Assembly.

7 years

Spain

13-member commission (Central Electoral Commission)

Other: Eight members appointed by General Council of Judicature, by lot, from among Supreme Court judges. Five members, from among professors of law, political science, and sociology, in active service, appointed through joint proposal of political parties, federations, coalitions, or groups of electors with representation in Congress.

Legislative term

Sweden

National Tax Board

Members public servants appointd by government.

NA

Table 4.1

(Concluded))

Country

Nature of electoral authority

Method of appointment

Term of office

Taiwan

1 2- to 20-member commission (Central Election Commission

Government: Members appointed by president on recommendation of Executive Yuan (cabinet). Also, there must be non-partisan commissioners. Number of members from same party cannot exceed 2/5 of total number of members of Commission.

NA

Trinidad and Tobago

3- to 5-member commission (Elections and Boundaries Commission)

Government and opposition: Members appointed by president, after consulting prime minister and leader of opposition.

5 years

United Kingdom

Minister (Secretary of state for home affairs)

Government

Legislative term

Uruguay

9-member commission (Electoral Court)

Legislature: Five members elected by General Assembly by two-thirds' majority vote. Four members elected by General Assembly, by a system of double simultaneous vote (simple majority).

NA

Vanuatu

3-member commission (Electoral Commission) and public official (secretary of Electoral Bureau)

Other (commission): Members of commission appointed by president, on advice of Judicial Service Commission, which consists of minister for justice, president of Public Service Commission, judge appointed by president of republic, and representative of the National Council of Chiefs, appointed by Council.

Commission: 3 years

Government (public official). Secretary of Electoral Bureau appointed by minister of interior.

Secretary: NA

Legislature: Members appointed by two-thirds' vote in jointchamber resolution.*

5 years

Venezuela

7-member commission (National Electoral Council)

NA = Not available. * The Venezuelian Senate was abolished in 1999.

94 Establishing the Rules of the Game Term of Office

The term of office of election authorities also varies from country to country. In 23 countries, the term is a fixed number of years - on average, five years. In four countries, the mandate is renewed before each election. In one, the mandate expires after six years or when the public official reaches the age of 65, whichever comes earlier. In just one country does the mandate automatically expire when the official turns 65. In another, the commission's members, who are judges, must step down when their term as judges ends. Finally, 16 countries make the term of office coterminous with that of the legislature; following an election, the electoral authority's mandate expires and a new authority is appointed. Electoral Commissions

An electoral commission is the central electoral authority in 43 of our countries (69 per cent).1 Table 4.1 shows numerous methods for nominating and appointing members; these may involve the government (with or without the opposition), the legislature, the judiciary, or some combination of these. In six countries (Bangladesh, India, Italy, Malta, the Netherlands, and Taiwan), the government nominates and appoints members of the commission. It can do so unilaterally, although in Malta the prime minister must consult the leader of the opposition. Other countries also involve the opposition (for example, Barbados), the judiciary (for instance, Australia and Brazil), the assembly (for example, Bolivia and the Philippines), other political parties (for instance, the Czech Republic), or a combination of those entities (for example, Panama). Another common method is to have the legislative assembly appoint all members of the commission, as occurs in nine countries. In two of these - Cape Verde and Uruguay - a qualified majority (2/3) is required.2 In Ecuador, Hungary, Mongolia, Namibia, and Sao Tome, a simple majority is enough. In Hungary, the minister of home affairs submits propositions to the legislature, taking into consideration the proposals of parties. In Namibia, the chair of the five-member Electoral Commission must be either a judge or a former judge of the Supreme Court or the High Court; nevertheless, the decision lies with the

Who Conducts the Election? 95

assembly, deciding by a simple majority vote. In Israel, members of the Central Elections Committee are appointed by parties having at least four seats in the Knesset.3 In Venezuela, a joint session of the legislature elects the seven members of the National Electoral Council by a two-thirds majority.4 The assembly's prerogative may be shared with the judicial branch (as in Latvia). Political parties may also directly appoint electoral commissions (as in Slovakia). The judicial branch of government can also be the main appointing body (for example, in Costa Rica and Poland). In Romania, seven members of the Central Electoral Bureau are judges of the Supreme Court chosen by lot by the president of the Supreme Court. However, the other 16 members are party representatives. Finally, in a number of countries, electoral commissions are appointed by various combinations of entities (as happens in Germany, Portugal, and Spain). Perhaps the most inclusive nomination processes are found in Benin and Mali. In the former, the National Autonomous Electoral Commission comprises 17 members - seven chosen by the government, seven by the national assembly, two by the Commission of Human Rights, and one by the 'Magistral du siege de 1'Ordre judiciaire elu par I'Assemblee Generate des Magistrats' (a magistrate selected by all other magistrates). In Mali, the government, the parliamentary majority, the opposition, the Catholic and Protestant churches, the judicial bar, two human rights associations, a women's group, and the union of magistrates all concur in the appointment of the 30-member Independent National Electoral Commission. Established democracies and former British colonies are less likely than other countries to empower a commission with the central administration of elections (Table 4.2). It is plausible that weak democracies create commissions in order to involve diverse factions in the process, as is the case in Benin and Mali. As for former British colonies, only Ireland follows Britain's lead by empowering a minister with the administration of elections. It will be interesting to follow developments in Britain, which in 2000 set up an Electoral Commission to oversee some aspects of the electoral process. The Swedish case is unique, with elections administered by the National Tax Board - a 10-member body appointed by the government - which also serves as the main taxation body and maintains the population register.

96

Establishing the Rules of the Game

Table 4.2 British colonialism, political rights, and the presence of an electoral commission Electoral commission conducts election British colonialism Former British colonies Other countries Political rights Established democracies

57%* (n = 23) 79% (n = 23) 50%t n = 22)

Other countries

80% (n = 40)

Total

69% (62 countries)

* Difference significant at 0.10 level (two-tailed f-tests). f Difference significant at 0.05 level (two-tailed Mests).

A Single Official A public official is solely in charge of the supervision of elections in 10 countries.5 In Cyprus and Jamaica the government appoints the person on its own initiative. In Canada and Micronesia, the legislative assembly nominates the public official. In Micronesia, the appointment is made by the president, with the advice and consent of Congress. In Chile, the president's appointee must be approved by a majority of senators. In Bahamas, the governor general (in effect the cabinet) appoints the Parliamentary Commissioner, acting on the advice of the Public Service Commission, which has three to five members chosen by the governor general, acting on the recommendation of the prime minister after consultation with the leader of the opposition. In New Zealand, the chief electoral officer is named by the secretary for justice, who is appointed by the State Service Commission, made up of four members appointed by the government. In Samoa, the chief electoral officer is appointed by the Public Service Commission, which is com-

Who Conducts the Election? 97

posed of not more than three persons appointed by the head of state, acting on the advice of the prime minister. A Minister Six countries, all in Europe, have a minister in charge of the conduct of elections: Belgium, Denmark, France, Luxembourg, Ireland, and the United Kingdom. In the first four, the minister of interior is in charge of electoral matters; in Ireland, the minister for the environment and local government; and in Britain, the secretary of state for home affairs. TWO CASE STUDIES In order to explicate the intricacies of the appointment process and the nature of the electoral administration's prerogatives, we now offer a summary of the central electoral administration in Canada (public official) and Australia (commission). Canada

In Canada, the conduct of elections in each federal constituency is under the supervision of a returning officer appointed by the cabinet. The necessity of having an impartial senior official supervising the work of returning officers led Parliament to create the office of chief electoral officer of Canada in 1920. The chief electoral officer (CEO) is appointed by a resolution of the House of Commons (the elected chamber of the Canadian Parliament) that requires a simple majority. The governor general may dismiss the CEO, but only for cause and at the request of both Houses of Parliament, including the unelected Senate. The same procedure is used for dismissing federally appointed judges and is deemed to guarantee their independence from the government. His or her salary is equal to that of a puisne judge of the Federal Court. The CEO is expected to hold this position until the age of 65. Since 1920, Canada has had only five CEOs, including Jean-Marc Hamel, who held the office from 1966 to 1990, and the present incumbent, Jean-Pierre Kingsley. The average tenure of Kingsley's predecessors was 17 years and 6 months. Although the appointment procedure would theoretically empower a majority government to impose its nominee on opposition parties, in practice all appointments have been agreed to by other parties and have not occasioned a vote in the House of Commons. Some

98 Establishing the Rules of the Game

appointments have even been seconded by the leader of the opposition, and the most recent one was supported by the two opposition parties. No cabinet minister has direct authority over the CEO, who is answerable to Parliament as a whole. However, the CEO usually communicates with the cabinet through a minister designated for that purpose, generally the government's House leader. The CEO supervises the work of an assistant CEO appointed by the cabinet and of a staff of 100, which is expanded during election campaigns. Duties include the general direction and supervision of the administrative conduct of elections and the enforcement by all election officers of fairness, impartiality, and compliance with the Elections Act The CEO may also implement public education and information programs to make the electoral process better known to the public. The office of the CEO is known as Elections Canada. In emergency situations during an election campaign when Parliament is not in session, the CEO may adapt the provisions of the act to the execution of its intent, to such extent as seems necessary to meet the exigencies of the situation. This includes the power to extend the time for doing any act and to increase the number of election officers or polling stations. Since 1951, following disastrous floods in Manitoba, the CEO has been empowered to certify that, by reason of a flood, fire or other disaster, it is impracticable to carry out the provisions of the act in an electoral district where an election has been called. In case such a certificate has been issued, the cabinet may order postponement of the election in these districts.6 Ensuring compliance with and enforcement of the act is the duty of the commissioner of Canadian elections, appointed by the CEO. If this official believes that an electoral offence has been committed and that the public interest requires it, he or she may institute or cause to be instituted a prosecution for the offence. No prosecution for an electoral offence may be instituted by a person other than the commissioner without the commissioner's prior written consent. The Canadian system of election management is a blend of approaches. At the highest level, strict political neutrality is ensured by the position of the CEO and the fact that the staff is recruited according to the usual procedures of the public service. At the local level, polling officials are appointed by the parties that finished first and second in the constituency. At the middle level - that of returning officers appointment is by the cabinet and may be affected by party consider-

Who Conducts the Election? 99

ations, but these officers are expected to work under the instructions of the CEO. Further, they are prohibited from engaging in politically partisan conduct, including contributing to a candidate or a party or being an employee of or holding a position in a party. While the staff of Elections Canada has been greatly expanded since its creation, the basic underlying principles have remained remarkably stable over the years. The chief electoral officer of Canada has no responsibility whatsoever with regards to provincial elections. All provinces and territories have their own autonomous CEOs. Australia Since 1984, federal elections and referendums in Australia have been conducted under the authority of a three-member body known as the Australian Electoral Commission (AEC). Prior to that date, this role was fulfilled by the Australian Electoral Office, under the authority of successive government departments. Distinct bodies oversee the conduct of elections in each of Australia's states and territories, under specific state or territorial legislation. However, all states and territories have a joint roll (electoral register) agreement with the AEC. Under this arrangement, electors' details on the AEC's roll-management system can be used by other electoral bodies to produce rolls for local and state-government elections and byelections. This method reduces the costs of electoral registers and means that electors have to complete only one form to enrol for elections at the three levels of government. The AEC consists of a chair (a judge or a retired judge of the Federal Court), the electoral commissioner, and one other non-judicial member (always, so far, the Australian statistician). Of these three people, only the electoral commissioner, described as the chief executive officer of the commission, has the powers of a deputy minister and serves full time, while the chair and the third, non-judicial member both hold office part time. In practice the electoral commissioner is in charge of day-to-day operations. All three members of the commission are appointed by the governor general (acting on the advice of the prime minister) for renewable terms of up to 7 years. No formal consultation with the opposition precedes appointments. The chair is selected from among a list of three judges or former judges of the Federal Court submitted by the court's chief justice. The non-judicial part-time member must be a deputy minister.

100 Establishing the Rules of the Game

In practice, the commission meets rarely and, when it does, deals with judicial or legislative matters. At meetings, decisions are taken by a majority of votes of members of the commission present and voting. The higher echelons of the commission include a deputy electoral commissioner and an Australian electoral officer (AEO) for each of the six states and for the Northern Territory. All these officials are appointed by the governor general for renewable terms of up to 7 years and must retire at 65. AEOs manage electoral activities within their state or territory, including statewide elections for the federal Senate. The AEC has a central office in Canberra and a head office in each state capital. The staff of the commission is appointed under the Public Service Act. For elections to the House of Representatives, Australia is divided into 148 electoral districts (known formally as divisions and informally as electorates). For each division, the AEC appoints a divisional returning officer (DRO), who is responsible for electoral administration within that division - in particularly maintenance of the electoral roll and preparations for the next electoral event. The AEC conducts elections to the House of Representatives and Senate, by-elections, or referendums and maintains and keeps up-todate the Commonwealth Electoral Roll. At least every two years the AEC conducts a roll-checking program called an Electoral Roll Review, in which AEC staff contact all households to check the enrolment details of electors. It publishes the detailed results of elections. It has the duty of promoting public awareness of electoral and parliamentary matters by way of education and information programs. The AEC is responsible for the implementation of the legislation governing election funding and financial disclosure. Also, it advises Parliament and the government on electoral matters. As voting in Australia is compulsory, it is up to the electoral commissioner to institute proceedings against individuals who have failed to vote and have not been otherwise excused for their absence. In addition, the AEC conducts up to 700 industrial (i.e., trade-union) elections each year and elections for the Aboriginal and Torres Strait Islander Commission (ATSIC). Union elections are held every three years for regional councils, zone representatives, and regional council office holders. The AEC is empowered, subject to the approval of the minister for foreign affairs and trade, to provide assistance in the conduct of elections and referendums to authorities of foreign countries or foreign organisations.

Who Conducts the Election? 101

So far, it is reported that the commission's performance has not attracted serious criticism from the opposition parties, and individual members of Parliament have complained very rarely. Following every election, the all-party Joint Standing Committee of Parliament on Electoral Matters reviews electoral issues, which provides members with an opportunity to voice concerns. CONCLUSION An electoral commission has been instituted in a majority of countries. It prevents the centralization of the conduct of elections in one individual's hands. In practice, many countries whose elections are administered by a public official or a minister have been successful in preventing fraud or other irregularities. As we have seen in this chapter, there are almost as many appointment procedures as there are countries. As the case studies have shown, political appointments do not necessarily result in elections being conducted in partisan ways. Political culture in the country is as important as the formal process. Emerging democracies wishing to set up a neutral administration should remember that there is no single magic recipe to ensure free and fair administration of elections.

Chapter Five

How People Vote

Once electors are included in the electoral register and candidates have been nominated, the electoral administration is ready to organize what is regarded as the most important and visible democratic endeavour: polling day. Each year, millions of electors go to the polls to elect presidents and those representatives who will sit in parliaments and enact legislation. The need to ensure efficiency at the polls requires logistical efforts that are carefully organized and supervised by the electoral administration. This chapter provides an overview of the electoral machinery on polling day. The first section looks at conditions on polling day. The second examines the types of ballots used in legislative and presidential elections around the world. The third describes the various special voting procedures that facilitate the act of voting. This chapter offers an overview of the plethora of ways in which voting is conducted in democracies. POLLING DAY This section addresses seven questions about conditions on polling day. • • • • • • •

On how many days does the election take place? Is polling day a weekday or a rest day? Must employers provide time off for voting? For how many hours are polling stations opened? How many days separate rounds of voting? Who are the polling officials? Need electors provide identification?

Table 5.1 provides a summary of these various polling-day indicators.

Table 5.1 Polling day

Country

Number of polling day(s)

Polling day: weekday or rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Argentina

1

Rest day

Yes

10

Maximum 30 days Presiding officer and assistants appointed by (presidential National Electoral Chamber, elections) Party observers may be present.

Yes

Australia

1

Rest day

Yes

10

Not applicable

No

Polling officials: nomination procedure and presence of party observers

Presiding officer, deputies, and assistants appointed by Electoral Commission. Party observers may be present.

Identification document required

Bahamas 1 Weekday Yes 1 0 Not applicable Presiding officer (officer of department of Yes parliamentary commissioner) and assistants (as far as practicable public officers) appointed by parliamentary commissioner. Party observers may be present. Bangladesh

1

Weekday

No

8

Not applicable

Presiding officer and assistants appointed by relevant returning officer. Party observers may be present.

Yes

Barbados

1

Weekday

Yes

12

Not applicable

Presiding officer appointed by Commission, Yes after consultation with supervisor of elections; poll clerks appointed by supervisor of elections. Party observers may be present.

Belgium

1

Rest day

No

5

Not applicable

Presiding officer appointed by relevant canton Yes office president. In each polling station four scrutineers appointed by presiding officer, from among literate voters of at least 30 years of age. A secretary (without a deliberative vote) is appointed by presiding officer. Party observers may be present.

Table 5.1

Country

(Continued) Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Polling officials: nomination procedure and presence of party observers

Identification document required

Belize 1 Weekday Yes 1 1 Not applicable Presiding officer and assistants appointed Yes by chief elections officer, with approval of Elections and Boundaries Commission, from within public service. Party observers may be present. Benin

1

Rest day

No

10

15 days (presidential elections)

Presiding officer and two scrutineers appointed Yes by National Autonomous Electoral Commission, Party observers may be present.

Bolivia

1

Rest day

Yes

8

Not applicable

Five members of polling station appointed by relevant Departmental Electoral Court. Party observers may be present.

Yes

Brazil

1

Rest day

Yes

9

21 days (presidential elections)*

Presiding officer, assistants, and secretaries appointed by electoral judges, Party observers may be present.

Yes

Bulgaria

1

Rest day

No

13

7 days (presidential elections)

Members of Sectional Election Commission Yes (chair, deputy chair, secretary, and three to seven members) appointed by relevant Municipal Election Commission (itself appointed by Central Election Commission, after consultation with political parties.) Party observers may be present.

Table 5.1

( Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Canada

1

Weekday

Yes

12

Not applicable

Deputy returning officer appointed by returning No officer from list of names supplied by candidate of party whose candidate finished first in electoral district in last election. Poll clerk appointed by returning officer from list supplied by candidate of party whose candidate finished second in electoral district in last election. Party observers may be present.

Cape Verde

1

Rest day

Yes

11

14 days (presidential elections)*

Presiding officer, secretary, and two scrutineers appointed by relevant municipal assembly, Party observers may be present.

No

Chile

1

Weekday declared holiday (presidential election: rest day)

Yes

9

35 days (presidential elections)*

Five members of polling station appointed by relevant Departmental Electoral Court, Party observers may be present,

Yes

Costa Rica

1

Rest day

Yes

11

28 days (presidential (elections)§

Every political party contesting election may appoint one person to polling station. No party observers.

Yes

Cyprus

1

Rest day

Yes

10.5

7 days (presidential elections)

Presiding officer and assistants appointed by returning officer (appointed by government), Party observers may be present.

Polling officials: nomination procedure and presence of party observers

Identification document required

Yes

Table 5.1 (Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Czech Republic

2

Weekday No and rest day

15

Not applicable

Ward Election Board (no less than five members) composed of persons nominated by parties (one per list). No party observers.

Yes

Denmark

1

Weekday

No

11

Not applicable

Polling supervisors (five to nine) elected by PR by local council in which polling district located, from among voters of district. They are assisted by appointed electors, elected by PR by local council in which district is located, from among voters residing in municipality. In practice, they are party representatives.

No

Ecuador

1

Rest day

No

11

42 days (presidential elections)

Three-member Ballot Reception Committee appointed by Provincial Electoral Tribunal, from among names submitted by parties. Tribunals composed of 7 members appointed by Supreme Electoral Tribunal, preferentially from political parties. Party observers may be present.

Yes

Estonia

1

Rest day

No

11

Not applicable

Division committee members (up to nine)

Yes

Polling officials: nomination procedure and presence of party observers

appointed by relevant local government council. No party observers.

Identification document required

Table 5.1

(Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

France

1

Rest day

No

10

7 days (legislative elections) 14 days (presidential elections)

Presiding officer mayor, adjunct mayor, or municipal councillor; scrutineers appointed by candidates (one per candidate). Party observers may be present,

Germany

1

Rest day

No

10

Not applicable

Electoral officer and deputy appointed by In some relevant Land government. Other Electoral cases only Board members (three to five) appointed by electoral officer. But Land government may order that members of Electoral Board be appointed by local authority. In appointment of board members, regard shall as far as possible be paid to political parties represented in polling district concerned. No party observers.

Guyana

1

Weekday

Yes

12

Not applicable

Presiding officer and poll clerks appointed by chair of Elections Commission. Party observers may be present.

Hungary

1

Rest day

No

13

14 days (legislative elections)

Returning boards elected by representative Yes boards of municipality, from among names proposed by heads of local electoral offices. In addition, each candidate or list may designate one member. No party observers.

Polling officials: nomination procedure and presence of party observers

Identification document required Yes

Yes

Table 5.1

(Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of Number of days granted time hours for between rounds off for voting voting of voting

India

Other

Rest day

No

9 or 10

Not applicable

Presiding officer appointed by relevant district election officer, deputies appointed by Election Commission in consultation with government of (corresponding) state. Party observers may be present

In some cases only

Ireland

1

Weekday

No

14

Not applicable

Presiding officer - and poll clerk appointed by relevant returning officer - sheriff or county registrar or person appointed by minister for environment. Party observers may be present

No

Israel

1

Weekday declared holiday

Yes

15

Not applicable

Polling Committees (PCs) appointed by Yes District Committees with approval of Central Elections Committee. Number of members and party composition of each PC prescribed each year by Central Committee. Members of PC may be chosen from any of party groups represented on Central Committee, but each PC shall include representatives of at least three party groups. Observers from parties not represented in Polling Committee may be present.

Italy

1

Rest day

No

NA

Not applicable

President of polling station appointed by president of court of appeal from among magistrates, state lawyers, prosecutors, or other public servants. Four scrutineers appointed by president of polling station. Party observers may be present.

Polling officials: nomination procedure and presence of party observers

Identification document required

Yes

Table 5.1

(Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Jamaica

1

Weekday

Yes

10

Not applicable

Presiding officer and poll clerk appointed by the governor general (in practice, government). Party observers may be present.

Yes

Japan

1

Rest day

NA

11

Not applicable

Superintendent of poll appointed by election management committee from among voters. NA (party observers)

NA

Latvia

1

Rest day

Yes (military 12 personnel only)

Not applicable

Seven members of polling commission appointed by relevant local council. Political parties may submit proposals. Party observers may be present.

Yes

Lithuania

1

Rest day

No

14

14 days (legislative and presidential elections)

Electoral committees of electoral districts appointed by electoral committees of electoral area from names proposed by parties represented in municipal council. Party observers may be present.

Yes

Luxembourg

1

Rest day

No

6

Not applicable

Presiding officer is judge or person appointed Yes by president of relevant principal office (local official), from among electors of commune. Four scrutineers appointed by presiding officer. Party observers may be present, but limit of three per polling station, chosen by lot from among all party representatives.

Polling officials: nomination procedure and presence of party observers

Identification document required

Table 5.1 (Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Madagascar

1

Rest day

No

11

NA (legislative elections) 56 days (presidential elections)"

Polling officer, assistant, and four clerks appointed by general assembly of city. Party observers may be present,

Yes

Malawi

1

Weekday

NA

12

Not applicable

Polling officers (at least five) appointed by Yes

Not applicable

Polling officials: nomination procedure and presence of party observers

Identification document required

Electoral Commission Party observers may be present.

Mali

1

Rest day

No

10

14 days (legislative and presidential elections)

Presiding officer appointed by local electoral commission. Scrutineers appointed by each candidate or list (1 per party), Party observers may be present.

Yes

Malta

1

Rest day

No

15

Not applicable

Assistant commissioners (at least three) appointed by Electoral Commission, from among names submitted by parties. Moreover, commission appoints one assistant commissioner on its own. Party observers may be present.

Yes

Micronesia

1

Weekday

No

10

Not applicable

Board of election members appointed by national election commissioner of each state, with approval of national election director. Both officials appointed by president with advice and consent of Congress. Party observers may be present.

No

Table 5.1

(Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Mongolia

1

Rest day

No

15

14 days (presidential elections)

Members of division election committee appointed by relevant local election committee, taking into account recommendations of president of local assembly. Party observers may be present.

Yes

Namibia

2

Weekday

No

28

Not applicable

Presiding officer and polling officers appointed by Electoral Commission. Party observers may be present.

Yes

Netherlands

1

Weekday

Yes

11

Not applicable

Members of electoral committee appointed by relevant burgomaster and aldermen (mayor and municipal councillors). In practice, not party representatives. No party observers.

In some cases only

New Zealand

1

Rest day

Yes

10

Not applicable

Deputy returning officer and poll clerks appointed by returning officer, appointed by chief electoral officer. Party observers may be present.

No

Panama

1

Rest day

Yes

9

Not applicable

Three polling officers appointed by Electoral Tribunal; each party may also appoint one polling officer. Party observers may be present.

Yes

Papua New Guinea

Other

Other

Yes

Other

Not applicable

Presiding officer, assistants, and poll clerks appointed by returning officer, appointed by Electoral Commission. Party observers may be present.

No

Polling officials: nomination procedure and presence of party observers

Identification document required

Table 5.1

(Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of Number of days granted time hours for between rounds off for voting voting of voting

Philippines

1

Weekday declared holiday

No

8

Not applicable

Board of election inspectors (three public school teachers) appointed by Commission on Elections Party observers may be present.

In some cases only

Poland

1

Rest day

No

16

14 days (presidential elections)

Members of district electoral commission appointed from among voters, by executive committee of commune, taking into account proposals submitted by election committees (parties). Party observers may be present.

Yes

Portug

al 1 Re

st day Ye

s11

21 dac

ys Polling (presidential elections)

officers appointed by parti Party observers may be present,

es. Yes

Romania

1

Rest day

No

15

14 days (presidential elections)

Ward Election Board composed of president Yes and assistant magistrates or jurists appointed by president of departmental tribunal. Other members are party representatives. No party observers present. However, candidates and delegates from press, TV, radio, and international organizations may be present.

St Lucia

1 Wee

kday Yes 1

1 .5 N

ot applicabl

e Presiding of

ficer and poll clerk appointed b Electoral Commission. Party observers may be present.

Polling officials: nomination procedure and presence of party observers

Identification document required

y Yes

Table 5,1

(Continued) Number of days between rounds of voting

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

St Vincent and the Grenadines

1

Weekday

Yes

10

Not applicable

Presiding officer and poll clerks appointed by supervisor of elections. Party observers may be present.

Yes

Samoa

1

Weekday declared holiday

No

9

Not applicable

Deputy returning officer and poll clerk appointed by chief electoral officer. Party observers may be present.

Yes

Sao Tome and Principe

1

Rest day

Yes

11

21 days (presidential elections)

Members of voting assemblies appointed by parties. Party observers may be present.

Yes

Slovakia

2

Weekday No and rest day (presidential elections: rest day)

15

14 days (presidential elections)

Members of Divisional Electoral Commission appointed by parties. No party observers.

Yes

Slovenia

1

Rest day

NA

12

NA (presidential elections)

Polling committees appointed by district electoral board, from among names submitted by parties. Party observers may be present.

No

South Africa

1

Weekday

No

NA

Not applicable

Presiding officer and voting officers appointed by chief electoral officer, in consultation with electoral commission. Party observers may be present.

Yes

Country

Polling officials: nomination procedure and presence of party observers

Identification document required

Table 5 . 1 ( Continued)

Country

Number of Polling day: polling weekday or day(s) rest day

Employees Number of granted time hours for off for voting voting

Number of days between rounds of voting

Spain

1

Rest day

No

11

Not applicable

Presiding officer (must have a high-school diploma or equivalent) and two members appointed by lot from among all literate electors under 65 years of age registered in corresponding ward. Party observers may be present.

Yes

Sweden

1

Rest day

NA

12

Not applicable

Electoral officers appointed by relevant Election Committee (local election authority in each municipality). NA (party observers)

In some cases only

Taiwan

1

Rest day

No

NA

Not applicable

Chief administrator and other administrators Yes appointed by election commission. Party observers may be present. Every candidate may appoint one, but if assigned observers are more than prescribed number for polling station, choices determined by drawing lots.

Trinidad and Tobago

1

Weekday

Yes

12

Not applicable

Presiding officer and poll clerks appointed by Elections and Boundaries Commission. Party observers may be present.

Yes

United Kingdom

1

Weekday

No

15

Not applicable

Presiding officer and poll clerks appointed by returning officer - an ex-officio county public servant. Party observers may be present.

No

Polling officials: nomination procedure and presence of party observers

Identification document required

Table 5 1

(Concluded)

Country

Number of polling day(s)

Polling day: weekday or rest day

Number of Employees granted time hours for off for voting voting

Number of days between rounds of voting

Polling officials: nomination procedure and presence of party observers

Uruguay

1

Rest day

Yes

11.5

28 days (presidential elections)*

Three members of Reception Committee appointed by relevant electoral junta. Party observers may be present.

Yes

Vanuatu

1

Weekday declared holiday

No

9

Not applicable

Presiding officer and scrutineers appointed by relevant returning officer, appointed by secretary of Electoral Bureau. Party observers may be present.

Yes

Venezuela

1

Rest day

Yes

10.5

Not applicable

Polling officers appointed by National Electoral Council. Party observers may be present.

Yes

Identification document required

NA = Not available * First round of presidential elections held on first Sunday of October; second round on last Sunday of same month. In the 1998 election, this would have meant 21 days between rounds. * Constitution says 'within 15 days' (section 121). * Based on dates of 1999-2000 presidential election. § First round held on first Sunday of February; second round on first Sunday of April. 11 Based on dates of 1996 election. * First round held on last Sunday of October; second round on last Sunday of November. In 1999, that amounted to 28 days between two rounds.

116 Establishing the Rules of the Game Number of Polling Days There are two issues here. First, as a general legislative election is legally a collection of local elections (at least in countries with numerous electoral constituencies), should all electoral districts vote on the same day or days? Second, assuming that they do, should balloting within each electoral district take place on a single day or be spread over a few days? Judging by the practice of most countries and given that, other things being equal, multiplying polling days increases costs, the answer to both questions may seem obvious. Yet such was not always the case, and some countries still have more than a single polling day. In earlier British elections held under a restricted franchise, constituencies voted on different days, and within each constituency polling took place over many days. This method helped the government. As voting was public, candidates and their organizers could be informed of the voting trends on a daily, indeed hourly, basis and had ample time to mobilize their supporters if needed or to resort to less lawful means. The government could schedule voting so that results from ridings leaning its way would be known earlier, presumably influencing doubtful or opposition-leaning constituencies that voted a few days or even weeks later. Such practices became the target of reformers in the 19th century and were subsequently eliminated. A less interested case can be made for opening polls on more than a single day, as some categories of voters may be unable to vote on the day appointed for polling. However, advance polls and/or voting by mail or by proxy (see the section on special voting procedures) can address this problem. An overwhelming majority of democracies (92 per cent) hold their legislative elections on a single day. Only three (the Czech Republic, Namibia, and Slovakia) have two days of voting, while India and Papua New Guinea allow for more than two. In India, however, there is a single day of polling in each constituency, so that no elector has more than one day to cast a vote.1 In presidential elections, only Namibia has two voting days. In all the above distinctive cases, ballots are not counted, and therefore results are not known, before all electors have had an opportunity to vote. Countries with more than one polling day do not present striking similarities among themselves. However, both the Czech Republic and Slovakia are the successor states to Czechoslovakia, which had two polling days.

How People Vote 117 Polling Day: Weekday or Rest Day?

On which day of the week should polling take place? Here the choice is between a weekday or a rest day (Saturday or Sunday). On a rest day, work will prevent fewer people from voting. In advanced Western countries, however, the growing practice of opening businesses on Sundays deflects this argument. The same objective can be met by obliging employers to grant their employees a specific number of hours off during the day so that they have the time to get to the polls - a practice that can be costly for employers. A supplementary argument is that schools, frequently chosen for polling purposes, are readily available on rest days, while they can be used on weekdays only if schoolchildren are granted a holiday. Some people in countries with large Christian populations object as a matter of principle to the opening of polls on Sundays, in view of its sacred character for them. This concern used to be common when elections routinely gave rise to disturbances, drunkenness, or even violence. Voting on a rest day may reduce turnout, as some voters may find leisure more attractive than civic duty. On a rest day, however, electors have more free time to go to the polls than on a weekday; thus turnout should in fact increase. Evidence on the impact of weekend voting is ambiguous. Franklin (2002) reports findings that suggest slightly higher turnout on the weekend, but a study that includes more countries (61) found no such effect (Blais, Dobrzynska, and Massicotte 2001). The rationale for holding legislative elections on rest days has been accepted by a majority (59 per cent) of democracies. The day is most often a Sunday, except in Australia, Bulgaria, Latvia, Malta, New Zealand, and Taiwan, which choose Saturday. In India, the latest election (1999) was spread over two Saturdays and three Sundays. In contrast, 18 countries (29 per cent), such as Denmark, South Africa, and the United Kingdom, hold their legislative elections on a regular weekday. In five countries (Chile, Israel, the Philippines, Samoa, and Vanuatu) this weekday is declared a public holiday for that purpose only. In the Czech Republic and Slovakia, with two days set aside for polling, one is a weekday (Friday) and the other is a rest day (Saturday). Papua New Guinea has more than one weekday and more than one rest day for voting. For presidential elections, 25 countries (83 per cent) hold their contest on a rest day. Only Guyana, Ireland, Malawi, Namibia, and the Philippines choose their president on a weekday The Philippines

118 Establishing the Rules of the Game

declares that weekday a public holiday. In many countries, however, the day(s) set for holding elections may vary from one contest to another. For instance, Denmark held elections on a Tuesday until 1988 and also in 2001 and on a Wednesday between 1988 and 2001. In Ireland, the last four legislative elections of the 20th century were held on different days (all weekdays).2 Harder to find are countries with elections sometimes on weekdays and other times on rest days. The 1999 legislative elections in Benin took place on a Tuesday, because a holiday fell on the preceding Sunday, the day usually set aside for polling. Holding legislative elections only on rest days is rare among former British colonies, which once again emulate their former ruler (see Table 5.2). Exceptions are Australia, Cyprus, India, Malta, and New Zealand. A rest day is the rule in Europe (74 per cent) and in countries with a Catholic majority (78 per cent). Time Off for Voting? As we saw above, one way to facilitate voting is to force employers to grant a certain amount of time off work for voting, without loss of pay. This privilege is found in 27 countries, such as Argentina, Australia, Latvia, and Portugal. In Latvia, it applies only to military personnel. Table 5.2 shows a surprising finding: former British colonies are more likely to adopt such a proviso than other countries, unlike Britain, which has no such practice. Number of Hours for Voting Another difference across countries is the number of hours polling stations are opened for voting. There is a tradeoff here again between the cost of more polling hours and the flexibility that they offer to electors. But closing polls earlier increases the likelihood that results will become public on the evening of the vote. The shortest period of polling (five hours) is in Belgium; the longest is 28 hours, in Namibia, where polls are open for two days. Among countries with a single polling day, the longest period allowed for voting is 16 hours (Poland). The average number of voting hours is 11.5.3 Number of Days between Rounds of Voting When more than one round of voting is necessary to determine the out-

Table 5.2 British colonialism, political rights, and polling day

British colonialism Former British colonies

Other countries

Poligical rights Established democracies

Other countries

Total

Polling on rest days only (legislative elections)

Employees granted time off for voting

Number of hours for voting

Identification document required for all electors

22%* (n = 23)

64% * (n = 22)

12.1 (n = 20)

74%

82%

36% (n = 36)

11.1 (n = 37)

79%

(n = 39)

59% (n = 22)

52%* (n = 21)

11.0 (n = 21)

59%f (n = 22)

59%

42% (n = 38)

11.7 (n = 37)

85%

(n = 41)

59%

46% (59 countries)

11.5 (58 countries)

76%

(63 countries) * Difference significant at 0.01 level (two-tailed f-test). 1 Difference significant at 0.05 level (two-tailed f-test).

(n = 23)

(n = 38)

(n = 40)

(62 countries)

120 Establishing the Rules of the Game

come of an election, the length of the 'second' campaign depends on the number of days separating both rounds. Waiting longer between the two rounds allows more time for electors to make up their mind among the remaining candidates. This is especially important for those who voted for a candidate who is not present in the second round. They need to assess whether they will bother to turn out and vote and, if so, for whom they will - sometimes reluctantly - cast their ballot. However, prolonging this process may increase costs for candidates and parties, wear out electors, reduce interest in the election, and lower turnout. For legislative elections, the two-round system is used in France, Hungary, Lithuania, and Mali. France allows for one week between the two rounds, and the other three countries, two weeks. In presidential elections, runoff elections are more widespread. Of the 20 countries where there may be more than one round of voting, eight hold the second round exactly two weeks after the first round.4 In two countries, Bulgaria and Cyprus, the second round of voting takes place only one week after the first one. In Argentina, the constitution provides that no more than 30 days can separate the first and second rounds. The longest time between rounds is eight weeks, in Madagascar.5 On average, rounds of voting in presidential elections are three weeks apart. Polling Officials

Who is in charge of conducting the election at the polling station on polling day? There are various appointment procedures, not unlike those for central electoral administrations. It is at times difficult to evaluate the role of political parties in the nomination process. Table 5.1 describes the various nomination processes. In our view, the crucial issue is whether representatives from political parties are entitled to sit in polling stations, either as election officials or as observers. Whereas in some countries parties are specifically involved in the nomination of polling officials, in most of them these nominations are seemingly non-partisan. Examples of the first arrangement are Canada, Costa Rica, Ecuador, Hungary, Mali, Panama, and Poland. In Canada, each polling station is under the authority of a deputy returning officer, nominated by the party that finished first in the district in the preceding election, and a poll clerk, nominated by the party that finished second.6 In Ecuador, the Provincial Electoral Tribunal appoints three-member ballot-reception committees from names submitted by

How People Vote 121

the parties. In Costa Rica, Hungary, Mali, and Panama, every candidate or list may appoint one member to the polling-station committee. In Poland, members of the district electoral commissions are appointed by the executive committee of the commune, taking into account proposals submitted by parties. Clearly these countries, involve political parties in the process. Elsewhere, the situation is more ambiguous. For example, in Jamaica the government appoints polling officers, which raises the possibility that partisan considerations may affect nominations. In Cyprus, polling officers are named by returning officers, themselves appointed by the government. Perhaps the most relevant piece of information concerns the presence or absence of party representatives as observers at the polling station on election day. Appointed by parties or candidates, these representatives are not election officials and are not allowed to interfere with the balloting process. However, they may sit in the polling station throughout polling day and may report on irregularities that they witnessed. The presence of such observers is provided for in 53 countries - a consensus that we find quite significant. Identification Documents The first step when a voter arrives at the polling station is for the person(s) in charge to ascertain whether the person is on the electoral register.7 In some countries, the elector states his or her name and, if it appears on the register, is allowed to cast a vote. This least-stringent procedure, employed in 10 countries such as Australia, Canada, Denmark, and Ireland, has the advantage of simplicity and rapidity. The obvious drawback is the greater chance of a registered elector being impersonated. To counter this possibility, a majority (47) of countries require all electors to prove their identity by presenting identification. Although this lowers the risk of fraud, voters may forget or lose their identification document, which may be a voter's card, a passport, a driver's licence, or a national identity card. Requiring a voter's card entails costs for the election administration, which must issue and distribute a card for each and every elector. It is, however, easier then to ascertain the voter's identity as the information printed on the card must correspond to that of the electoral list. In another five countries, electors must prove their identification on certain occasions only. For instance, in Germany this requirement applies only to electors who do not produce their notice of polling or to those voting in a constituency

122 Establishing the Rules of the Game

other than the one where they are registered. India has voter's cards in some electoral districts only. In the Philippines, the chair at the polling station may require voters to produce an authentic identification document when doubt arises over their identity. In Sweden, electors voting at post offices, at special voting stations, or at diplomatic missions must first show their voter's card. Interesting patterns develop here. First, three-quarters of former British colonies, unlike Britain, require all electors to produce some means of identification before voting. Second, established democracies are less likely to require voters to identify themselves other than verbally. - Non-established democracies probably worry more about electoral fraud. In the next two sections, we look at the kinds of ballots used in elections and at the voting procedures in democracies. BALLOT PAPERS For many people today, democracy is epitomized by the image of a voter dropping a piece of paper in a ballot box. Ballot papers and ballot boxes embody one of the few unchallenged principles governing elections - namely, secrecy.8 Like many other basics of electioneering, secret voting emerged through trial and error. In earlier elections, voters publicly declared their support for one of the candidates, and their preferences were duly recorded in the poll book. It was quickly realized that this practice, while it might fulfil the wildest dream of some election organizers and psephologists, defeated the very purpose of free elections by creating a fertile ground for corruption and intimidation. If an elector had been bribed for voting for a candidate, the latter had every means to check whether the deal had been respected. People known on the basis of their voting record to be supporters of a party could be prevented from voting for dubious reasons. Further, people who were in a position of dependence towards their employers or landlords were exposed to economic sanctions should they vote their own mind instead of their master's. In particular, at a time when women were legally subordinated to men, there was a widespread conviction that no woman should be allowed to 'spoil' her husband's vote by choosing a different candidate. Finally, in socially cohesive groups, voters were exposed to intimidation from their peers. Secret voting was advocated as a way to ensure that voters could express their choices freely and reveal their own

How People Vote 123

mind, rather than merely reflecting the constraints imposed by their environment. Voting became secret in Germany in 1867, in New Zealand in 1871, in Britain and Switzerland in 1872, in Canada in 1874, in Belgium in 1877, in Luxembourg in 1879, in Norway in 1884, in Denmark in 1901, in Austria in 1907, and in France in 1913. Today, ballot papers (or voting machines) are perceived as a prerequisite for free and fair elections and help to ensure the free expression of the voters' will. However, not all democracies use the same kind of ballot paper, and a few dispense with it in favour of mechanical devices or computers. In every country, electors may cast a vote on a first-come, first-served basis. In peak hours, this means several minutes, if not hours, of waiting. In Bolivia and Brazil, some categories of electors may cast a ballot as soon as they arrive. These include, in Brazil, candidates, electoral judges, party representatives, electors aged more than 75, electors suffering from a disability, and pregnant women. In Bolivia, such a provision covers polling officers, candidates, electors aged 70 or more, voters suffering from a disability, pregnant women, and sight-impaired electors. We find two basic types of ballot papers - 'Australian' and 'French' which dictate the steps whereby electors cast their vote. Tables 5.3 and 5.4 list the type of ballot papers used in legislative and presidential elections, respectively. After looking at these two ballots in turn, this section compares them, considers alternative methods of voting, compares ballots for legislative and presidential elections, and examines the use of symbols and/or pictures on ballot papers. Australian-type Ballots The first type is known as the 'Australian' ballot paper, because it was pioneered by Australian colonies in the 1850s and was advocated under that name in the United States, although it began to triumph there only in the 1890s. Under this system each elector receives a single ballot paper, which lists the names of all candidates and/or all parties in the electoral district, each followed by a blank spot. Electors indicate their choice(s) by marking one of these blank spaces and then fold the ballot paper and insert it in the ballot box. This simple scheme involves many specific arrangements. Some countries list candidates or parties in alphabetical order, while others list specific parties (for example, the government party) first, on the assumption that they will derive a slight advantage from this position.

124 Establishing the Rules of the Game Table 5.3 Ballot papers in legislative elections

Country

Number of ballot papers

Electronic voting or voting machines

Argentina Australia Bahamas Bangladesh Barbados Belgium Belize Benin Bolivia Brazil Bulgaria Canada Cape Verde Chile Costa Rica Cyprus Czech Republic Denmark Ecuador Estonia France Germany Guyana Hungary India Ireland Israel Italy Jamaica Japan Latvia Lithuania Luxembourg Madagascar Malawi Mali Malta Micronesia Mongolia Namibia

One per candidate/party One per elector One per elector One per elector One per elector One per elector One per elector One per elector One per elector One per elector One per candidate/party One per elector One per elector One per elector One per elector One per elector One per candidate/party One per elector One per elector One per elector One per candidate/party One per elector One per elector One per elector One per elector One per elector One per candidate/party One per elector One per elector NA One per candidate/party One per elector One per elector One per candidate/party One per candidate/party One per candidate/party One per elector One per elector One per elector One per elector

No No No No No Yes No No No Yes No No No No No No No No NA No Yes Yes No No Yes No NA No No NA No No No No NA No No No No No

Party symbols

Pictures

Yes No Yes Yes No Yes No Yes Yes Yes NA No Yes Yes Yes Yes No No Yes No Yes No Yes Yes Yes No NA Yes Yes NA No No Yes Yes Yes NA Yes No No Yes

No No No No No No No No Yes Yes NA No No No No No No No No No NA No No No No No No No No NA No No No Yes No NA No No No Yes

How People Vote 125 Table 5.3

(Concluded)

Country

Number of ballot papers

Electronic voting or voting machines

Netherlands New Zealand Panama Papua New Guinea Philippines Poland Portugal Romania St Lucia St Vincent and the Grenadines Samoa Sao Tome and Principe Slovakia Slovenia South Africa Spain Sweden Taiwan Trinidad and Tobago United Kingdom Uruguay Vanuatu Venezuela

One One One One One One One One One

Yes No Yes No No No No No No

No Yes Yes No No NA Yes Yes Yes

No No Yes Yes No NA No No No

One per elector One per elector

No No

Yes Yes

No No

One One One One One One One

per elector per candidate/party per elector per elector per candidate/party per candidate/party per elector

No No No No No No No

Yes Yes Yes Yes Yes Yes No

No No No Yes No No Yes

One One One One One

per elector per elector per candidate/party per candidate/party per elector

No No No No Yes

Yes No Yes Yes Yes

No No No Yes No

per elector per elector per elector per elector per elector per elector per elector per elector per elector

Party symbols

Pictures

NA = Not available.

The political affiliation of candidates may or may not be indicated, although the historical trend has been to include this most relevant piece of information. What kind of mark should be allowed on the ballot paper (a cross, a check, and so on) has raised much argument among legislators and judges, because a faulty mark may void the ballot paper, while an unconventional mark may be intended to reveal how a specific voter had voted. The ballot paper handed to the voter may or may not include a numbered counterfoil that is checked and detached from the ballot paper before the latter is put into the ballot

126 Establishing the Rules of the Game Table 5.4 Ballot papers in presidential elections Country

Number of ballot papers

Party symbols

Pictures

Argentina Benin Bolivia Brazil Bulgaria Cape Verde Chile Costa Rica Cyprus Ecuador France Guyana Ireland Lithuania Madagascar Malawi Mali Mongolia Namibia Panama Philippines Poland Portugal Romania Sao Tome and Principe Slovakia Slovenia Taiwan Uruguay Venezuela

One per candidate/party One per candidate/party One per elector One per elector One per candidate/party One per elector One per elector One per elector One per elector One per elector One per candidate/party One per elector One per elector One per elector One per candidate/party One per candidate/party One per candidate/party One per elector One per elector One per elector One per elector One per elector One per elector One per elector One per elector One per elector NA One per elector One per candidate/party One per elector

Yes Yes Yes Yes NA No Yes Yes Yes Yes No Yes No No Yes Yes NA No Yes Yes No NA No Yes

No Yes Yes Yes NA Yes No Yes Yes Yes No No No No Yes NA NA No Yes Yes No NA Yes No Yes No NA Yes No Yes

No No NA No Yes Yes

NA = Not available.

box - a safety device that helps to ensure that voters hand back the same ballot paper they were given.9 Finally, alternative or preferential electoral formulas may require voters to mark the ballot by numbering candidates in order of preference. Romania offers an interesting variant of the Australian ballot by printing on a booklet the name and symbol of parties, one of which voters mark with a rubber stamp supplied by polling officials. This type of ballot paper, however, is not immune

How People Vote 127

to misdeeds by electoral authorities. For example, in the Duvaliers' Haiti, voters were obliged to write down on the ballot paper the name not only of the candidate for which they were voting, but also of the electoral district and of the commune - quite a challenge in a country mostly illiterate.10 French-type Ballots

The other option known sometimes as the French model, or the 'ballot and envelope' model - provides for one ballot paper for each party or candidate. Electors must insert the ballot of the party or candidate for which they wish to vote into a standard opaque envelope (in order to preserve the secrecy of their choice) and then drop it into the ballot box; ballot papers for other candidates are discarded. Counting the votes is done by opening each envelope and identifying which candidate's or party's ballot was inside. The results of the counting are entered on a summary sheet. Again, many variations on this basic scheme can be imagined, with differences in elements from the quantity of information printed on each ballot paper and their colour to whether voters must bring to the polling station the ballot paper previously distributed by the party or candidate of their choice or can use ballot papers handed to them at the polling station. It is certainly preferable for the state to be in charge of printing and distributing ballot papers and envelopes in polling stations, rather than relying on parties and candidates, as weaker candidates are unlikely to have the means to fulfil this duty as effectively as others. Australian-type ballot papers are less costly, since voters receive a single ballot, irrespective of the number of candidates or parties contesting the election. The state prints the ballot paper, which saves money for candidates or parties. Further, following the count, ballot papers so marked can be inserted into sealed envelopes and kept for an indefinite period. In case a recount is requested, the envelopes are opened and the validity of each ballot paper, as well as the number of votes cast for each candidate, can be double-checked. However, because the ballot papers must be marked, the risk of a ballot being rejected is higher, especially when, under stringent rules, its validity is determined by over-zealous scrutineers or judges. Australian Senate elections suggest that a ballot paper of that kind can become quite unwieldy in districts electing ten members, especially if the ballot must have a picture of candidates as well as party symbols.

128 Establishing the Rules of the Game

The French-type ballot costs the state more, as ballot papers for each candidate must be printed in huge quantities (unless parties have to provide the ballots). Enough ballot papers for each candidate must be available in case all the voters registered in the electoral district show up and vote for that candidate. This requirement results in huge quantities of ballot papers being printed, at great cost, even if some are unlikely to be cast by more than a few dozen voters. After scrutineers have opened the envelope and counted the ballot, no physical evidence remains of the voter's will, except the statistical summary of polling operations. Should the validity of the official statement be challenged - conflicting statements for the same polling station have not been unknown in some jurisdictions - there is no possibility of recounting the votes in order to settle the dispute, as only invalid ballot papers and envelopes are kept after counting.11 It is theoretically possible to direct the ballots cast for each candidate to be inserted into an envelope after the count is over, but such ballots, being unmarked, would be identical to any ballot paper printed. However, this system normally results in a smaller number of invalid ballots, because it eliminates disputes over whether a ballot paper has been properly marked, unless a voter has mistakenly inserted ballot papers for different candidates in the envelope, in which case the vote is invalidated.12 Alternative Methods

The pace of technology, which makes ballot papers look somewhat quaint to many people, and the appetite of candidates and of the public for quick results, as well as the administrative problems generated by the use of ballot papers at elections involving a multiplicity of offices, have led some jurisdictions to consider alternative means of voting - namely, voting machines or computers. The former have become a common feature of American elections because of the U.S. custom of electing on a single day numerous public officials. Voting machines emerged as an answer to the problems generated by lengthy ballot papers or the confusion that the use of one ballot paper for each office might generate and to the time lag between the close of the poll and the publication of results. In other countries, alternative means of voting are still seen as imperfect solutions. Only eight countries have some electronic or mechanical method of voting, and in some cases it is still at an experimental stage. Brazil introduced electronic voting in the 1996 local elections in 57 municipal-

How People Vote 129

ities, covering one-third of the electorate. The proportion of electors voting by electronic means almost doubled, to 57.6 per cent, in the 1998 federal election, and in the local elections in 2000, all Brazilian municipalities used electronic ballot boxes (Chang Mota 1998). The presidential, congressional, and state (both gubernatorial and legislative) elections of October 2002 were conducted entirely with electronic ballot boxes, which work as follows. Voters face a keyboard and a screen. In order to vote for a candidate, they must press a specific number on the keyboard. The candidate's picture appears on the screen, and voters can then either confirm, reject, choose another candidate, or start the process again.13 The use of voting machines has been shown to decrease voter fatigue or 'roll-off' in elections where voters face several decisions (Nichols and Strizek 1995).14 Also, the lack of ballot papers to be tampered with decreases the risk of fraud. However, the efficiency of voting machines and electronic voting rests on the quality and reliability of the technology; lack of absolute trust in the latter has led many countries to put off such methods until they are foolproof. Legislative versus Presidential Elections

Roughly three out of four countries (48) use Australian-type ballot papers in legislative elections,15 while 21 countries (72 per cent) use it in presidential elections. Although Australian-type ballots predominate in all six continents except Africa, some interesting variations can be observed. The most striking is that the preference of Britons for this type is shared by all former British colonies, except Israel,16 Malawi, and Vanuatu. Although Spain uses ballots and envelopes, only two of its former Latin American colonies (Argentina and Uruguay) do so, while seven Latin American countries (including Brazil and Venezuela) provide for a single ballot paper. Even in continental Europe, France's lead has been followed by only six countries, including Spain, Sweden, and the Czech Republic, while all other countries (including Germany, Italy, and Poland) have opted for Australian-type ballots, as have all Asian countries in our sample (except Israel). The drawbacks of Australian ballots in multi-member districts do not appear to have prevented many PR countries from using them. Almost all countries use the same balloting procedures for legislative and presidential elections; exceptions are Benin and Slovakia. The former employs the French model for presidential elections and Aus-

Table 5.5 British colonialism, political rights, and ballot papers in legislative elections Australian-type ballot papers

Electronic or mechanical voting

Party symbols on ballot papers

Candidate pictures on ballot papers

87% (n = 23)

5%* (n = 21)

73% (n = 22)

17% (n = 23)

71% (n = 38)

19% (n = 37)

66% (n = 35)

15% (n = 34)

86% (n = 22)

18% (n = 22)

55% (n = 22)

0%f (n = 21)

Other countries

73% (n = 40)

11% (n = 37)

75% (n = 36)

24% (n = 37)

Total

77% (62 countries)

14% (59 countries)

67% (58 countries)

16% (58 countries)

British colonialism Former British colonies

Other countries

Political rights Established democracies

* Difference significant at 0.10 level (two-tailed Mest). f Difference significant at 0.01 level (two-tailed Mest).

How People Vote 131

tralian-type ballots for recent legislative elections (though not for the first one, held in 1991). The reverse holds for Slovakia. Symbols and Pictures In order to facilitate electors' choice and to ensure that illiterate voters may knowledgeably cast their vote, legislators have come up with two solutions: party symbols and candidates' pictures. Forty of our countries (69 per cent) allow parties or candidates to be identified by a symbol printed on ballot papers in legislative elections, and 16 (62 per cent) provide the same for presidential elections. Established democracies tend to shun the use of party symbols. Symbols undeniably facilitate voting, since electors may easily spot their preferred political party. This may prove extremely helpful in areas where illiteracy is high. Indeed, only three countries (Belize, Mongolia, and Papua New Guinea) where the literacy rate is below 90 per cent do not allow party symbols on ballot papers. More than a third (38 per cent) of countries with high rates of literacy (90 per cent and over) do not use such symbols. Symbols take up space, and the ballot paper becomes a dazzling sheet of coloured symbols, which may confuse voters. There have been instances, as in Romania in 1992, when different parties had similar symbols on their ballots (four had a map of the country, three had roses, three had a house). Also, some countries allow for the electoral administration to assign symbols to parties. What ensues is competition to acquire the 'best' symbol. Both Australian- and French-type ballot papers may include party symbols, although the latter seems more amenable because of its larger size and because using a distinct colour may make it easier to identify the ballot paper for a specific candidate. Printing pictures of candidates on ballot papers in legislative elections is less frequent: only nine countries (16 per cent) do so. Seven of these countries have literacy rates lower than 90 per cent. Once again, the argument in favour of candidates' pictures is that they help electors to spot their preferred candidate, provided that the printing is well done. However, in multi-member districts, it is nearly impossible to print the pictures of all candidates on ballot papers. Furthermore, some observers may fear that this practice focuses the election on candidates (and their looks) rather than on issues. Perhaps is it more appropriate to print presidential candidates' pictures on ballot papers, since there is a single official to elect and usually fewer candidates than in legislative elections. This may explain why a majority (56 per cent) of coun-

132 Establishing the Rules of the Game

tries with presidential elections print the candidates' pictures on ballot papers. Printing pictures for legislative elections seems confined to developing countries. Quite a few of these countries, however, do not (Benin, Ecuador, India, and the Philippines, for instance). None of the established democracies in our sample puts candidates' pictures on ballot papers. Many countries print the pictures of presidential but not legislative candidates. Such is the case in Benin, Cape Verde, Costa Rica, Cyprus, Ecuador, Portugal, and Sao Tome. SPECIAL VOTING PROCEDURES Election statutes in force a century ago typically provided a single way for electors to express their preferences: they had to go in person to the polling station and cast their vote. Yet employment or physical disability prevented some electors from doing so. It was up to electors to adjust to the electoral procedure, not the other way round. A more enlightened attitude prevails today, and although most electors cast their vote personally on election day, many special voting procedures now accommodate those unable to do so. These procedures owe their emergence and development to the complaints of many voters unable to cast a ballot because of circumstances such as physical disability, employment, and service abroad. In this section, we examine six types of special voting procedures: postal voting, advance voting, assisted voting, proxy voting, mobile polling stations, and other voting procedures.17 To obtain a clearer picture of each country's willingness to facilitate voting, we construct an index of such procedures. We call them 'special' because most countries use these methods for a relatively small proportion of the electorate. An exceptional case is Sweden, where more than a third of the electorate vote by (not-so) special procedures (Choe 1997, 191). We then discuss other means of voting, such as voting by internet, phone, or fax. Although these procedures are infrequent in nationwide contests, there are instances, such as local elections, where they have become standard. Table 5.6 lists the various procedures used in different countries. Postal Voting.18

With postal voting, electors need not show up at the polls and can mail their ballot(s) to election authorities. This procedure apparently origi-

How People Vote 133

nated in the United States during the Civil War, when huge numbers of men would lose the right to vote because the war took them away from their residence. During the First World War, general elections and referendums held in Australia, Canada, and New Zealand led to the adoption of similar statutes. Conscription for military service overseas, a measure that soldiers were presumed to support, was a hot issue in these countries - a factor that allegedly encouraged governments to extend voting rights in their case. Canada offers a typical illustration of the gradual extension of this wartime privilege. Postal voting was first provided for soldiers in 1915 and was allowed to lapse after the war ended. It was reinstated in 1940 and maintained thereafter for troops stationed abroad. Unsurprisingly, soldiers' wives pressed for the same advantage, which they received in 1955. The rise in the number of civil servants serving abroad and of their dependants resulted in their addition to the list in 1970. In 1993, postal voting was finally extended to all Canadians abroad and to those residing in the country who are unable to vote at their polling stations on election day. In effect, anyone may apply for a postal ballot, provided of course that he or she is on the electoral register. Procedures in force in most countries involve the following elements. Registered electors must notify election officials of their need for a postal ballot. On receiving voting kits, they mark the ballot paper and return it to election officials by mail. Such votes are counted only if they are received, or postmarked, before the close of regular polls. Elaborate precautions must preserve the secrecy of the postal vote while avoiding fraud. In Canada, an elector inserts the ballot, once marked, into an 'inner' envelope bearing no identification. He or she seals this envelope and inserts it into a slightly larger 'outer' envelope, which includes the voter's name and address. In turn, the elector inserts this envelope into a third envelope, which he or she sends to the chief electoral officer. Counting postal votes, election officials first examine the outer envelope to see whether the elector's name and address match the indications on the electoral register. If this is the case, they open the outer envelope and put the inner envelope into a ballot box, which preserves the vote's secrecy. This ballot box is opened only after the close of the polls on election day, and the results of the count are then added to those of other polling stations. In Australia, the procedure is similar, except that electors must mark their postal ballot in the presence of a witness. The ballot is then sent

134 Establishing the Rules of the Game Table 5.6 Special voting procedures Country

Postal voting

Advance voting

Assisted voting

Proxy voting

Mobile polling stations

Argentina Australia Bahamas Bangladesh Barbados Belgium Belize Benin Bolivia Brazil Bulgaria Canada Cape Verde Chile Costa Rica Cyprus Czech Republic Denmark Ecuador Estonia France Germany Guyana Hungary India Ireland Israel Italy Jamaica Japan Latvia Lithuania Luxembourg Madagascar Malawi Mali Malta Micronesia Mongolia Namibia Netherlands New Zealand

No Yes No Yes No No No No No No No Yes Yes No No No No No No Yes No Yes No No Yes Yes No No No Yes Yes Yes Yes No No No No Yes No No Yes Yes

No Yes Yes No No No No No No No No Yes No No No No No Yes No Yes No No Yes No No Yes Yes No Yes Yes No Yes No No No No Yes No No Yes No Yes

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

No No No No No Yes Yes Yes No No No No No No No No No No No No Yes No Yes No No No No No No No No No No No No Yes No No No No Yes No

No Yes No No No No No No No No No Yes No No No No Yes Yes No Yes No Yes No Yes No Yes No Yes No No Yes Yes No No No Yes No No No Yes No Yes

How People Vote 135 Table 5.6

(Concluded)

Country Panama Papua New Guinea Philippines Poland Portugal Romania St Lucia St Vincent and the Grenadines Samoa Sao Tome and Principe Slovakia Slovenia South Africa Spain Sweden Taiwan Trinidad and Tobago United Kingdom Uruguay Vanuatu Venezuela.

Postal voting

Advance voting

Assisted voting

Proxy voting

Mobile polling stations

No Yes No No No No No

No No No No Yes No Yes

Yes Yes Yes Yes Yes Yes Yes

No No No No No No No

No No No No No Yes No

No No Yes No Yes NA Yes Yes NA No Yes No No No

No No Yes No Yes NA No Yes NA Yes No No No No

Yes Yes Yes Yes Yes Yes Yes Yes NA Yes Yes Yes Yes Yes

No No No No No NA No Yes NA No Yes No Yes No

No No No Yes Yes Yes No No NA No No No No No

NA = Not available.

not to the central administration, but directly to the divisional returning officer. In Estonia, electors residing abroad must mail their ballots to the Estonian embassy in the country in which they reside. Postal voting may be used by electors who are temporarily away from their residence, which poses no problem as to the district for which their ballot will be counted. Electors not residing in the country may also use postal voting. Such votes need to be assigned to a particular electoral district. A common option is for electors abroad to be deemed to be residing in the district where they lived before leaving the country. Such address appears on the outer envelope and allows election officials to ensure proper assignment of postal votes.19 Postal voting allows voters to cast a ballot personally without having to go to a polling station. The chief argument against it is fear of

136 Establishing the Rules of the Game

fraud, which probably explains its confinement to a rather small minority of electors in most countries. Twenty-one countries use postal voting. In Canada, Lithuania, and Spain, any voter may request a postal ballot. Other countries permit only some categories of electors to vote by mail, such as soldiers, prisoners, electors outside their district on polling day, electoral officers, people working on polling day, and hospital patients. In some countries, however, the conditions are not very stringent, and most voters may vote by mail provided that they say that they will be unable to go to the polling station on election day. Wary of the risk of fraud associated with postal voting, non-established democracies are slightly less likely to allow it than established ones (see Table 5.7). No South American country allows electors to vote through the mail. Canada is the only country in the Americas in our sample that allows postal voting, although the United States also permits it. We find few jurisdictions with only postal voting in a national election. Oregon's 1996 election for the U.S. Senate was the first such contest. In favour of this method, holding elections is cheaper, turnout is higher, and voting is easier for electors unable to go to a polling station. Opponents see greater risk of fraud and of undue influence, loss of the camaraderie of the polling place, and more voting by uninterested persons (see Southwell and Burchett 1997, 53). Southwell and Burchett (1997, 56-7) show that Oregon voters surveyed after the election were mostly in favour of postal voting and that its use did not favour any one party. Theirs is one of several empirical studies showing the positive effect of postal voting on turnout (Southwell and Burchett 2000; Hamilton 1988; Magleby 1987; Oliver 1996; see also Sparrow 2000). Another instance of all-mail ballot was the election in 1997 of half of the delegates to the Australian Constitutional Convention (the other half were appointed). Every registered elector was mailed an election kit containing a ballot paper and information on the election. Turnout was lower than for an average election, as less than half of the electorate cast a ballot, partly because, unlike in legislative elections, voting was not compulsory. The total cost for the election was just over $24 million, compared to more than $60 million for legislative elections held a few months later.20 At the local level, Americans are no strangers to all-mail elections. In the decade leading to 1988, approximately 1,000 local elections were held using an all-mail ballot procedure (Hamilton 1988). In New Zealand, postal ballots are now used in all municipal elections.

Table 5.7 British colonialism, political rights, and special voting procedures Index of special voting procedures

Postal voting

Advance polling

Proxy voting

Mobile polling station

32% (n = 22)

55%* (n = 22)

14% (n = 22)

26% (n = 23)

1.23 (n = 22)

34% (n = 38)

21% (n = 38)

16% (n = 38)

32% (n = 38)

1.03 (n = 38)

45% (n = 22)

45% (n = 22)

27% (n = 22)

32% (n = 22)

1.50* (n = 22)

Other countries

28% (n = 39)

26% (n = 39)

10% (n = 39)

28% (n = 40)

0.90 (n = 39)

Total

34%

33% (61 countries)

16% (61 countries)

29% (62 countries)

1.11 (61 countries)

British colonialism Former British colonies

Other countries

Political rights Established democracies

(61 countries)

* Difference significant at 0.05 level (two-tailed Mest).

138 Establishing the Rules of the Game Advance Voting Advance voting requires electors to go to a polling station and cast a ballot personally on specific days prior to election day, as if it were polling day. On the close of advance polling, ballot boxes are sealed, and they are opened only at the end of the regular polling period. This procedure preserves all the elements of a regular polling station, except that it takes place on a different day. For the voter, it is not as flexible as postal voting, but it allows those who are unable to go to the polling station on election day to participate in the democratic process, provided that they can do so on the days appointed for advance polling.21 In some countries, such as Canada, advance polling takes place within special districts, including numerous polling stations, while in Sweden post offices serve as advance polling stations. In most countries, electors who wish to vote in advance must prove that they will be unable to go to the polling station on polling day. In some countries, such as Canada, Estonia, and Sweden, any voter can take advantage of advance polling. In other countries, only a small number of electors may vote before polling day - for instance, in Guyana, Jamaica, and St Lucia, members of the armed forces; in Malta, assistant electoral commissioners; in Ireland, residents of islands where a single polling day will not be enough; and in Israel, electors on vessels. The names of electors who vote in advance are struck from the list of electors for regular polling day, to prevent double voting. An exception is Sweden, where electors who use either option (postal or advance voting) may rescind their choice and cast a new ballot on polling day. They may change their mind, and their vote, before polling day - a privilege unavailable, to our knowledge, anywhere else in the world. All in all, advance voting is offered to all or some electors in 20 countries, half of which also have postal voting. Established democracies and former British colonies seem more prone to letting electors vote in advance (see Table 5.7). Assisted Voting The most common special procedure on polling day is assisted voting, whereby electors who cannot vote without some help may be assisted by another voter or a polling officer. Often provisions, however, deny candidates or their agents the right to assist a voter.

How People Vote 139

Democracies seem to have concluded that denying assistance to an elector is entirely unjustifiable; assisted voting is permissible everywhere. The remote possibility that 'assistants' will not express the voter's true preference is outweighed by the benefit of helping sightimpaired, illiterate, or otherwise disabled electors to mark their ballot. Another way of helping sight-impaired voters is to use templates, which dispense with the need for assistance. Proxy Voting Proxy voting allows an elector who cannot vote in person to appoint another voter to cast a ballot on his or her behalf at the polling station. Voters contact election officials before polling day to establish that they are eligible to vote by proxy and to indicate the name of the proxy. On election day, the proxy goes to the polling station and casts a vote on behalf of the voter. This procedure assumes that proxies will receive instructions to vote for a specific party or candidate and will vote accordingly. Given the need to preserve voting secrecy, neither assumption can be checked. Yet in most instances, proxies are relatives or trusted friends of those who appoint them. Proxy voting is less widespread and is used in only 10 countries, such as Belgium, Benin, France, and the United Kingdom. In Canada, this procedure formerly allowed prisoners of war in enemy camps to vote. Mobile Polling Stations 'If you are unable to come to the polling station, the polling station will come to you/ seems to sum up the rationale and working of mobile polls. They are used when an elector requires that polling officers visit him or her with ballot papers and a ballot box. They may visit institutions where many confined but otherwise registered electors wish to vote (such as prisons and hospitals). They differ from regular polling stations in hospitals or other institutions: they are temporary settings that are moved around by polling officers on polling day. Mobile stations are open for shorter periods. They are found in 18 countries, such as Australia, the Czech Republic, South Africa, and Sweden. Sweden sets them up before polling day as, in effect, a procedure for advance voting. For all special voting procedures (excluding assisted voting), established democracies tend to be more accommodating than weak ones

140

Establishing the Rules of the Game

Table 5.8 Political rights and the index of voting procedures Number of special voting procedures

0

1

2

3

Established democracies 14% (n = 3)

46% (n = 10)

18% (n = 4)

23% (n = 5)

Other countries

41% (n = 16)

36% (n=14)

15% (n = 6)

8% (n = 3)

Total

31% 39% (19 countries) (24 countries)

16% (10 countries)

13% (8 countries)

Gamma = 2.49 (p < 0.05).

(although differences are not statistically significant). We constructed an index of special voting procedures - postal voting, advance voting, proxy voting, and mobile polling stations. For each country, the index takes on a value equal to the number of special procedures in force. For instance, Sweden receives a score of 3 because it has postal, advance, and proxy voting. Nearly a third of our democracies (31 per cent) have no special voting procedures besides assisted voting. Thirty-nine per cent have one special voting procedure, 16 per cent have two, and 3 per cent have three. No country has every procedure (see Table 5. Only 14 per cent of established democracies do not offer any such procedure (besides assisted voting), while the proportion among nonestablished democracies exceeds 40 per cent. A gamma test shows a significant relationship between political rights and multiple special voting procedures (p < 0.05). On average, established democracies allow 1.5 special voting procedures, and other countries, 0.9 (see Table 5.7). Other Voting Procedures

Technological breakthroughs such as the internet offer alternative voting procedures that not only require less from voters but also do away with ballot papers. These include voting by telephone, fax, and the internet. These methods are used at different types of elections. For instance, some Canadian provincial parties use 'tele-voting' (phone) for the election of their leader (Cross 1996). Concern that these meth-

How People Vote 141

ods may not protect the secrecy of the vote has impeded their adoption at higher levels, but they are seriously considered in many countries. For the first time electors voted via the internet in a public legally binding election - Arizona's Democratic presidential primary in 2000. Also, in New Zealand, overseas electors who do not have access to special voting facilities may vote by fax. Experiences with new election technologies have multiplied in recent years, with Brazil emerging as a leader. Critics (Mercuri 2002) focus on the difficulty of reconciling privacy (preserving the secrecy of the vote) and auditability (the ability, when necessary, to recount the votes cast). Yet their use may spread over the coming decades. CONCLUSION This chapter has outlined numerous ways for electors to vote and shown that virtually no method is followed by every democracy. While the typical democratic country holds an election on a single rest day, with polling stations opened for about half a day, many do not follow this pattern. Parties usually participate at the polling station, whether through polling officials or through representatives. And obliging an elector to show identification is fairly standard. The predominant type of ballot paper is the Australian. Each elector receives a single ballot, on which the names of all the candidates and/ or parties appear. The French model, with a ballot paper for each party and/or candidate, prevails in only one country in four. Electronic or mechanical methods of voting at the polling station are still exceptional. The use of party symbols and/or candidates' pictures is confined mostly to countries with high rates of illiteracy, except in presidential elections, for which a majority of countries have candidates' pictures on ballots. Finally, most countries try to facilitate voting by providing special voting procedures. However, assisted voting is the only one universally accepted in democracies; other procedures are less widespread. Perhaps election officials perceive the act of a voter depositing his or her ballot paper in the ballot box on election day as essential to the democratic process. After all, it is the only way to ensure that voters will personally mark their ballot without outside interference. But the recent trend has been to extend the use of special voting procedures, as demonstrated by the increasing use of the internet.

Chapter Six

Sorting Out Winners and Losers

Once electors have cast their vote, their preference(s) must be ascertained and counted. Transparency, honesty, and efficiency are the essential goals of legislative provisions governing that part of the process. In addition, election law establishes a procedure for adjudicating electoral disputes, either by recounting ballot papers or by trying an election petition. This chapter contains two sections. The first describes the various settings and conditions under which ballot papers are counted. The second looks at the possibility, after announcement of the results, of a recount of any or all ballot papers and at the ways in which election petitions are tried. COUNTING THE VOTES1 Counting the votes typically involves the following operations. In the polling station, the ballot box is opened: one by one (preferably), ballot papers or envelopes are taken out of the box, and a decision is made as to their validity.2 Then, votes for each candidate are counted, and the results are entered on official tally sheets and communicated to the electoral body responsible for the electoral district, so that voting figures for the district may be tabulated and the final result proclaimed. In countries using Australian type ballots, ballot papers cast for each candidate, as well as rejected ballot papers, are inserted into distinct sealed envelopes and preserved for further examination, should a recount be held at a later date. In countries where people vote by inserting a ballot into an envelope, both ballots and envelopes are usually disposed of on the spot, and the official tally sheet becomes the

Sorting Out Winners and Losers 143

only supporting evidence of the result. In some cases, however, ballots may be kept for further review - for example, when electors may express a preferential vote within a party's list. Vote counting differs from one country to another, depending on the electoral system used. In countries with plurality systems, majority (two-ballot) systems, or PR systems with closed list, all that is needed is to count the ballots cast for each candidate or party. The use of alternative voting or single transferable vote (STV) involves numerous preferences, only the first one being normally counted on election night. Subsequent preferences are counted later and transferred if necessary. List systems of proportional representation (PR) involving either panachage or the expression of preferences for individual candidates make vote counting quite complex. In this section, we do not review all the details of vote counting but focus on three dimensions of this process. First, we examine where vote counting takes place: at the polling station, at the electoral-district level, or at the national level. For this purpose, we consider the decisive vote count. Second, we look to see whether a single official or more than one person determine whether a ballot paper is valid. Third, we inquire about the presence of party observers at the count. Table 6.1 summarizes our findings. In developed countries, technology has made voting results available at various geographical levels: polling stations, electoral districts, regional subdivisions, and the country as a whole. Legally speaking, the only results that matter for the distribution of parliamentary seats are those for electoral districts, as these are the units that legislators are elected to represent. Politically speaking, election statistics for the whole country are highly relevant aggregates and catch the attention of the media and of political analysts. Yet, unless the whole country is a single electoral district, as is normally the case for presidential elections or in Israel and the Netherlands for legislative elections, national figures are not decisive. Where Are Ballot Papers Counted?

The simplest option is for ballots to be counted on the spot at the polling station, immediately when polling is over, and not to be counted again unless the results for the electoral district are challenged or are so close that the legislation calls for a recount. This approach is generally recommended by international organizations, for many reasons. An

144 Establishing the Rules of the Game Table 6.1 Counting the votes

Country

Location of decisive count

Number of persons to judge validity of ballot papers

Party observers may be present during count

Argentina Australia Bahamas Bangladesh Barbados Belgium Belize Benin Bolivia Brazil Bulgaria Canada Cape Verde Chile Costa Rica Cyprus Czech Republic Denmark Ecuador Estonia France Germany Guyana Hungary India Ireland Israel Italy Jamaica Japan Latvia Lithuania Luxembourg Madagascar Malawi Mali Malta Micronesia Mongolia Namibia Netherlands

Polling station Electoral district Electoral district Polling station Electoral district Other (see text) Electoral district Polling station Polling station Polling station Polling station Polling station Polling station Polling station Polling station Polling station Polling station Other (see text) Polling station Other (see text) Polling station Polling station NA Polling station Electoral district Electoral district Polling station Polling station Electoral district Other (see text) Polling station Polling station Polling station Polling station Polling station Polling station Country Other (see text) Polling station Electoral district Polling station

One One One One One More than one One More than one More than one More than one More than one One One More than one More than one One More than one More than one More than one More than one More than one More than one One More than one One One More than one One One NA More than one More than one More than one More than one NA NA More than one More than one More than one One More than one

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes No Yes Yes No Yes No Yes No Yes Yes Yes Yes Yes NA Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No

Sorting Out Winners and Losers 145 Table 6.1 (Concluded)

Country

Location of decisive count

Number of persons to judge validity of ballot papers

Party observers may be present during count

New Zealand Panama Papua New Guinea Philippines Poland Portugal Romania St Lucia St Vincent Samoa Sao Tome Slovakia Slovenia South Africa Spain Sweden Taiwan Trinidad and Tobago United Kingdom Uruguay Vanuatu Venezuela

Electoral district Polling station Electoral district Polling station Polling station Polling station Polling station Electoral district Electoral district Country Polling station Polling station Polling station Polling station Polling station Other (see text) Polling station Polling station Electoral district Electoral district Polling station Polling station

One More than one One More than one More than one One More than one One One One More than one More than one More than one One One More than one More than one One One More than one One More than one

Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes No Yes Yes Yes NA Yes Yes Yes Yes Yes Yes

NA = Not available

election involves suspense, and suspense should not be allowed to last for too long. Counting the votes at a higher level than the polling station will delay proclamation of the results, if only while the ballot boxes are carried to another location. This is a serious inconvenience for candidates, for the media, and for the public, and in tense political contexts it may foster public mistrust. Politicians and their organizers also like to know their own areas of strength and of weakness. Counting at a higher level limits the information available to party strategists and politicians.3 However, there may be sound reasons for counting the votes at higher levels. The country may be very small, making it easy to carry ballot boxes to a central place. There may be doubts as to the ability of polling officers to tabulate results, especially if the voting system is

146 Establishing the Rules of the Game

complex. In developing countries, polling stations are often located in open areas or in poorly lit buildings, while counting normally takes place after sunset. Transporting the ballot box to a centrally located, well-equipped building may make sense, provided that some precautions are taken to avoid tampering, ballot-box stuffing, or simply the box being lost! Keeping secret the details of the vote for smaller geographical units may prevent reprisals against areas that supported defeated candidates. Benin, where counting was done at a higher level at the first elections in 1991, provides a good example of the precautions needed for moving ballot boxes. On completion of the balloting process, the opening at the top of the ballot box was shut and locked, while the lid of the box was crafted so that it would be impossible for an envelope to be inserted into the box during transportation. Ballot boxes were carried by police trucks and watched over throughout the trip to the counting office by gendarmes and party scrutineers alike. No serious doubt was expressed as to the integrity of the process, but Benin later followed foreign observers' advice and provided for counting in polling stations. The decisive vote count takes place at polling stations in 40 of the 62 countries for which the information is available. It is conducted at the level of electoral districts in 14 countries. In Belgium, counting is done in-between, at 'counting bureaus.' On average, there is one counting bureau for three polling stations. In Japan, the decisive count is conducted at a central place in the town, city, or village. In Micronesia, ballots are counted at the level of the four states, which are electoral districts for the election of four Senators. Two countries, Malta and Samoa, go further, as votes are counted for the whole country in a central location. This procedure is possible only in small countries where no ballot box has to be carried thousands of kilometres from the polling station. Malta's STV system requires numerous and complex procedures for counting and transferring preferences, which can better be made at the central level. The information above describes the decisive vote count. Some countries have a preliminary scrutiny conducted earlier. Such a count can provide the media with preliminary results. In case of discrepancies between both counts, the second count is decisive. We find such arrangement in ten countries. In Australia, the Bahamas, Jamaica, New Zealand, St Lucia, and St Vincent, there is a preliminary count at the polling station (including, in Australia, an interim assessment of prefer-

Sorting Out Winners and Losers 147

ences between the two leading candidates), while the returning officer conducts the decisive count at the electoral-district level. In Denmark, there is a preliminary count at the polling station, while the decisive count occurs at the nomination-district level - a unit smaller than an electoral district. In Estonia, the preliminary count at the polling station is followed by a decisive count at the county level. In Sweden, counting takes place first at the polling station; each county's administrative board handles the decisive count. Finally in Samoa, the country-level vote count follows preliminary scrutiny in each polling station. British influence seems to determine the location of the decisive vote count (see Table 6.2). While only 36 per cent of former British colonies conduct that count at the polling station, 82 per cent of other countries do so. Most erstwhile British colonies follow the United Kingdom's lead and conduct their decisive tabulation at the electoral-district level. Among the 13 countries, in addition to Britain, that conduct their decisive count at the electoral-district level, all but Uruguay are former British colonies. For two of these countries, the use of the alternative vote (Australia) or of the single transferable vote (Ireland) appears to be a supplementary factor, as transfers of second and subsequent preferences are best done at that level, though in Australia a 'two candidate preferred count,' as noted above, is conducted in polling stations on election night in order to give an indication of the likely outcome of the poll in each division. As for multiple vote counts, almost one-third of former British colonies have them, while under one-tenth of other countries hold a preliminary count. This finding belies our expectations, since the United Kingdom does not conduct a preliminary vote count at the polling station. However, a clear majority (70 per cent) of former British colonies do not conduct preliminary vote counts. We also find that established democracies are less likely to conduct their decisive count at the polling station. Who Has the Right to Declare a Ballot Invalid?

Counting the ballots involves first deciding whether a ballot must be excluded from the count as invalid. This is an important prerogative. If held by a single official, it can be used in mischievous ways. If this official is appointed by the government, he or she can give unfair advantage to the government's candidates, by allowing doubtful ballots cast for the 'right' candidate, while overzealously invalidating other ballots. This possibility can be diminished, and even eliminated, by awarding

Table 6.2 British colonialism, political rights, and counting procedures Decisive vote count conducted at polling station

Decisive count preceded by preliminary vote count

Decision on validity of ballot papers lies with single official

36%* (n = 22)

30%t (n = 23)

91%* (n = 22)

82% (n = 39)

8% (n = 39)

16% (n = 37)

50%* (n = 22)

23% (n = 22)

59%* (n = 22)

Other countries

73% (n = 40)

12% (n = 41)

37% (n = 38)

Total

65% (62 countries)

16% (63 countries)

45% (60 countries)

British colonialism Former British colonies

Other countries

Political rights Established democracies

* Difference significant at 0.01 level (two-tailed Mest). f Difference significant at 0.05 level (two-tailed f-test). * Difference significant at 0.10 level (two-tailed f-test).

Sorting Out Winners and Losers 149

this prerogative to more than one person. This prevents such unilateral decisions. Ideally, board members should be neutral or appointed by various factions to ensure representation of more than one party. However, this arrangement can foster endless argument about the validity of a ballot paper. Countries are almost evenly divided between the two options (see Table 6.1). In 27 countries, the power to declare a ballot invalid lies with a single person, usually the presiding officer at the polling station, or the returning officer for the electoral district. In 33 countries, it lies with more than one person (the information was not available for Japan, Malawi, and Mali). A committee decides on the validity of a ballot paper, usually by majority vote. An overwhelming majority (91 per cent) of former British colonies empower a single official to decide whether a ballot paper is to be rejected or not. They seem to view efficiency as the main consideration. Coupled with the fact that most of these countries conduct their decisive count at the electoral-district level, this gives enormous power to a single official - the returning officer - who holds the fate of the candidates in his or her hands. However, legal procedures available to candidates or parties reduce fears that a dishonest official will tip the balance of an election. Non-established democracies are less likely so to empower a single official and turn a count at the polling station, conducted by more than one individual. Party Observers

It is common practice for vote counting, like balloting, to be supervised by party appointees. Candidates appoint their own representatives by way of a written document signed by them. On presentation of this document to the presiding officer, representatives are allowed to observe the process and, in some countries, to have their objections formally recorded in the official report of the proceedings. They may be paid for their efforts by the candidate or by the state. Their presence is widely argued to be the best guarantee that the process will not be tainted by fraud. In all but eight of the countries surveyed (the relevant information is unavailable for Japan and Sweden), political parties may appoint representatives to observe the decisive count (see Table 6.1). In fact, every country that allows party representatives to observe the proceedings of the election at the polling station also allows representatives to observe

150 Establishing the Rules of the Game

the count. In the eight other countries, political parties are involved in the nomination and appointment of those actually counting the ballots. It seems that the presence of different political interests, in this form of either observers or officials, around the table when votes are counted is universally acknowledged to be a worthwhile precaution. Parties may appoint observers in all cases where a single official can decide on the validity of ballot papers. RECOUNTS AND PETITIONS Elections are battles that are fought, won, and lost. They involve powerminded individuals inclined to suspect cheating from the other side, and the results will inevitably be difficult for the losers to accept. Established democracies have been so successful at eliminating electoral corruption and reducing the amount of litigation associated with electoral campaigns that even knowledgeable observers have forgotten that in the 19th century it was common for the results of British elections to be challenged in scores of seats. Students of elections in developing countries, where electoral results are frequently disputed and where election days are followed by a flurry of accusations of fraud, are routinely reminded that a hard-fought election whose results are accepted without demur by those who lost is a precious collective achievement that should not be taken for granted. If the electoral game is to be played fairly, its rules must be followed by all participants. The right to challenge the results of the election on the basis of a breach of the electoral law may be seen as a sine qua non for holding free and fair elections. In the absence of such a right, the results may remain illegitimate for a large section of the public, who want to be assured that no corrupt or illegal practices took place. Procedures in that area are highly complex, vary from one country to another, and have been little studied. We do not claim to describe in detail the procedures in force in various countries. We rather examine two standard procedures used by losing parties or candidates who feel that the electoral process has not been fair - namely, the possibility of demanding a recount of ballot papers, and the procedure (referred to as 'election petitions' in British electoral parlance) allowing candidates to challenge the validity of the election on the ground that some provision of the election law was not respected. Recounts normally take place shortly after ballot papers have been counted and entail not only a new compilation of ballots, but also a

Sorting Out Winners and Losers 151

review of the validity of the votes cast. It can lead either to a confirmation of the outcome already arrived at or to the election of another candidate. Election petitions, in contrast, go far beyond by aiming at the invalidation of the election of a candidate on the ground of corrupt or illegal practices perpetrated on his or her behalf at any stage of the electoral process. Final decision on such petitions may take a long time, depending on the procedure. There are instances of elections being voided many years after the accused legislator had been defeated at a subsequent election.4 Recounts We are interested here in what happens when parties, candidates, or their representatives request a recount of ballot papers after the initial counting of the votes. We do not include recounting of ballot papers at the initial stage of the tabulating process; although that is not a trivial prerogative, we consider here the recounting of ballot papers by a person or body other than those responsible for the initial count. By 'recount/ we mean here not the mere arithmetical task of revising the tally of votes but a fresh scrutiny of the votes and/or the possibility of revising a previous decision with respect to the validity of any or all ballot papers. Table 6.3 lists the various bodies empowered to conduct or to order the conducting of a recount.5 We find 37 countries in our sample that provide for recounts. No less than 31 of them use Australian-type ballot papers, while six use party ballots and envelopes. All established democracies for which we could find information allow for recounts. The body in charge of a recount is most often a judge or court of justice. In Belgium, the House of Representatives can recount ballot papers. In Micronesia, the counting and tabulating committee, which did the initial counting of the votes, may be asked to recount ballot papers. However, if a petition for recount is denied, the candidate may appeal to the Supreme Court. In some countries, recounts are done automatically when certain conditions are met. For example, in Bangladesh, each returning officer recounts ballot papers declared invalid by the presiding officers. Canada provides for an automatic judicial recount whenever the majority of the leading candidate is lower than one-thousandth of the valid votes. In some other countries (for example, Hungary), such automatic recount is also provided for at the initial count.

152 Establishing the Rules of the Game Table 6.3 Recounts and petitions

Country Argentina Australia Bahamas Bangladesh Barbados Belgium Belize Benin Bolivia Brazil Bulgaria Canada Cape Verde Chile Costa Rica Cyprus Czech Republic Denmark Ecuador Estonia France Germany Guyana Hungary India Ireland Israel Italy Jamaica Japan Latvia Lithuania Luxembourg Madagascar Malawi Mali

Bodies empowered to conduct or order recount of ballot papers

Election Election petitions may petitions may be tried by be tried by ordinary courts the legislature

District electoral commission Australian Electoral Officer Court of Disputed Returns Election court (two justices of Supreme Court) Returning officer Electoral court (three judges) House of Representatives Justice of the Supreme Court NA No recounts Regional electoral tribunal No recounts Judge of a provincial court A court (Tribunal da Comarca) NA NA Supreme Court NA Folketing (Parliament) NA Territorial committee or National Election Committee Judge or court NA NA NA High Court High Court NA Central constituency office Resident magistrate for parish NA NA Electoral committees NA Constitutional Court Electoral Commission High Court NA

No Yes

No No

Yes

No

Yes Yes No Yes Yes No No Yes Yes Yes Yes No Yes Yes No No Yes

No No Yes No No No No No No No No No No No Yes No No

Yes Yes Yes Yes Yes Yes No No Yes Yes Yes Yes No Yes Yes

No Yes No No No No Yes Yes No No Yes No Yes No No

Yes

No

Sorting Out Winners and Losers 153 Table 6.3

(Concluded)

Country Malta Micronesia Mongolia

Namibia Netherlands New Zealand Panama Papua New Guinea Philippines Poland Portugal Romania St Lucia St Vincent Samoa Sao Tome Slovakia Slovenia South Africa Spain Sweden Taiwan Trinidad and Tobago United Kingdom Uruguay Vanuatu Venezuela

Bodies empowered to conduct or order recount of ballot papers

Election petitions may be tried by ordinary courts

Election petitions may be tried by the legislature

Court Counting and tabulating committee, Supreme Court NA

Yes No

No Yes

Yes

NA Election Council District Court judge No recounts Electoral Commission National Court A court NA General count assembly NA Supreme Court High Court A judge General count assembly No recounts NA Electoral Commission Electoral Court NA Electoral Causes Committee NA Chief Election Officer High Court NA Supreme Court NA

Yes No Yes No Yes

Yes (presidential elections only) No Yes No No No

Yes Yes Yes No Yes Yes Yes Yes Yes No Yes

Yes No No No No No No No No Yes No

Yes No Yes Yes Yes No Yes No

No No No No No No No No

NA = Not available.

Although many cases are missing, one pattern is evident from our data. Former British colonies overwhelmingly empower courts to proceed to recounts (Table 6.4). In this respect, their legislation is similar to that of Britain, where the High Court recounts ballots.

154 Establishing the Rules of the Game Petitions Whenever it is alleged that the election campaign has been tainted with fraud or other irregularities, it is crucial that there be the possibility to petition a court, an electoral institution, or the legislative assembly in order to secure a review of the validity of the election result. In theory, findings of serious irregularities could lead to invalidation of the whole electoral process and to the calling of another general election. In practice, election petitions result either in the dismissal of the petition or in the indictment of a few individuals, the invalidation of the election in specific districts, and the replacement of the legislators whose election was tainted with fraud. In this section, we address two questions on the trial of election petitions. First, may an aggrieved party petition ordinary courts for redress? The procedure for adjudicating electoral disputes involves the judicial branch of government, which in principle operates independently and transparently. By 'ordinary courts/ we mean courts whose jurisdiction exceeds that of electoral tribunals. The petition is thus taken away from the electoral administration and placed with judges. This has symbolic and practical implications. Symbolically, the petition moves to a body whose area of jurisdiction is wider and that often has more prestige. Practically, it implies the participation of judges who, coming from outside the electoral hierarchy, can take a fresh look at the evidence.6 Second, may the validity of election returns be reassessed by the legislature? If so, the process is left in the hands of the people's representatives who are accountable to the public for their actions. This practice was common in earlier elections. The matter could be referred to an allparty committee that heard witnesses and made a final decision as to the legality of the election. As late as 1956, the election of quite a few legislators from the poujadiste party was declared invalid by the French National Assembly (see Williams 1972,164). Like any club, Parliament had the privilege of determining whether its members had been lawfully admitted. That practice is open to challenge on various grounds. First, any decision reached by a committee dominated by the governing party is bound to be denounced as the 'justice of the winner' rather than the outcome of a fair process. Second, it was observed that in practice, when elections were followed by dozens of serious charges against legislators from all sides, political parties tended to bargain secretly among themselves so that most of the charges were dropped (a practice known as 'saw-off'), thus securing the position of members of

Sorting Out Winners and Losers 155

the parliamentary club. Hence a tendency among electoral reformers to advocate the transfer of disputes over the validity of elections to administrative bodies or to the courts. An ordinary court of justice is empowered to decide on election petitions in no less than 45 countries (see Table 6.3). In general, the courts involved tend to be of a high status. In 13 countries, the Supreme Court may try election petitions; 11 former British colonies follow the United Kingdom's example by empowering the High Court to resolve electoral disputes; and 10 countries give the Constitutional Court jurisdiction over electoral matters. In the Philippines, election petitions for presidential elections are tried by the Supreme Court, while for elections to the House of Representatives they are tried by an Electoral Tribunal composed of three justices of the Supreme Court and six members of the House. The legislature or some of its members are empowered to try election petitions in only 12 countries, eight of them in Europe. The nonEuropean countries are Israel, Micronesia, Mongolia (for presidential elections only), and the Philippines (see the above description). In Germany, the privilege enjoyed by the Bundestag is mitigated by the possibility of an appeal to the Federal Constitutional Court. Some interesting patterns emerge (Table 6.4). First, former British colonies are prone to leave the trial of petitions to ordinary courts. In fact, only Israel does not use the ordinary courts. As a result, these former colonies are less likely to allow the legislature to meddle in the trial of election petitions. This last procedure is more likely in European nations (35 per cent do so). Most Latin American countries, as well as Romania and Sweden, do not fall in either category. Adjudication of electoral disputes in these countries is left to electoral commissions or to other electoral bodies set up for the conduct of elections. So even in Brazil, where the Superior Electoral Tribunal as well as regional electoral tribunals are involved, the trial of election petitions is still in the hands of the electoral administration. These tribunals are composed of judges, but they are also responsible for many other matters related to elections. In Sweden election petitions are tried by the Electoral Causes Committee, composed of a chair who is required to be, or to have served as, a judge, and six other members, all appointed by Parliament. In Cyprus, New Zealand, and Samoa, it is possible to trace back a vote to the person who cast it. This possibility is used when someone is convicted of corrupt practice or dual voting. When a guilty person's

Table 6.4 British colonialism, political rights, and election petitions Recounts done by ordinary courts

Election petitions may be tried by ordinary courts

Election petitions may be tried by the legislature

90%* (n = 20)

96%* (n = 23)

4%* (n = 23)

25% (n = 20)

58% (n = 38)

28% (n = 39)

61% (n = 18)

71% (n = 21)

27% (n = 22)

Other countries

57% (n = 23)

73% (n = 41)

15% (n = 41)

Total

59% (41 countries)

73% (62 countries)

19% (63 countries)

British colonialism Former British colonies Other countries Political rights Established democracies

Difference significant at 0.01 level (two-tailed f-test).

Sorting Out Winners and Losers 157

vote is to be disallowed, the court matches the consecutive number appearing beside the name of the elector in the electoral register with the number on the ballot paper. This is a minor breach in the secrecy of the vote, but it allows the result to be altered through the invalidation of an illegally cast ballot. In New Zealand, there are occasional complaints about the traceability of votes, but there have not been any known instances where the secrecy of the vote was infringed illegally.7 In practice, tracing a person's vote to invalidate it rarely occurs. CONCLUSION As with other issues in electoral administration, there is no single best way to count ballot papers. Yet there is a general consensus that the presence of party observers contributes to the integrity of the counting process, and so, in almost all the countries examined, party representatives may watch the actual counting of the vote. Conducting the decisive count of ballots on the spot at the polling station is also the rule in nearly two-thirds of democracies, while ten countries conduct as well a preliminary count at the polling station. There thus seems to be a perception that such a procedure reduces the risk of fraud. All but one of the countries where the decisive count takes place at the electoral-district level are former British colonies. There is no consensus, however, about whether the authority to declare a ballot paper invalid should be given to a single individual (in 27 countries) or to many (in 33 countries). So empowering a single official may put the process at risk if there is no possibility for parties or candidates to ask for a recount. Adjudicating electoral disputes is the last step in the electoral game. Most often, ordinary courts are empowered to recount ballot papers or try election petitions. In only 12 countries may the legislature try election petitions - a procedure that may cast doubts on the legitimacy of the process. Once again, however, much depends on how the law is actually applied. It is striking that there are few systematic differences in the procedures used by established and by non-established democracies. Two processes may be at work here. On the one hand, new democracies may have been keen to adopt rules that had been deemed necessary to prevent fraud in a democracy. On the other hand, established democracies may have come to believe that electoral fraud is not really a serious threat and may not have felt the need to review old electoral legislation.

Conclusion

Patterns and Future Directions

Like any other game, elections can be understood as a set of rules designed to constrain the actions of its players. We have shown the great variety of rules that govern the electoral process in contemporary democracies. It should be evident by now that there is no unique way to conduct free and fair elections. Democracies have developed substantially different rules for the electoral game. One might have thought issues pertaining to the administration of elections to be sorted out, yet most of them are anything but resolved. Free and fair elections can be held under various settings, and many issues about how to conduct a democratic election need re-examination. Consensus, however, does emerge on some aspects of election laws. For instance, every democracy recognizes universal adult suffrage, and almost all agree that adulthood begins at 18. Most countries reserve the right to participate in national elections - either as a voter or as a candidate - to their citizens only. Thus, even with the easing of barriers and the ever-increasing flow of individuals between countries, rulers in democracies still value a concept sometimes depicted as outmoded and exclude non-citizens from the political process. Almost every democracy uses electoral registers to ensure that all electors can vote. Concerned with transparency, most countries also allow political parties to appoint observers at the various stages of the electoral process. On other matters, democracies are nowhere near a consensus. For example, chapter 4, on the central electoral administration, shows that its personnel are appointed in widely varying ways. Special voting procedures have been devised, but besides assisted voting, countries have adopted different approaches. As with the study of electoral systems, scholars need to start by

Conclusion: Patterns and Future Directions 159

defining relevant dimensions and categorizing each set of rules on those dimensions. This is what we did with regards to electoral administration. We can distinguish countries in terms, for example, of those that require deposits from candidates or parties and those that do not or in terms of those where counting takes place at the polling station and those where it does not. Without such categorization, efforts to generalise any findings pertaining to electoral administration are pointless. PATTERNS Patterns in election law arise when we compare a vast array of countries. In general, former colonies tend to adopt the same rules as their former mother country, not unlike what we noted with regard to electoral formulas (Blais and Massicotte 1997). It might be tempting for colonies going through a democratization process to emulate what appears to 'work' in their former rulers' country. This especially holds true if the political elite has been trained in the mother country. Established democracies tend to be more inclusive than non-established ones. Voting rights and the right to be a candidate can be exercised by more people in established democracies. This reveals the importance of suffrage as a yardstick of a society's level of democracy. Established democracies allow for a wider array of special voting procedures than non-established ones. In recent years, democracies (both established and non-established) have expanded not only the right to vote, but also the opportunities and ways for voters to exercise that right. But even among established democracies, as we have seen, there is not widespread agreement on many issues. It seems perfectly acceptable to some to disfranchise prison inmates, for example, even if others consider this unfair. Besides inclusiveness, there are not many differences, however, between established and non-established democracies. This is one of our surprising findings. One would expect these two groups of countries to diverge on many dimensions, but they seem to converge more often than not. Maybe the explanation lies in the fact that countries within each category cannot agree among themselves. Also, new democracies may be inclined to emulate established ones. Their leaders may assume that international observers will deem their electoral arrangements more credible if they are consonant with practice in older democracies.

160 Establishing the Rules of the Game FUTURE DIRECTIONS There remain many puzzling findings that warrant further investigation. For instance, one-quarter of democracies do not require electors to show an identification document before being allowed to vote. This is surprising, because it is relatively easy to impersonate someone else, especially with the rural exodus of the last century and the relative anonymity of large cities. In the same vein, how do we explain the persistence of French-type ballot papers, given the substantial advantages that the Australian-type ballot seems to offer? Another puzzling trend, or rather lack thereof, is that compulsory voting does not seem to gain or lose any ground. We might expect that with the general decline of deference in democratic societies (Nevitte 1996), voters would resist being told by political authorities that they do not have the right not to vote. Still, three-tenths of democracies oblige electors to go to the polls. It will be interesting to see whether, in the next decades, compulsory voting will be able to weather the liberal emphasis on individual rights and liberties. One important aspect of elections that we could not deal with in as much depth as we would have liked is the independence of election authorities. We have provided some descriptive information and have looked more closely at the situation prevailing in two countries, Australia and Canada. But clearly we have only scratched the surface here. We need more detailed case studies of how election authorities are appointed and of how and why these nomination procedures affect their independence, which appears to be a fundamental prerequisite of a fair election. Another aspect that needs further examination pertains to the process for challenging election results. It should be clear, in the light of the U.S. presidential contest in 2000, that the legitimacy of the electoral game hinges to a large extent on what happens after election night. We have stressed from the very start the exploratory character of our study. We hope to have provided a useful overview of the wide variety of rules established by contemporary democracies, of the main arguments advanced for and against the various options, and of the broad sociological patterns that may explain why some rules tend to be more popular in certain types of countries. Obviously, much more work needs to be done on factors that have led to the adoption of particular rules in specific countries. Electoral rules are endogenous; they are determined by the very same people who will be governed by them in

Conclusion: Patterns and Future Directions 161

the future. The macroscopic perspective that we have employed has prevented us from examining in detail the political dynamics that come into play and, more specifically, the strategies adopted by the various political parties, both in power and in opposition, about determining the rules of the game. How and why, for instance, are parties that feel disadvantaged by existing rules able to change them? What is the relative importance of values such as fairness and pragmatic party interest in the actual determination of electoral rules? We also need detailed studies about the consequences of election laws. There is some evidence, for instance, that turnout is enhanced by compulsory voting and postal voting. The next step is probably to assess to what extent the various electoral rules contribute to the overall legitimacy of the electoral process. After all, electoral arrangements can be considered to be successful only if they are perceived to be fair by all (or the overwhelming majority), especially by the losers of the electoral process. We thus need strong survey evidence about how voters themselves feel about who should and should not have the right to vote and to be a candidate, who should nominate election officials, how the vote should take place, and how it should be counted (see, for instance, Blais and Gidengil 1992). We also need good cross-national studies indicating in which countries elections are most and least likely to be perceived as fair. We excluded the United States and Switzerland from our study for good reasons: election laws vary from one state (canton) to the other. Further research should definitely look at the variations within these countries, as they would provide a natural experiment for assessing the consequences of the 'other' election laws. Especially in the United States, important questions pertaining to the administration of elections have been raised in the last few years. The debates on whether felons should be permanently disfranchised or not, on the practicability of all-mail elections and internet voting, and on the role of the state in registering electors have raised many of the issues examined in our study (not to mention the omnipresent debate on campaign-finance reform). Clearly, then, there are many questions that we have not addressed. But we hope that our study will put the conduct of elections once again at the forefront of the political science agenda.

This page intentionally left blank

Appendix

Countries, Sources, and Variables Included in the Study

List of the 63 Countries Covered Argentina Australia Bahamas Bangladesh Barbados Belgium Belize Benin Bolivia Brazil Bulgaria Canada Cape Verde Chile Costa Rica Cyprus Czech Republic Denmark Ecuador Estonia France

Germany Guyana Hungary India Ireland Israel Italy Jamaica Japan Latvia Lithuania Luxembourg Madagascar Malawi Mali Malta Micronesia Mongolia Namibia The Netherlands New Zealand

Sources on Election Law Choe 1997 Constitutions

Panama Papua New Guinea The Philippines Poland Portugal Romania St Lucia St Vincent and the Grenadines Samoa Sao Tome and Principe Slovakia Slovenia South Africa Spain Sweden Taiwan Trinidad and Tobago United Kingdom Uruguay Vanuatu Venezuela

164 Appendix Electoral laws and regulations Government documents Interparliamentary Union's internet website (www.ipu.org) Interviews with experts on specific countries LeDuc, Niemi, and Norris 1996 Nix 1995 Responses to our questionnaire Description and Sources of Independent Variables Former British Colonies

We deemed countries to be former British colonies if they were once ruled by Britain: Australia Bahamas Bangladesh Barbados Belize Canada Cyprus Guyana

India Ireland Israel Jamaica Malawi Malta Namibia New Zealand

Papua New Guinea St Lucia St Vincent and the Grenadines Samoa South Africa Trinidad and Tobago Vanuatu

Established Democracies

We assign a dummy variable that equals 1 for those countries that had, according to Freedom House 1997, a political rights rating of 1 uninterruptedly from 1987 to 1996: Australia Barbados Belgium Belize Canada Costa Rica Cyprus Denmark

France Germany Ireland Italy Luxembourg Malta Netherlands New Zealand

Portugal St Lucia Spain Sweden Trinidad and Tobago United Kingdom

Appendix

165

Other Countries

Argentina Bahamas Bangladesh Benin Bolivia Brazil Bulgaria Cape Verde Chile Czech Republic Ecuador Estonia Guyana Hungary

India Israel Jamaica Japan Latvia Lithuania Madagascar Malawi Mali Micronesia Mongolia Namibia Panama Papua New Guinea

Philippines Poland Romania St Vincent and the Grenadines Samoa Sao Tome and Principe Slovakia Slovenia South Africa Taiwan Uruguay Vanuatu Venezuela

This page intentionally left blank

Notes

Introduction: Establishing the Rules 1 Other causes may generate that kind of paradoxical outcome. For example, a candidate may lose even if he or she has more votes when his or her votes are too heavily concentrated in a few states. 2 The Administration and Cost of Elections (ACE) Project - a joint endeavour of the Institute for Democracy and Electoral Assistance (IDEA), the International Foundation for Electoral Systems (IFES), and the United Nations includes much relevant information on the existing legislation in democracies. This database can be reached, in English, French, and Spanish, at the following internet address: http://www.aceproject.org 3 The same approach is followed by Cox (1997) and Blais and Dobrzynska (1998). 4 See the list of countries in the Appendix. 5 The information is, to the best of our knowledge, correct as of December 1999. Whenever possible, we cross-checked our information with all other available sources. Chapter 1: Who Has the Right to Vote? A first version of this chapter appeared in Electoral Studies 20 (2001), p. 41-62. 1 Autonomy is distinct from competence when it is defined in financial terms. But restrictions on the basis of financial dependence no longer prevail. 2 We find no countries with a maximum voting age. 3 Alhough this is outside the ambit of our inquiry, for municipal elections

168 Notes to pages 17-28

4

5

6 7

8

9

only, five German Lander now enfranchise citizens from the age of 16. In each case, candidates must be at least 18. Katz 1997, 218-29, enumerates instances where similar distinctions based on marital status were found: Brazil (1824-89), Chile (1833-77), Colombia (1821-18 and 1853-86), Ecuador (1884-1946), Honduras (1894-1957), Mexico (1932-69), Peru (1860-1933), Portugal (1826-78), and Uruguay (18301918). Britain's 1918 legislation enfranchised men at 21 and women at 30. This anomaly was removed in 1928. Such a double standard existed in Ireland between 1918 and 1923. In Finland between 1869 and 1906 farmers were enfranchised at 21 and town residents at 24. In most countries the electoral law does not specify the details of disqualification based on mental deficiencies. Procedures for obtaining citizenship vary among countries, and it would be important to compare how demanding their requirements are. To our knowledge, no study of this kind exists on this topic, which is beyond the scope of the current volume. In Australia, the franchise is extended to British subjects whose names were included on an electoral roll before 26 January 1984. The law does not extend the 'Commonwealth clause' to more recent newcomers. Formerly, Canada also granted the right to vote to other British subjects; in 1970 it narrowed this right to those who were eligible to vote at the 1968 election, and the right lapsed in 1975. The experience of these two countries suggests that 'Commonwealth clauses' might disappear in countries where they exist. Furthermore, the Maastricht treaty sets rules for elections to the European Parliament and local elections. EU citizens residing in another EU country may vote at local and European elections in their country of residence. We also find countries that allow non-citizens to vote in local and regional elections. In Denmark, citizens of non-EU countries may vote in elections for county and municipal councils, if they have been residing in Denmark for three years prior to election day. In Ireland, non-EU citizens may vote in local elections only. In the Netherlands, non-Dutch nationals may vote in municipal and provincial elections, if they satisfy certain requirements. In Sweden, citizens of EU countries, Iceland, and Norway may vote in elections for county and municipal councils under the same conditions as Swedish nationals; citizens of other states may vote for county and municipal councils if on election day they have been registered residents of Sweden for three years uninterruptedly. In Venezuela, non-citizens who have been residing for at least 10 years in the country may vote in local elections.

Notes to pages 29-39 169 10 The New Zealand Electoral Act specifies that those who do not meet the residence requirement retain the right to vote in the previous district of residence. The same rule applies in Australia and, presumably, in other countries. 11 We consider only those cases that explicitly require residence in the country. 12 In Sweden, there is no minimum period of country residence required. However, one must have resided in the country at some point in order to have the right to vote. 13 We did not include Uruguay in the calculation of the median residence requirement, because it distinguishes three categories of electors. 14 When there were different time limits for different types of electors, we took into account the one that applied to the majority of the electorate. 15 In Canada, the electoral law removes the right to vote of prisoners sentenced to imprisonment for two years or more. This restriction was declared unconstitutional in 1996, and as a consequence prison inmates were allowed to vote in the 1997 federal election. The government launched an appeal, however, and won its case. Finally, in November 2002, the Supreme Court of Canada, by a 5-to-4 decision, decided that all prisoners should be enfranchised. 16 In the United States, 14 states impose permanent disfranchisement on their convicts (Economist, 24 Oct.1998, 30). 17 See also Mike Steketee, 'Voluntary Voting a Big Boost for Coalition,' Australian, 8 Aug. 1996, 2. The article is based on a paper prepared by Malcolm Mackerras and Ian McAllister for the Australian Election Study. 18 They are Argentina, Belgium, Bolivia, Brazil, Chile, Costa Rica, Ecuador, Italy, Panama, Portugal, Uruguay, and Venezuela. 19 These are countries where a majority of the population is Catholic, according to the Encyclopaedia Universalis 1997. 20 Countries that do not follow this pattern are Australia and Cyprus - two former British colonies located in Oceania and Asia, respectively - where voting is compulsory. Australia instituted compulsory voting because low turnouts (between 45 per cent and 60 per cent of votes) obliged parties to divert resources to take voters to the polls. Compulsory voting was also assumed to increase honesty in elections (for a full compendium of the arguments raised for and against compulsory voting in the Australian Parliament and state legislatures, see Hughes 1966, 81-3). As for Cyprus, the existence of compulsory voting may be explained by the influence of Greece, where voting is compulsory. 21 For an interesting argument for compulsory voting, see Lijphart (1998) and Hill (2002).

170 Notes to pages 40-59 Chapter 2: Who Has the Right to Be a Candidate? 1 However, in Madagascar, women who have received citizenship by marriage become eligible only ten years thereafter. 2 Provisions obliging holders of judicial positions to retire at a specified age are common. For example, judges of the highest courts must retire at 70 in Australia and 75 in Canada. In the latter country, 75 is also the retirement age for members of the (appointed) Senate. 3 Lower house of Parliament only. 4 The Australian Constitution prevents those who are not citizens of Australia exclusively from sitting in either House of Parliament. During the debate over a constitutional amendment that would have extended the same rule to future presidents, it was pointed out that this provision actually excluded as many as one-quarter of Australians; Tony Abbott, 'In Republic, Some Are More Australian Than Others/ Australian, 17 Sept. 1999,15. 5 In another, Namibia, candidates must be citizens by birth or descent. 6 See, respectively, section 38(1) of the German Basic Law, section 67 of the Italian constitution, and section 152(3) of the Portuguese constitution. 7 Standing in more than one electoral district was common practice for politicians during the 19th century either as a self-promoting political tactic or, more prosaically, as an insurance policy against defeat. In France, Adolphe Thiers's election in 1871 in about 30 departments was instrumental in bringing his appointment as president. In Canada, standing in two districts was standard practice among party leaders, until it was prohibited in 1919 (Ward 1963). 8 Most countries with mixed electoral systems allow another kind of double candidacy, in which candidates can stand in single-member districts as well as on party lists. The proposal put forth in Britain in 1998 by the Report of the Independent Commission on the Voting System (chaired by Lord Jenkins of Hillhead) stated that double candidacies should be allowed, but not encouraged. Ecuador and Panama forbid such double candidacies. 9 A few countries are missing. In many instances, the electoral law and the constitution are silent on this topic. Consequently, we did not perform any quantitative analyses. 10 Some scholars in the public choice literature have gone even further, suggesting that because of conflict of interest, public-sector employees should not have the right to vote (Tullock 1970). In the province of Quebec (Canada), both federal and provincial civil servants were disfranchised from 1889 to 1895, while municipal and provincial policemen were disfranchised for even longer (see Bernard and Laforte 1969, 94; Gow 1986, 73-4).

Notes to pages 60-74 171 11 In some countries, such as Barbados and Luxembourg, the absence of elected regional structures guaranteed no incompatibility, while the documents that we consulted did not mention such an incompatibility for other countries. 12 In this vein, low public trust of politicians in Australia has led to suggestions that members of Parliament who resign between general elections should be obliged to pay for the cost (estimated at A$300,000) of the ensuing by-election; see Fia Gumming, 'We Pay for MPs to Retire Early/ SunHerald, 3 Oct. 1999, 28. 13 In Namibia, the deposit for independent candidates is half the amount for party candidates. 14 Deposits are non-refundable in Mali, Micronesia, and Vanuatu. 15 This provision was inserted in 1993 and was in force during the period examined in this book. In 2000, the law was amended so that the refund of a candidate's deposit no longer depends on the number of votes received. The deposit is refunded in full if the law's reporting requirements are met. 16 In the Quebec provincial election of 1998, for example, no fewer than 58 candidates (out of 657) got less than 100 votes; the law on nomination papers required 100 signatures. 17 Using the data from the 1997 legislative election, this would amount to some 40,000 signatures. 18 There is an old joke in Latin America that the mother of a newly elected president, when interviewed about her son's political success, exclaimed : 'Had I known he would become president, I would have sent him to school.' 19 The presence in Baltic countries of a sizeable contingent of unilingual Russian speakers probably explains this requirement. Chapter 3: Registering Voters 1 Different countries use different terminologies: electoral list, voters' register, electoral roll. 2 Voters show up on election day at a polling station and get their passport stamped, whereupon they may cast a vote. A similar method was used in the South African elections in 1994. 3 Independent of the formal registration period, there is almost always the possibility of registering at some time before election day. When we categorize a country's registration process as 'continuous,' 'annual/ or 'monthly/ we do so on the basis of first registration. 4 It has been estimated that a nationwide law allowing registration on elec-

172 Notes to pages 74-116

5 6 7 8

tion day in the United States would increase turnout by 5 percentage points (Fenster 1994). They can exercise their right to vote at an electoral contest once they have attained the voting age. In three countries, one finds a combination of party representatives and observers. However, the first registration under the new system was done on the basis of a door-to-door canvass. This is the general principle set out in the act. However, if new electors request their own inclusion through their federal tax statement, no confirmation will be demanded.

Chapter 4: Who Conducts the Election? 1 What we call a 'commission' may have different names: Electoral Bureau, Electoral Committee, Electoral Council, Electoral Court, Electoral Office, Electoral Tribunal. Regardless of its appellation, we consider a 'commission' to be a body composed of more than one person, appointed for the sole purpose of supervising the electoral process. 2 In Uruguay, five members of the nine-member Electoral Court are elected with a two-thirds' majority vote. 3 Following the latest (1999) election, this meant as many as twelve parties. 4 This information was collated before the referendum held on 15 December 1999 that modified the constitution. 5 Moreover, in Germany and Vanuatu, the authority is shared between a public official and a commission. 6 In 1997, a general election was called after the start in Manitoba of the worst flood witnessed during the 20th century, and a party leader demanded the application of that clause. The chief electoral officer visited the flooded areas and concluded that there was no evidence that it would be impracticable to carry out the provisions of the act in any of the nine or ten constituencies affected. As it happened, polling went smoothly in the areas affected. Chapter 5: How People Vote 1 The reason for multiple days of polling in India is given by the Electoral Commission: Tolling is normally held on a number of different days in different constituencies, to enable the security forces and those monitoring the election to keep law and order and ensure that voting during the election is fair' (http://www.eci.gov.in/infoeci/elec_sys/elecsys_fs.htm).

Notes to pages 118-27 173 2 However, the 2002 election was held on a Friday, like the previous one. 3 We excluded India because different electoral districts have different polling hours. We excluded Papua New Guinea for lack of data. 4 We excluded Slovenia for lack of information. 5 This is based on the dates of Madagascar's presidential election in 1996. Neither the constitution nor the electoral law stipulates the exact number of days separating the two rounds. 6 This rule grants entitlements to parties based on strength in electoral districts rather than nationwide, which means that an opposition party may be entitled to nominate the deputy returning officer and the government party the poll clerk. The government party may even not be entitled, in areas where it is weak, to nominate either official. Entitlements are based on the results of the previous election - a feature that may complicate implementation if and when the party system is changing rapidly. 7 Chile formerly had an original practice of providing for men and women to be entered in distinct registers and to vote in distinct boxes. The existing law still allows this practice, excepcionalmente. 8 Brennan and Pettit (1990), however, argue that open voting is preferred to secret voting because it provides maximum incentives for electors to vote responsibly. In their own words, '(the secret vote) is a blemish in the democratic crown, not a jewel' (1990, 328). 9 This device was invented to prevent a fraudulent practice known as 'telegraphing,' whereby a corrupt voter, in the secrecy of the polling booth, substituted for the blank ballot paper provided by election officials a ballot paper already marked on behalf of a candidate and then, outside the polling station, handed the former back to the agent who had corrupted him or her as evidence that he had 'secretly' voted for the 'right' candidate. In turn, this blank ballot paper was marked for the candidate and given to another corrupt voter, and the chain went on; hence the expression 'telegraphing.' 10 'Write-in' ballot papers - i.e., ballot papers on which voters must write down the name of the candidate for whom they are voting - are used in some jurisdictions in exceptional circumstances. For example, Canadian voters overseas use this procedure because it is cumbersome, even impossible, to supply them in time with a specimen of the ballot paper used in their respective constituencies; voters are, however, given a list of the candidates standing in their constituencies. A variant is used in Estonia, where lists of candidates are posted in the polling place and the polling booth. Electors write on the ballot paper the registration number of the candidate for whom they vote. In the Philippines, electors must write the name of the various candidates that they support for the various offices to be filled.

174 Notes to pages 128-38 11 This excludes instances where electors have to indicate their preferences on the ballot paper; in this case, the validity of a ballot may still be subject to a later review. 12 There is another balloting procedure, formerly in use in India and Africa under British rule in the 1950s: voters were offered a ballot, but behind the voting screen they had to choose from among numerous ballot boxes, one for each candidate. One voted by inserting the ballot in the box corresponding to the preferred candidate. 13 See the news release 'National Semiconductor and Unisys Equip Brazil with New Voting Machines for Fast and Accurate Election Results in the Fall,' issued by the National Semiconductor Corporation on 6 May 2002, on website http:/ /www.national.com/news/item/0,1735,757,00.html, accessed on 18 October 2002, and Holli Riebeek, 'Brazil Votes Electronically' on website Wysiwyg:/ /13/http:/ /www.spectrum.ieee.org/ WEBONLY/wonews/oct02/brazil.html, accessed on 18 October 2002. A detailed description outlining the advantages of the procedure has been issued by the Brazilian government: see 'The Electronic Ballot Box,' on website http:/www.mre.gov.br/boletim/urna-i2.htm, accessed on 18 October 2002. 14 Roll-off occurs when electors indicate their preference for the higher positions to be filled and leave blank the lower contests on the ballot. 15 We include Estonia in this group. 16 Not a typical British colony, as Palestine was under a League of Nations mandate. 17 Another procedure is to allow electors to vote in a polling station other than the one to which they were assigned. We did not survey the use of this option, since it generally does not change the steps to be followed by electors when casting a ballot. 18 'Absentee voting' is also used, mostly in the American literature, to describe postal voting. 19 Allegations in France that, in the close parliamentary election of 1973, postal votes cast by electors overseas had been directed by unscrupulous officials to districts where they could be expected to tip the balance in favour of the ruling party led to the abolition of this procedure a few years later (see Masclet 1992, 91). 20 This figure excludes public funding of candidates. These data can be found on the Australian Electoral Commission's website (http:/ / www.aec.gov. au). 21 Until 1993 in Canada, in order to remedy the problem that voters might be unable to vote on advance polling days as well as on polling day, a stand-

Notes to pages 13&-57 175 ing advance poll was opened at the office of the returning officer each day during the weeks preceding polling day. Chapter 6: Sorting Out Winners and Losers 1 We call 'vote count' the actual counting, one by one, of ballots papers. Under this definition, the mere tabulation of previous counts does not amount to a vote count. 2 Some presiding officers open the box and turn it upside down. This practice is strongly discouraged, as it may generate confusion if some ballots fall to the ground and may foster fraudulent behaviour. 3 At the British referendum on Europe in 1975, the decision was made to count ballots not in constituencies, as is normally done for parliamentary elections, but in 68 larger, regional subdivisions - i.e., English and Welsh counties, Scottish regions, and Northern Ireland as a whole. This was decided by a free vote in the House of Commons against the government's preference for a nationwide count and against the preference of some MPs for a constituency count (Butler and Kitzinger 1976, 55). This decision spared MPs the inconvenience of seeing their own stand on the issue eventually 'repudiated' by their own constituents. At the referendums on devolution in 1997, votes were counted in local government areas in Scotland and in county or county boroughs in Wales. 4 In Canada, a member of the Quebec legislative assembly elected in 1966 saw his election being challenged in a court. His election was voided in 1971, one year after he had been defeated in a general election. 5 Information on recounts is at times difficult to obtain, as electoral legislation is not always clear on the matter. 6 We reckon that the line between an 'ordinary' court and an electoral administration composed of judges, as exists in Brazil, is sometimes hard to draw. 7 Private communication with Dr Paul Harris, chief executive of the New Zealand Electoral Commission.

This page intentionally left blank

References

Accoce, Pierre, and Pierre Rentchnick. 1996. Ces malades qui nous gouvernent. Paris: Stock. Beaud, Olivier. 1992. 'Le droit de vote des etrangers: 1'apport de la jurisprudence constitutionnelle allemande a une theorie du droit de suffrage.' Revue frangaise de droit administratif 8: 409-24. Benoit, Kenneth, and John W. Schiemann. 2001. 'Institutional Choice in New Democracies: Bargaining over Hungary's 1989 Electoral Law.' Journal of Theoretical Politics 13:153-82. Bernard, Andre, and Denis Laforte. 1969. La legislation electorate au Quebec 17901967. Montreal: Editions Ste-Marie. Blais, Andre. 1988. 'The Classification of Electoral Systems.' European Journal of Political Research 16: 99-110. - 2000. To Vote or Not to Vote? The Merits and Limits of Rational Choice Theory. Pittsburgh: University of Pittsburgh Press. Blais, Andre, and Agnieszka Dobrzynska. 1998. Turnout in Electoral Democracies.' European Journal of Political Research 33: 239-62. Blais, Andre, Agnieszka Dobrzynska, and Louis Massicotte. 2001. 'Les variables institutionnelles affectent-elles la participation electorate en regime democratique?' Ottawa: Elections Canada. Blais, Andre, and Elisabeth Gidengil. 1992. Representative Democracy: The Views of Canadians. Ottawa: Royal Commission on Electoral Reform and Party Financing. Blais, Andre, and Louis Massicotte. 1997. 'Electoral Formulas: A Macroscopic Perspective.' European Journal of Political Research 32: 107-29. - 2002. 'Electoral Systems.' In Lawrence LeDuc, Richard G. Niemi, and Pippa Norris, eds., Comparing Democracies 2: New Challenges in the Study of Elections and Voting. Thousand Oaks: Sage Publications.

178 References Boix, Charles. 1999. 'Setting the Rules of the Game: The Choice of Electoral Systems in Advanced Democracies.' American Political Science Review 93: 609-24. Bollen, Kenneth. 1993. 'Liberal Democracy: Validity and Method Factors in Cross-National Measures.' American Journal of Political Science 37:1207-30. Brennan, Geoffrey, and Philip Pettit. 1990. 'Unveiling the Vote.' British Journal of Political Science 20: 311-33. Burkhart, Ross E., and Michael Lewis-Beck. 1994. 'Comparative Democracy: The Economic Development Thesis.' American Political Science Review 88: 903-10. Butler, David E., and Bruce Cain, eds. 1992. Congressional Redistricting: Comparative and Theoretical Perspectives. New York: Macmillan. Butler, David, and Uwe Kitzinger. 1976. The 1975 Referendum. London: Macmillan. Butler, David, Howard R. Penniman, and Austin Ranney. 1981. 'Introduction: Democratic and Nondemocratic Elections.' In David Butler, Howard R. Penniman, and Austin Ranney, eds., Democracy at the Polls: A Comparative Study of Competitive National Elections. Washington, DC: American Enterprise Institute for Public Policy. Chang Mota, Roberto. 1998. 'Utilizac.ao da urna eletronica (Brasil 1996).' In Roberto Chang Mota and Francisco Silvino De Jesus Ferreira Matos, eds., La automatizacion de los processes electorates. San Jose, Costa Rica: Institute Interamericano de Derechos Humanos, Centro de Aseroria y Promocion Electoral (CAPEL). Choe, Yonhyok. 1997. How to Manage Free and Fair Elections: A Comparison of Korea, Sweden and the United Kingdom. Goteborg: Goteborg University. Courtney, John C. 2001. Commissioned Ridings: Designing Canada's Electoral Districts. Montreal: McGill-Queen's University Press. Cox, Gary W. 1997. Making Votes Count: Strategic Coordination in the World's Electoral Systems. Cambridge: Cambridge University Press. Cross, William. 1996. 'Direct Election of Provincial Party Leaders in Canada, 1985-1995: The End of the Leadership Convention?' Canadian Journal of Political Science 29: 295-315. Cunningham, Dayna L. 1991. 'Who Are to Be the Electors? A Reflection on the History of Voter Registration in the United States.' Yale Law and Policy Review 9: 370-404. Dahl, Robert. 1989. Democracy and its Critics. New Haven, Conn.: Yale University Press. De-Mesquita, Ethan Bueno, 2000. 'Strategic and Nonpolicy Voting: A Coalitional Analysis of Israeli Electoral Reform.' Comparative Politics 33: 63-80. Denoncourt, Yves. 1991. 'Reflections on Criteria for Excluding Persons with

References

179

Mental Disorders from the Right to Vote.' In Michel Cassidy, ed., Democratic Rights and Electoral Reform in Canada. Vol. 10 of the Royal Commission on Electoral Reform and Party Financing. Toronto: Dundurn. Dion, Stephane. 1986. La politisation des mairies. Paris: Economica. Fenster, Mark J. 1994. 'The Impact of Allowing Day of Registration Voting on Turnout in U.S. Elections from 1960 to 1992. A Research Note.' American Politics Quarterly 22: 74-87. Franklin, Mark N. 1999. 'Electoral Engineering and Cross-National Turnout Differences: What Role for Compulsory Voting?' British Journal of Political Science 29: 205-24. - 2002. 'The Dynamics of Electoral Participation.' In Lawrence LeDuc, Richard G. Niemi, and Pippa Norris, eds., Comparing Democracies 2: New Challenges in the Study of Elections and Voting. London: Sage. Freedom House. 1997. Freedom in the World. New York: Freedom House. Gastil, Raymond D. 1979. Freedom in the World: Political Rights and Civil Liberties. New York: Freedom House. Gosnell, Harold F. 1935. 'Voting.' in The Encyclopaedia of the Social Sciences, vol. 15. London: Macmillan & Co. Gow, James Iain. 1986. Histoire de I'administration publique quebecoise 1867-1970. Montreal: Presses de 1'Universite de Montreal. Hamilton, Randy H. 1988. 'American All-Mail Balloting: A Decade's Experience.' Public Administration Review 48: 860-6. Helliwell, John F. 1994. 'Empirical Linkage between Democracy and Economic Growth.' British Journal of Political Science 24: 225-48. Hill, Lisa. 2002. 'On the Reasonableness of Compelling Citizens to "Vote": The Australian Case.' Political Studies 50: 80-101. Hughes, Colin A. 1966. 'Compulsory Voting.' Politics 1: 81-95. Jackman, Simon. 1999. 'Non-Compulsory Voting in Australia? What Surveys Can (and Can't) Tell Us.' Electoral Studies 18: 29-48. Jaensch, Dean. 1995. Election! How and Why Australia Votes. St Leonards: Allen & Unwin Pty Ltd. Kaminski, Marek M. 2002. 'Do Parties Benefit from Electoral Manipulation? Electoral Laws and Heresthetics in Poland, 1989-93.' Journal of Theoretical Politics 14: 325-58. Katz, Richard S. 1997. Democracy and Elections. Oxford: Oxford University Press. - 2000. The "Other" Aspects of Electoral Systems: Empirical Research and Democratic Theory/ Paper presented at the 18th World Congress of the International Political Science Association, Quebec City, 1-5 Aug. Laponce, Jean Antoine. 1969. People vs. Politics: A Study of Opinions, Attitudes,

180 References and Perceptions in Vancouver-Burrard, 1963-1965. Toronto: University of Toronto Press. LeDuc, Lawrence, Richard G. Niemi, and Pippa Morris. 1996. 'Introduction: The Present and Future of Democratic Election.' In Lawrence LeDuc, Richard G. Niemi, and Pippa Norris, eds., Comparing Democracies: Elections and Voting in Global Perspective. Thousand Oaks: Sage Publications. Lijphart, Arend. 1994. Electoral Systems and Party Systems: A Study of Twentyseven Democracies, 1945-199Q. Oxford: Oxford University Press. - 1998. 'Unequal Participation: Democracy's Unresolved Dilemma.' American Political Science Review 91:1-14. Lutz, Georg, and Dirk Strohmann, 1998. Wahl- und Abstimmungsrecht in den Kantonen. Droits politiques dans les cantons. Bern: Verlag Paul Haupt. Mackenzie, W.J.M.. 1958. Free Elections. London: George Allen & Unwin. Magleby, David B. 1987. 'Participation in Mail Ballot Elections.' Western Political Quarterly 40:79-91. Masclet, Jean-Claude. 1992. Le droit des elections politiques. Paris: Presses Universitaires de France. Massicotte, Louis, 1999. 'Parliament in the 1990s.' In James Bickerton and Alain-Gustave Gagnon, eds., Canadian Politics, 3rd ed. Peterborough, Ont: Broadview Press. Massicotte, Louis, and Andre Blais. 1999. 'Mixed Electoral Systems: A Conceptual and Empirical Survey.' Electoral Studies 18: 341-66. McLean, Iain, and David E. Butler, ed. 1996. Fixing the Boundaries: Defining and Redefining Single-Member Electoral Districts. Aldershot: Dartmouth. Mill, John Stuart. 1972. Considerations on Representative Government. London: J.M. Dent & Sons Ltd. Nadeau, Richard, and Andre Blais. 1993. 'Accepting the Election Outcome: The Effect of Participation on Losers' Consent.' British Journal of Political Science 23: 553-63. Nevitte, Neil. 1996. The Decline of Deference: Canadian Value Change in CrossNational Perspective. Peterborough, Ont.: Broadview Press. Nevitte, Neil, ed. 2002. Value Change and Governance in Canada. Toronto: University of Toronto Press. Nichols, Stephen M., and Gregory A. Strizek. 1995. 'Electronic Voting Machines and Ballot Roll-off.' American Politics Quarterly 23: 300-18. Nix, Stephen B., ed. 1995. Electoral Law Compendium of Central and Eastern Europe. Kiev: International Foundation for Electoral Systems. Oliver, J. Eric. 1996. 'The Effects of Eligibility Restrictions and Party Activity on Absentee Voting and Overall Turnout.' American Journal of Political Science 40: 498-513.

References

181

Planinc, Zdravko. 1987. 'Should Imprisoned Criminals Have a Constitutional Right to Vote?' Canadian Journal of Law and Society 2:153-64. Powell, G. Bingham, Jr. 1986. 'American Voter Turnout in Comparative Perspective.' American Political Science Review 80:17-43. Pye, Lucian W. 1985. Asian Power and Politics: The Cultural Dimension of Authority. Cambridge, Mass.: Belknap Press of Harvard University. Rae, Douglas W. 1967. The Political Consequences of Electoral Laws. New Haven, Conn.: Yale University Press. Report of the Royal Commission on the Electoral System. 1986. Towards a Better Democracy. New Zealand. Robertson, Gerald B. 1994. Mental Disability and the Law in Canada. 2nd ed. Scarborough, Ont: Carswell. Robson, William A. 1923. 'Compulsory Voting.' Political Science Quarterly, 38: 569-77. Southwell, Priscilla L., and Justin Burchett. 1997. 'Survey of Vote-by-Mail Senate Election in the State of Oregon.' PS: Political Science and Politics 30: 53-7. - 2000. 'The Effect of All-Mail Elections on Voter Turnout.' American Politics Quarterly 28: 72-9. Sparrow, Andrew. 2000. 'Voting by Post Boosts Turnout.' Daily Telegraph, 12 May. Taagepera, Rein, and Matthew Soberg Shugart. 1989. Seats and Votes: The Effects and Determinants of Electoral Systems. New Haven, Conn.: Yale University Press. Tullock, Gordon. 1970. Private Wants, Public Means: An Economic Analysis of the Desirable Scope of Government. New York: Basic Books. Ward, Norman. 1963. The Canadian House of Commons: Representation. 2nd ed. Toronto: University of Toronto Press. Williams, Philip M. 1972. Crisis and Compromise: Politics in the Fourth Republic. London: Longman.

This page intentionally left blank

Index

Administration and Cost of Elections (ACE) Project, 167n.l advance voting, 79, 132, 134-5, 137-8 Africa, 11, 38, 64, 129 African League of Human Rights, 89 Argentina, 18, 35, 38, 42, 50, 53, 56-8, 68, 73, 76, 84, 103, 118, 120, 124, 126, 129, 134, 144; National Electoral Chamber, 84, 103; vote counting practices of, 152 Aruba, 31 Asia, 11, 17, 38, 53, 55, 73, 129 Australia, 18, 28-9, 31-6, 38, 42, 60, 67-8, 74, 76, 80-1, 84, 94, 99, 103, 117-18, 121, 124, 133-5, 139, 144; Aboriginal and Torres Strait Islander Commission, 100; alternative vote, 147; Australian Constitutional Convention, 136; Australian Electoral Commission, 80, 84, 99, 103; Australian electoral officer (AEO), 100, 160; Commonwealth Electoral Roll, 100; Electoral Roll Review, 100; electoral office, 99; Federal Court, 99; Joint Standing Commission of Parliament on

Electoral Matters, 101; Public Service Act, 100; two-candidate preferred count, 147; vote counting practices of, 146-7, 152 Austria, 123 Bahamas, 18, 42, 62, 68, 84, 96, 103, 124, 134, 144; Parliamentary Commissioner, 84, 96; Public Service Commission, 84, 96; vote counting practices of, 146, 152 ballot and envelope model. See ballots, French model ballot boxes, 122-3, 125-9, 133,1389, 141-2, 145; electronic, 129; prevention of loss of, 146; prevention of tampering with, 146; transportation of, 146 ballot papers, 13, 122-32, 136,13842; alternatives to, 6, 123-5,12730, 139, 141; distribution of, 124-7; invalidation of, 147,149; marking of, 123, 125-7; official tally sheet, 142; recasting of, 138; recounting of, 127, 141-3; tampering with, 129; validity of, 149, 151; voiding of, 125, 127; with party symbols,

184 Index 123-7, 130-2, 141; with pictures of candidates, 123-7, 130-2, 141 ballot papers, counting of, 142-5, 157; at polling stations, 143, 157; political party observers to, 143, 149, 157; recounting of, 150-2, 157 ballot validity, 142-3, 151, 157; as determined by a single official, 143, 150,157 ballots, Australian model, 123,125-9, 131-2, 141-2, 160; cost of, 127; counting of, 127, 142; manipulation of, 123, 127; marking of, 123, 125, 127; recounting of, 127, 142, 151; validity of, 127; voiding of, 125, 127 ballots, French model, 61, 123, 126-9, 131-2, 141-2, 160; cost of, 127; counting of, 127; official tally sheet, 127, 142 Bangladesh, 18, 42, 57, 68, 84, 94, 103, 124, 134, 144; Election Commission, 84; vote counting practices of, 151-2 Barbados, 18, 28-9, 31, 42, 62, 68, 84, 94, 103, 124, 134, 144; Electoral and Boundaries Commission, 84, 103; vote counting practices of, 152 Belgium, 18, 32, 35, 38, 42, 68, 73, 84, 97, 103, 118, 123-4, 134, 139, 144; Minister of Interior, 84; vote counting practices of, 146, 151-2 Belize, 18, 28, 30, 42, 68, 84, 104, 124, 131, 134, 144; Election and Boundaries Commission, 84, 104; vote counting practices of, 152 Benin, 18, 42, 50, 55-6, 58, 63-4, 68, 74, 82, 85, 95, 104, 118, 124, 126, 129, 132, 134, 139, 144, 146; Autonomous National Electoral Com-

mission, 85, 94, 104; ballot boxes, transportation of, 146; College des Magistrals du siege, 85; Commission of Human Rights, 85, 94; I'Assemblee Generates des MagisPtrats, 94; Magistrat du siege de I'Ordre judiciaire elu par, 94 Bolivia, 17, 19, 31, 35, 38, 43, 50, 55, 60, 68, 85, 94, 104, 123-4, 126, 134, 144; National Electoral Court, 85, 104; vote counting practices of, 152 Brazil 6, 17, 19, 28, 31-2, 35-6, 38, 43 50, 53, 55, 57-8, 60, 68, 74, 85, 94, 104, 123-4, 128-9, 134, 141, 144; election petitions in, 155; Superior Electoral Court, 85; Superior Electoral Tribunal, 155; Superior Justice Court, 85; Supreme Federal Court, 85; vote counting practices of, 152 Britain (see also United Kingdom): former colonies 14, 26, 28-30, 32, 36-7, 39, 53-4, 56, 58, 61, 65, 75, 81, 95- 6, 118-19, 122, 129, 130, 137-8, 153; former colonies, defined, 164 Bulgaria, 19, 27, 43, 50, 60, 63-4, 68, 85, 104, 117, 120, 124, 126, 134, 144; Central Election Commission, 85, 104; Municipal Election Commission, 104; Sectional Election Commission, 104; vote counting practices of, 152 Bush, George W., 7 Canada, 6, 19, 27, 30-1, 43, 60, 62, 678, 73, 78-9, 82, 83, 85, 96-7, 105, 120-1, 123-4, 133-4, 136, 138-40, 144, 160; Canada Elections Act, 8, 98; chief electoral officer, 85, 97-8; Customs and Revenue Agency, 78;

Index 185 Elections Canada, 98-9; Public Record Office, 79; Record Office for Persons Detained in Penal Institutions and Persons at Large, 79; Report of the Canadian Royal Commission on the Electoral System, 16; Royal Commission on the Electoral System, 16; vote counting practices of, 151-2 candidates, election, 3, 11, 40-65; age restrictions, 40-54, 58, 65; citizenship restrictions, 40-52, 54-6, 58, 65; conflicts of interest, 41, 59; corruption, 59; deposit reimbursement, 42-52, 62-3; disqualification of clergy, 65; dual mandate, 59-60; education requirements, 64; election expenses, 62-3, 65; election mandate incompatibilities, 40-9, 58-60, 65; incompatibilities with public service, 41-9, 59-60, 65; ineligibilities, 40, 42-9, 64; legislative, 13, 61-2, 65; monetary deposits, 40-52, 54, 61-3, 65; nominations, 57, 64; presidential, 61, 63-5; residence restrictions, 40, 42-52, 54, 56-8, 65; restrictions implicit in voting laws, 41; supporting signatures, 40-1, 50-2, 63-5 Cape Verde, 19, 35-6, 38, 43, 50, 64, 68, 85, 94, 105, 124, 126, 132, 134, 144; National Election Commission, 85; vote counting practices of, 152 carpetbaggers, 57 Catholic church, 37, 95 Chile, 19, 28, 30, 35, 38, 43, 50, 53, 578, 64, 68, 85, 96,105,117,124,126, 134,144; Departmental Electoral

Court, 105; Director of Electoral Service, 85; vote counting practices of, 152 Commonwealth, 18-24,28,30,37,45, 48 Commonwealth clause, 28-30, 55-6 Communism, fall of, 4 Communist countries, former, 5,11 compulsory voting, 16, 33-9,100, 136,160-1 constitutions, 10,12, 27, 35, 41, 56-7, 65,120 Costa Rica, 19, 30, 35, 38, 43, 50, 58, 69,86,95,105,120-1,124,126,132, 134,144; Supreme Tribunal of Elections, 86; vote counting practices of, 152 Cote d'lvoire, 56 Cyprus, 19,35,38,43,50,60-1,69,86, 96,105,118,120-1,124,126,132, 134,144; election petitions in, 155; General Returning Officer, 86; vote counting practices of, 152 Czechoslovakia, 33,116 Czech Republic, 19, 31, 44, 60, 69, 86, 94,106,116-17,124,129,134,139, 144; Central Election Commission, 86; Ward Election Board, 106 democracy: definition of, 6,12, 37; degree of, 11 democratic countries, 4,11,14; established, 5,12,14,17,26-30,32,35-9, 53-4,56, 58, 65, 74-5, 81-2,95-6, 119,122,130-2,136-40,143,147, 151,157; non-established, 5,12,14, 17,26-30, 32, 35-9, 53-4, 56, 58,65, 74-5,81, 95-6,101,119,122,130-2, 136-40,143,149,157 Denmark, 19, 43, 60, 69, 86, 97,106,

186 Index 117-18,121,124,134,144; Minister of the Interior, 86; vote counting practices of, 147,152 deposits. See candidates, election disfranchisement. See voting rights dual mandate, 59-60 Dutch Antilles, 31 Ecuador, 19, 35, 39,43, 50, 55, 57-8, 69,86, 94,106,120,124,126,132, 134,144; Ballot Reception Committee, 106; National Electoral Tribunal, 86; Provincial Electoral Tribunal, 106; Supreme Electoral Tribunal, 106 election laws, 4-6,8-13,30,35,41,83; analysis of, 3; consequences of, 161; patterns in, 159 election observers: from political parties, 102-15,121,143; international, 5,10,143,160; local, 10 election officials, 10,133; appointment of, 160; independence of, 160; nomination procedures, 160-1 election results, 5,11,142-3; announcement of, 142,145; challenging of, 143,150,154,160; determination of, 13; invalidation of, 13,151; petition of, 13,142,151, 154; secrecy of, 146 elections: conduct of, 13; corruption of, 122,150; cost of, 8,136; credibility of, 8, 9; fraud, 5, 9, 66, 74,101, 122,129,133,136,150,154; freedom and fairness of, 10, 34, 58, 83, 98,101,122-3,150,158; legitimacy of, 9-10, 34, 83,155 elections, legislative, 28, 31, 41-9, 53-

8,102,116-18,120,123,128-32, 136,141,143 elections, presidential, 28, 31,41, 503, 58,102,113,116-17,120,123, 128-9,131-2,136,141,143 electoral administration, 11,13, 83102,121,157-8; appointed by government, 83-91, 93-6, 99; appointed by judiciary, 83-6, 88, 90-1, 94—5; appointed by legislature, 83, 85-98 electoral authority, 83-101,127,154; electoral commission, 5, 83-96,99101; governmental ministers, 83-4, 86-8, 94-7; public officials, 83-7, 89-91, 96-8; term of office, 84-94, 97, 99-100 electoral disputes, 155 electoral districts, 29,127 electoral process, 8-9, 66,158 electoral reform, 5-6, 8,116 electoral register (see also voter registration), 11-12, 66-82, 99-100, 102, 121,133,136,138,158; accuracy of, 66-7; cost of, 66; date of closure, 68-72, 74, 76, 77, 79-82; establishment of, 3,13; exclusion from, 73, 138; exclusion of prison inmates, 67, 77, 79-80; inclusiveness, 66, 734, 76; political party involvement in, 66, 68-72, 75-8, 80-1,158; updating of, 13, 67-73 Electoral Register Office: Canada, 80; Spain, 79 electoral systems: proportional representation (PR), 29, 37, 61,129, 143; studies of, 4 electronic voting machines, 6,123-5, 128-30,141 England. See United Kingdom

Index 187 Estonia, 20, 27, 44, 62, 64, 69, 86,106, 124,134-5,138,144; National Election Commission, 86; vote counting practices of, 152 Europe, 11, 64, 73, 74, 97,118,129; Council of, 31, 33 European Union, 28 franchise. See voting rights France 20, 29, 31-2, 44, 50, 53, 60-1, 63-5, 67, 69, 73-4, 86, 97,107,120, 123-4,126,129,134,139,144; former colonies, 64; Minister of the Interior, 86; National Institute of Statistics and Economic Studies, 73; petition of 1956 election, 154; vote counting practices of, 152 Freedom House, 10-12,164 Germany, 20, 30-2, 44, 56-7, 60, 67, 69, 73, 76-7, 86, 95,107,121,123-4, 129,134,144; Electoral Board, 107; Federal Returning Office, 86; Land government, 107 Gore, Al, 7 Guyana, 20, 27-8, 44, 50, 55, 61, 64, 69, 74, 76, 87,107, 117,124,126, 134, 138, 144; Election Commission, 87,107 Haiti: Duvalier regime, 127 Hungary, 20, 44, 69, 82, 87, 94, 107, 120-1,124,134,144; Minister of Home Affairs, 94; National Electoral Committee, 87; vote counting practices of, 151 India, 20, 32, 44, 53, 57, 60, 62, 69, 87, 94,108, 116-18,122,124,132,134,

144; Election Commission, 87,108; vote counting practices of, 152 ineligibilities, of candidates, 40,42-9, 64 Ireland, 20, 27-8,45, 51, 61-2, 64, 69, 73,87,95, 97,108,117-18,121,124, 126,134,138,144; Minister of the Environment, 87; single transferable vote, 147; vote counting practices of, 152 Israel, 20,45, 60-1, 69, 73, 87, 95,108, 117,124,129,134,138,143,144; Central Committee, 108; Central Elections Committee, 87, 95; District Committees, 108; election petitions in, 155; Knesset, 95; Polling Committees, 108 Italy, 20,27, 31, 35, 39, 45, 57, 69, 87, 94,108,124,129,134,144; National Central Electoral Office, 87; vote counting practices of, 152 Ivory Coast, 56 Jamaica, 21, 27-8, 45, 55, 69, 73, 87, 109,121,124,134,138,144; Chief Electoral Office, 87; vote counting practices of, 146,152 Japan, 17,21, 32, 45, 53, 70, 73, 88, 109,124,134,144; vote counting procedures, 146 Latvia, 21, 45, 62, 64, 66, 70, 88, 95, 109,117-18,124,134,144; Central Electoral Commission, 88 Lithuania, 21, 45, 51, 63, 70, 82, 88, 109,120,124,126,134,136,144; Central Electoral Committee, 88; Lawyers' Society, 88; Minister of Justice, 88; vote counting practices of, 152

188 Index Luxembourg, 21,35,39,46,65,70,88, 97,109,123-4,134,144; Minister of the Interior, 88 Maastricht treaty, 168n.9 Madagascar, 21, 46, 51, 63, 64, 70, 88, 110,120,124,126,134,144; National Electoral Commission, 88; vote counting practices of, 152 Malawi, 21,28, 30, 46, 51, 56, 61, 64, 70, 73-4, 89,110,117,124,126,129, 134,144; Electoral Commission, 89,110; Political Parties Act 1993, 89; vote counting practices of, 152 Malaysia, 17 Maldives, 17 Mali, 21, 29, 31-2, 46, 51, 64, 70, 74, 89, 95,110,120-1,124,126,134, 144; AGEMPEM (Protestant group), 89; Autonomous Union of the Magistracy, 89; CAFO (women's group), 89; Human Rights Association, 89; Independent National Electoral Commission, 89, 94 Malta, 21, 32, 46, 58, 62, 70, 89, 94, 110,117-18,124,134,138,144; Electoral Commission, 110; vote counting practices of, 146,153 Marcos, Ferdinand, 10 Micronesia, 22, 46, 53, 56,58, 70, 76, 89,96,110,124,134,144; election petitions in, 155; National Election Director, 89; vote counting procedures, 146,151,153 Mongolia, 22,46, 51, 55, 61-3, 70, 89, 94, 111, 124,126,131,134,144; election petitions in, 155; General Election Committee, 89

Namibia, 22, 32,46,51, 56, 61, 63-4, 70, 90, 94, 111, 116-18,124,126, 134,144; Electoral Commission, 90, 94-5, 111; High Court, 94-5; Supreme Court, 94-5 Netherlands, 22,27, 31, 33, 37,47,60, 62, 70, 90, 94, 111, 125,134,143-4; Electoral Council, 90; vote counting practices of, 153 New Zealand, 22,28-9, 31,47, 70, 734, 76, 90, 96, 111, 117-18,123,125, 133-4,136,141,145; 1993 election, 7; Chief Electoral Officer, 90; election petitions in, 155,157; Electoral Act, 169n.lO; National Party, 7; Secretary for Justice, 96; State Service Commission, 96; vote counting practices of, 146,153 North America, 11, 38 Northern Ireland, 29 Norway, 123 Oceania, 11, 73 Pakistan, 17 Panama, 22, 29, 35, 39,47, 51, 56-8, 70, 76, 90, 94, 111, 120-1,125-6, 135,145; Electoral Tribunal, 90, 111; vote counting practices of, 153 Papua New Guinea, 22, 29, 47, 57-8, 60, 62, 70, 74, 90, 111, 116-17,125, 131,135,145; Electoral Commission, 90, 111; Electoral Commission Appointments Committee, 90; Public Services Commission, 90; vote counting practices of, 153 Philippines, 10,22,29,31-2, 35-9,47, 51, 53, 55, 57-8, 60, 64, 67, 71, 73, 90, 94,112,117,121-2,125-6,132, 135,145; Commission on Appoint-

Index 189 ments, 90; Commission on Elections, 90,112; election petitions in, 155; vote counting practices of, 153 Poland, 22,47,51,58,63,71,82,91,95, 112,118,120-1,125-6,129,135,145; Constitutional Tribunal, 91; district electoral commissions, 121; executive committee of communes, 121; National Electoral Commission, 91; Supreme Administrative Court, 91; Supreme Court, 91 poll book, 122 polling day, 13,102-15,128,136,1389,141; as a public holiday, 105,108, 112-13,115,117-18; cost, 116-18, 120-1; days between rounds, 10215,118,120; employees time off on, 102-15,117-19; hours of operation, 102-15,118-19,138,141; irregularities, 121; manipulation of, 116; number of polling days, 102-16, 138,141; on a rest day, 102-15, 117-19,141; on a week day, 102-15, 117-18; political party observers, 102-15,121; voter identification, 102-15, 119, 121-2, 141; voter impersonation, 121 polling officials, 98,102-15, 120, 126, 141; appointment of, 103-15; committees, 121; from political parties, 120,149; polling officers, 121,138, 145 polling stations, 122,127,132,135-6, 138-42,146-7; advance, 79,132, 134-5,137-8; at diplomatic missions, 122; in post offices, 122,132, 138; in schools, 117; location of, 146; mobile, 132, 134-5,137,139; special, 77,122,132 Portugal, 23, 28, 31-2, 35, 37-9, 47,

51, 57, 60-1, 65, 71, 73, 91, 95,112, 118,125-6,132,135,145; former colonies, 37-8, 61, 65; National Elections Commission, 91; vote counting practices of, 153 postal voting. See voting, postal proxy voting. See voting, by proxy Romania, 23, 47, 51, 63, 71, 74,91, 95, 112,125-6,131,135,145; Central Electoral Bureau, 91, 95; election petitions in, 155; Supreme Court, 94; Ward Election Board, 112 St Lucia, 23, 28, 30-1,48, 56, 58, 64, 71, 91,112,125,135,138,145; Electoral Commission, 91,112; vote counting practices of, 146,153 St Vincent and the Grenadines, 23, 28, 30, 48, 55, 64, 71, 91,113,125, 135,145; Public Service Commission, 91; Supervisor of Elections, 91; vote counting practices of, 146, 153 Samoa, 17, 23, 30, 48, 62, 71, 91, 96, 113, 117,125,135,145; election petitions in, 155; Public Service Commission, 91, 96; vote counting practices of, 146-7,153 Sao Tome and Principe, 23, 32, 35-6, 38-9, 48, 52, 64, 71,92, 94,113,1256,132,135,145; vote counting practices of, 153 Singapore, 17 Slovakia, 24, 31, 48, 52, 60, 71, 82, 92, 95,113,116-17,125-6,129,131,135, 145; Divisional Electoral Commission, 113; Electoral Commission, 92; vote counting practices of, 153 Slovenia, 24, 48, 52, 55, 71, 92,113,

190 Index 125-6,135,145; State Electoral Board, 92 South Africa, 24, 48, 62, 71, 92,113, 117,125,135,139,145; Commission on Gender Equality, 92; Constitutional Court, 92; Electoral Commission, 92; Human Rights Commission, 92; vote counting practices of, 153 South America, 55, 61, 64-5,136 South Korea, 17 Spain, 24, 32, 38, 48, 60-1, 65, 67, 71, 74, 76, 79-80, 92, 95,114,125,129, 135-6,145; Central Electoral Commission, 79, 92; electoral law of 1907, 38; Estado autonomico, 60; former colonies, 36, 37, 38, 55, 61, 65,129; General Council of Judicature, 92; National Statistical Office, 79 Sweden, 24,27, 32, 49, 71, 73, 92, 95, 114,125,129,132,135,138-40,145; election petitions in, 155; Electoral Causes Committee, 155; National Tax Board, 92, 95, vote counting practices of, 147,153 Switzerland, 11, 33,123 Taiwan, 17,24,28,49,52,55,57-8,60, 63-4, 71, 93-4,114,117,125-6,135, 145; Central Election Commission, 93 Trinidad and Tobago, 24, 28, 30, 49, 57-8, 72, 93,114,125,135,145; Elections and Boundaries Commission, 93,114; vote counting practices of, 153 United Kingdom, 6, 24,28-9, 31-3, 37, 49, 65, 72, 93, 95, 97,114,117-

18,122-3,125,129,135,139,145; Electoral Commission, 95; Report of the Independent Commission on the Voting System, 170n.8; Secretary of State for Home Affairs, 93; vote counting practices of, 1467,150,153 United States, 11, 33, 41, 56, 67,123, 128,133,136; 2000 presidential election, 7, 9,160; electoral college, 7; federal voting rights acts, 7; former colonies, 36; voting restrictions, 7 Uruguay, 24,28, 30, 35, 39, 49, 52, 67, 72-3,82, 93-4,115,125-6,129,135, 145,147; Electoral Court, 93, 172n.2; electoral junta, 115; Reception Committee, 115 Vanuatu, 25,29,49, 72,93,115,117, 125,129,135,145; Electoral Bureau, 93,115; Electoral Commission, 93; Judicial Service Commission, 93; National Council of Chiefs, 93; Public Service Commission, 93; vote counting practices of, 153 Venezuela, 25, 31-2, 35, 39,49,52,55, 65, 72, 93, 95,115,125-6,129,135, 145; National Electoral Council, 93,115 vote counting, 10,127,142-5,157; at the electoral-district level, 143,145; at the national level, 143,145; at the polling station, 143,146; by party members, 149; decisive, 143, 146-9; preliminary scrutiny, 146; recounts, 150-2 voter registration, 8, 66-82,161 (see also electoral register); automatic,

Index 191 67; by proxy, 78; collective applications, 77; compulsory, 67-72, 74-5, 79-81; continuous, 66, 68-72, 75, 78, 80; cost, 67, 73; house-to-house canvass, 68-75, 80-1; impersonation, 121; in Australia, 80-1; in Canada, 78-9; in Germany, 76-7; in Spain, 79-80; on election day, 74, 79, 81-2; of citizens abroad, 77-9, 81; of disabled voters, 77; of the military, 77, 79; periodic, 67-72, 78-9; premature registration, 76, 81; sanctions for not registering, 74, 80; self-initiated, 66-73, 80; state-initiated, 66-79; under-registration, 67, 74; voluntary, 67, 77-8, 81 voters, illiterate, 127,131,139,141 votes: alternative, 147; manipulation of, 5,10; single transferable vote (STV), 143,147 voting: advance, 132, 134-5,137-40; alternative, 143; as legitimization of electoral process, 8; assisted, 132,134-5,138-9,141,158; by fax, 132,140-1; by internet, 132,140-1, 161; by phone, 132, 140; by proxy, 116,132, 134-5,137, 139-40; compulsory, 16, 33-9, 100,136,160-1; decline in turnout, 6, 33-4,117, 120,136; electronic, 6,124-5,12830,141; fraud, 8,10, 31,129,133, 136, 138, 149, 151,157; postal, 6, 78-9,116, 132-8,140,161; procedure for, 11,123; sanction exemptions, 35,100; sanctions for not voting, 16, 33, 35-6, 38-9,100;

secret, 122-3,127,133,139,141; special procedures for, 13,102,116, 123,132-41,143,158 voting age: arguments for lowering, 17; minimum, 17-26, 37,158 voting machines, 6,123-5,128-30, 141 voting rights, 3,11,13,161; and level of democracy, 159; based on community membership, 15,29, 30-2; based on competence, 15-27, 37; by age, 12,16-26, 37,158; by citizenship, 12,15-16,18-28, 37,158; by country of residence, 12,16,1826,28,30-2,37; by electoral district of residence, 12,15-16,18-26,29, 37; by period of residence, 28-30, 37; consensus on, 16, 36; for African Americans, 7, 8; for citizens, 12,15-16,18-28, 37; for citizens abroad, 13,18-26, 30-2, 37,133; for citizens by birth, 6, 8,18-28, 30-1, 37; for disadvantaged groups, 7; for naturalized citizens, 18-28, 30-1, 37; for non-citizens, 6, 8,15,18-28, 30-1, 37; for the mentally handicapped, 6,13,15-16, 18-27, 32,37; for the military, 30-1, 133,138; for the poor, 7,15; for prison inmates, 6, 8,13,15-16,1826, 32-3, 37,41,159,161; for women, 7,15; historical importance of, 7; universal suffrage, 12, 15,158 West Germany, 53