1 - Introduction

Back to Cybervote Reports List

• 1.1 Scope of analysis
• 1.2 What is an online voting system?
• 1.3 Desirability of Internet Voting
• 1.4 Feasibility of Internet Voting
• 1.5 Method of Analysis

Report on electronic democracy projects, legal issues of Internet voting and users (i.e. voters and authorities representatives) requirements analysis

Synopsis

This document is the second volume of the report on electronic democracy projects, legal issues of Internet voting and users (i.e. voters and authorities representatives) requirements analysis.

The purpose of this volume is to analyse the legal implications of using a cyber voting system in Europe.

File name

KUL-WP2-D4V2-v1.0.doc

Amendment History

Version	Date	Description
0.1	22 September 2000	First draft outline for comment
0.2	2 October 2000	Use of the CyberVote template
0.3	9 October 2000	Second draft outline for comment
0.4	5 December 2000	First draft
0.5	22 December 2000	Second draft
0.6	24 January 2001	Third draft
0.7	28 February 2001	First final draft
0.8	23 March 2001	Updated draft
0.9	28 March 2001	Modification by ICRI according review comments
0.91	4 April 2001	Modification by KISTA according review comments
0.10	15 May 2001	Modification by ISSY according review comments
0.11	28 March 2001	Modification by BREMEN according review comments
0.12	17 May 2001	2nd final draft
1.0	01 June 2001	Final version sent to the EC

Disclaimer

The information in this document is provided as is and no guarantee or warranty is given that the information is fit for any particular purpose. The user thereof uses the information at its sole risk and liability.

Scope of CyberVote

CyberVote, an innovative cyber voting system for internet terminals and mobile phones, is a research and development (RDT) programme being funded by the European Commission, with additional funding from the companies and organisations undertaking the work. It is part of the Information Society Technologies (IST) 1999 programme for research, technology development and demonstration under the fifth framework programme (5^th PCRD). It is attached to Key Action 1 "Systems and Services for the Citizens".

The objective of CyberVote is to demonstrate the first highly secure cyber-voting prototype using mobile and fixed internet technologies. The project will define and implement a CyberVote prototype embedding an innovative voting protocol relying upon the use of advanced cryptographic tools that will be developed to ensure integrity, privacy and authentication of the voters. The prototype will be demonstrated and evaluated on 3 trial applications.

The project is carried out by a consortium led by MATRA Systèmes & Information (FR) and grouping together: British Telecommunications (UK), NOKIA Research Centre (FI), K.U.Leuven Research & Development (BE), Technische Universiteit Eindhoven (NL), Freie Hansestadt Bremen (DE), Mairie d'Issy-les-Moulineaux (FR) and Kista Stadsdelsnämnd (SE).

The project officially started on 1 September 2000. It will end on 1 March 2003. The overall budget of the project is 3 243 629 and the total effort is 27.4 man-years.

Executive summary

This document is the second volume of the report on electronic democracy projects, legal issues of Internet voting and users (i.e. voters and authorities representatives) requirements analysis.

The purpose of this volume is to analyse the legal implications of using a cyber voting system in Europe.

This report assesses if an online voting system can comply with existing election regulations. Both polling place voting systems and remote online voting systems are considered.

The analysis mainly focuses on describing the legal requirements CyberVote will have to comply with. The analysis also explains how the legal system should be modified in order to enable implementation of CyberVote.

An important part of this report considers if online voting could conform to the basic election rights laid down in international and regional convents and in the national constitutions: non-discrimination, freedom and secrecy of the vote, one-person-one-vote, and security, reliability, verifiability and confidentiality of the voting system.

Our investigation shows that it will be difficult for a remote online voting system to comply with a stringent application of the requirements of "the vote to be free and secret "and of "one person to be able to vote only once". However, exceptions to these requirements actually do exist (vote by proxy and vote by mail for instance) and there seems to be a tendency to gradually extend the field of application of these exceptions.

Also other legal requirements for democratic public elections are analysed. To that end, the current voting procedures in a few Member States are described.

We see that these prescriptions correspond with and are limited to the election techniques used (paper ballots, polling stations, voting booths, electronic voting machines, vote by mail, vote by proxy, manual count, etc.) and can therefore not simply be applied to (polling place or remote) online voting procedures.

These existing rules do however provide us with a guideline when considering the implementation of an online voting system.

TABLE OF CONTENTS

1 Introduction *

1.1 Scope of analysis *

1.2 What is an online voting system? *

1.3 Desirability of Internet Voting *

1.4 Feasibility of Internet Voting *

1.5 Method of Analysis *

2 Elections: Basic principles *

2.1 The Principle of Non-discrimination *
2.1.1 General *
2.1.2 Regulation *
2.1.3 Discussion *
2.1.4 Conclusion *

2.2 The Principle of Democratic Elections *
2.2.1 General *
2.2.2 Regulation *
2.2.3 Discussion *
2.2.4 Conclusion *

2.3 Summary: Legal requirements for CyberVote *
2.3.1 Equality, non-discrimination and fairness *
2.3.2 Democratic elections *
2.3.3 General *

3 The Election regulations today *

3.1 Introduction *

3.2 The Netherlands *
3.2.1 Introduction *
3.2.2 Traditional Voting Procedure *
3.2.3 Alternative Voting Procedures *
3.2.4 Possibilities of Internet Voting in the Netherlands *

3.3 Belgium *
3.3.1 Introduction *
3.3.2 Traditional Voting Procedure *
3.3.3 Alternative Voting Procedures *
3.3.4 Possibilities of Internet Voting in Belgium *

3.4 France *
3.4.1 Introduction *
3.4.2 Traditional Voting Procedure *
3.4.3 Alternative Voting Procedures *
3.4.4 Other Important Regulations *
3.4.5 Possibilities of Internet Voting in France *

3.5 Sweden *
3.5.1 Introduction *
3.5.2 Existing Voting Procedures *
3.5.3 Other Important Regulations *
3.5.4 Possibilities of Internet Voting in Sweden *

3.6 Germany *
3.6.1 Introduction *
3.6.2 Traditional Voting Procedures *
3.6.3 Alternative Voting Procedures *
3.6.4 Possibilities of Internet Voting in Germany *

3.7 Summary: Legal requirements for CyberVote *
3.7.1 General *
3.7.2 Election legislation *

4 Summary of the characteristics of the European voting systems *

5 Conclusion *

6 References *

7 Points of Contact for further information *

1 INTRODUCTION

This document is the second volume of the report on electronic democracy projects, legal issues of Internet voting and users (i.e. voters and authorities representatives) requirements analysis.

The purpose of this volume is to analyse the legal implications of using a cyber voting system in Europe.

1.1 Scope of analysis

The present report assesses if the use of an online voting system for public elections could comply with existing election regulations.

In this analysis, public elections (or political elections) are defined as general elections for citizens' representatives in local, regional, national or supranational legislative bodies. These elections are subject to particular procedural laws and rules concerning for instance ballot secrecy, fairness, verifiability, accuracy, democracy, etc.

1.2 What is an online voting system?

First of all, the subject of present report has to be clearly identified: what is an online voting system?

Elections may be organised in many different ways. Paper-based elections make use of paper ballots, while automated elections make use of some kind of voting machines which automate the voting and/or tabulation procedures. When these voting machines are computers, we talk about electronic voting.

Electronic voting systems may be further divided into offline and online voting systems. In offline voting systems the computers used may in essence be viewed as a stand-alone computer. In online voting systems the computers used are connected by a (closed or open) network, leading to an essential distinction between clients and servers.

Electronic Democracy (or e-Democracy) systems, which may in turn be viewed as parts of e-Government systems, form a broad class of systems related to public services targeted at informing citizens on political issues. Such systems may also cater for online discussions between citizens, possibly involving politicians too. These systems may also include mechanisms for online polls, which may be used for conducting informal surveys, without seeking a high level of accuracy of the result. Online voting systems are much more formal than online polling systems, because they seek (or should seek) to accurately reflect the voters' preferences.

The present report focuses on online (electronic) voting systems. If the network used is the Internet, this fact may be emphasised by using the term Internet voting instead.

The California Internet Task Force (The Internet Voting Report¹) describes Internet voting as a voting process that enables voters to cast a secure and secret ballot over the Internet. Two main types of Internet voting systems can be recognised: polling place Internet voting and remote Internet voting.
A polling place Internet voting system uses Internet voting computers at traditional polling places staffed by election officials who assist in the authentication of voters before ballots are cast.
A remote Internet voting system uses unsupervised Internet voting computers to cast a ballot over the Internet using a computer not necessarily owned and operated by election personnel. CyberVote also extends to voting by Mobile Phone. As far as the legal analysis is concerned, Mobile Phone voting is similar to remote Internet voting and will therefore be treated likewise.
The technical architecture of these two systems may differ substantially, because the technological and security concerns are more daunting in the second system.
The polling place voting system would not require a digital authentication: the authentication can be done physically, similar to traditional or electronic election. After the voter is authenticated, she can cast his/her vote anonymously. The authentication procedure and voting process for polling place Internet voting system would therefore (to a certain extent) not differ substantially from the electronic elections organised today. However, the use of digital signatures based on public/private keys² could be important to enhance the reliability of the election process. For instance, election officials would no longer have the possibility to illegitimately cast votes instead of voters who have not presented at the polling station. However the acts of the election officials are controlled, not only by their colleagues, but also by all voters present at the polling station and by the witnesses (which can appointed by the candidates in the election). When the votes are collected and/or counted electronically, it could prove to be easy to illegitimately manipulate the electronic vote collecting and/or vote tabulating machines without someone noticing. The use of an authentication method, based on the public/private key pair, could therefore also provide an additional guarantee for polling placed voting systems as to the reliability of the election procedure.
The remote Internet voting system would require electronic (for instance digital) authentication: the voter would need to have a personal key (password, digital signature) to distantly identify him/herself as a legitimate voter. The authentication is necessary to guarantee the one man, one vote principle. But at the same time, the link between the authenticated voter and the ballot cast must be cut, so as to make it not retraceable.
It is clear that a digital authentication procedure requires a highly sophisticated technical solution. The authentication is indeed described as one of the most difficult challenges for an Internet voting system.
The California Internet Voting Task Force therefore believes that election officials should initially test Internet voting technology through the use of Internet voting machines that are under direct control of election personnel in traditional polling places. Then if remote voting is eventually adopted, it should initially be modelled on the current absentee ballot.
According to the California Internet Voting Task Force, an Internet voting system should undoubtedly be introduced gradually and should therefore be implemented in four stages:

· Phase one: supervised use of an Internet Voting Machine

_ Stage one: Internet voting at voters' polling place

This is the polling place Internet voting system. Voters are required to vote at an official polling station in their home precinct.

_ Stage two: Internet voting at any polling place

A polling place Internet voting system is used. However, voters are allowed to vote at any official polling place within a county or at centrally located Internet polling locations available to all voters in the county.

· Phase two: Remote Internet Voting

_ Stage three: Remote Internet voting from county computers or kiosks

The voter is provided a password and/or a digital signature from the election officials and can use any polling place established throughout the community. Poll workers are not necessary for voter authentication, so voting can take place at any time the Internet voting machines are open for use.

_ Stage four: Remote Internet voting from any Internet connection

Remote Internet voting (same as stage three), and the voters are also allowed to vote on their own computers (at home or work) or on their mobile phones.

Each of these four stages provide greater convenience to voters, but each step requires increasingly advanced technology and security solutions: voter authentication, ballot secrecy, ballot integrity, reliable vote transport and storage, prevention of multiple voting, defence against attacks on Internet voting machines and systems, etc.³

1.3 Desirability of Internet Voting

Today, the Internet offers a vast range of opportunities, not only for private and business, but also for governmental application.

First of all, the Internet is an ideal medium to provide citizens with information of all kinds: not only practical information and information of general interest, but also databases of regulations, laws and jurisprudence. In the near future, it is likely that citizens will be able not only to find and download forms on the Internet, but also to fill them out online, using a digital signature.

The Internet also enhances transparency of the legislative process. In Belgium and the Netherlands, it is actually possible to follow online the activity of Parliament: proposals, questions and answers, debates, commissions' work, etc. are provided on the Internet.

Participation is another important condition for a genuine democracy. However, we have to acknowledge that a lot of people do not regularly participate in public elections today. In Belgium for example, however voting is obliged, 9.4% did not vote in the 1999 elections.

Couldn't participation be increased by allowing citizens to use the Internet to cast their vote?

For an everyday growing number of people, the Internet indeed becomes an established part of their way of living. Therefore, Internet voting would certainly fit in well with today's so-called information society. This society is characterised by easy access to all kinds of services. Internet voting could therefore capture people's interest as being a modern and contemporary alternative for traditional elections: the vote can easily be cast, through the very same medium which today guides a rising number of people in their everyday life.

Internet voting could also be an appealing alternative for people residing abroad and ill or disabled people. In some countries these people can cast an absentee ballot through ordinary mail. In other countries, like Belgium, they have to vote by proxy.

Considered like this, the idea of voting through the Internet is not as revolutionary as it might have seemed at first sight. It is most of all a logical consequence of the evolutions in today's society, with an everyday growing role for new communication techniques. Instead of questioning if people are ready today for Internet voting, we should ask if today's traditional elections aren't outdated in today's information society.

Internet voting has other benefits: tabulation time and accuracy, cost-effective equipment and reduced cost of elections.

Because tabulation is done electronically, not manually, it is significantly faster than hand-counting or machine-counting paper ballots. In addition, it is not subject to human error, whether that occurs by miscounting or by mishandling of paper ballots or punch card ballots.

The cost of purchasing, maintaining, and storing conventional poll site equipment is expensive. With Internet voting, computer equipment that already exists today can be used.

When implementing a system that allows Internet voting from a PC at home or work, there will be no additional capital equipment cost.

Finally, the costs of conducting and participating in elections will decrease. Savings can be found everywhere from the cost of creating, printing, and distributing paper ballots to the stamps for mail-in absentee ballots, and even the cost of gas for getting to the designated poll-site.

1.4 Feasibility of Internet Voting

Although Internet voting can be a valuable alternative for today's elections, one has to consider the legal constraints for organising public elections.

As it often is the case, law lags behind rapid changes in society and not always turns out to be well adapted to the Internet reality.

But we can observe that the (European) legislator is ready today to recognise the Internet to be a valuable medium and to adapt present legal framework to this new reality. The E-commerce Directive⁴ and the Electronic Signature Directive⁵ are just a few examples.
The continuation of the present report investigates if an Internet voting system could comply with present legislation and, if so, under which conditions. If not, the question will be raised whether it is desirable to review legal constraints and how this is to be done. This may in part be a policy decision.
Mainly the following legal constraints for public elections will have to be overcome by an Internet voting system: public elections in a democratic Europe require the elections to be equal accessible (non-discrimination), democratic (one man, one vote), free (fair, secret and without influence or coercion) and reliable (secure and verifiable).
These legal constraints will have to be addressed on a technical level. The Internet voting system will have to enable a correct authentication of the voter but will also have to safeguard his/her anonymity, the voting and counting process will have to be made transparent and verifiable, and the system has to be secure and reliable.

1.5 Method of Analysis

An important part of this report considers if Internet voting could conform to the basic election rights laid down in international and regional convents and in the national constitutions. These rights are that important and that basic in a democratic society, that they cannot easily be altered. Therefore, violation thereof cannot be tolerated. These rights should however be interpreted on the background of social changes.

Also other legal requirements for democratic public elections are analysed. Therefore, the current voting procedures in a few Member States is described. We shall have to conclude that these rules are created in the light of traditional voting systems, in which people physically move to a polling station and cast their vote in secret in a voting booth, filling out an anonymous paper ballot. In such systems, the paper ballots are put in a ballot box and all ballots are physically collected and counted.

In some countries (for instance Belgium), it is actually possible to vote using an electronic voting machine. A new law has been drafted for this new voting system. The vote is cast with an optical pen on a computer screen; the vote then is registered on a magnetic card, which the elector puts in an electronic ballot box. The votes are registered and recorded on a floppy disc and then counted electronically. Although innovative, the voting-process doesn't basically change: the voter still casts his vote at a polling station in the secrecy of the voting booth, the vote is registered on a (anonymous) physical (magnetic) card and the vote is not transferred online.

Nevertheless, the system shows similarities with Internet voting, being the legal recognition of an electronic or digital voting system. The experience, both on legal as on practical level is useful for assessing the legal requirements for an Internet voting system.

In the end of course we shall assess the legal possibilities and/or restrictions for an Internet voting scheme. This report therefore assesses the legal requirements CyberVote will have to comply with and/or how the legal system could be adapted, so that CyberVote could be implemented.

1 Internet Voting Report, California Internet Voting Task Force, January 2000 (USA) - http://www.ss.ca.gov/executive/ivote/

2 For a description of the digital signature technique, see D4 Volume 1: "Review of electronic democracy projects".

3 This study intends to report on the requirements for organising legally binding democratic public elections. It does not intend to offer technical solutions to these problems. These are provided in the technical part of the CyberVote project, by offering a state of the art voting system, which is technically able to address these firm requirements.

4 Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market;

5 Directive 1999/93/EC of the European Parliament and of the Council of 13 December 1999 on a Community framework for electronic signatures;

Back to Cybervote Reports List