Online Publications 

Towards Cybersociety and "Vireal" Social Relations

Geser Hans:
E-voting projects in Switzerland. In: Sociology in Switzerland: Towards Cybersociety and Vireal Social Relations. Online Publikationen. Zuerich, August 2002.


E-voting projects in Switzerland

Hans Geser

University of Zürich, Switzerland
August 2002



1. The ambivalent role of democratic traditions for e-politics

2. The postal vote as a major step toward "virtual voting"

3. The initiating and facilitating role of Federal government

4. The ambiguous implications of decentralized Federalism

5. The three pilot projects currently in the stage of experimentation

5.1 Neuchâtel: a comprehensive strategy encompassing all plebiscitarian rights

5.2 Zürich: Overcoming the obstacles of decentralized administrative structures

5.3 Geneva: the most comprehensive and pioneering endeavour

5.3.1 Introduction

5.3.2 Favourable preconditions

5.3.3 Goals and driving motivations

5.3.4 On the framework of planning and implementation

5.3.5 Voting procedure

5.3.6 Future perspectives of amelioration

6. Some longer-run concerns

6.1 The collisions with norms of transparency and lay control

6.2 The precarious role of government in the web of online information channels

6.3 Questionable potential impacts on political participation



1. The ambivalent role of democratic traditions for e-politics

In the governmental and administrative realm, Switzerland has developed a number of rather advanced Internet applications: e.g. by implementing electronic systems of transit traffic control or tariff payments, and by offering online options for filling out census questionnaires and tax declarations or for following parliamentary debates in real-time on the Net. [1]

In the sphere of political communication, however, the use of online media has remained outstandingly peripheral. This is astonishing in the light of the extensity and intensity of grassroots-level political discussion and deliberation going along with the frequent initiatives and referenda which lead to a multitude of plebiscitarian votings (on the Federal, cantonal and communal level) every few weeks. Thus, participation in online discussions and chat hours (organized by reputable politicians) is notoriously weak, and no cases are known where the Internet has been effectively used for petitions or other sorts of political campaigns.

Consistent with these trends, the “Selects”-survey conducted in 1999 has shown that only about 11% of all Swiss Net users are visiting any websites of political parties, and a tiny minority of them declares to be influenced by any kind of content on these pages. And the Vox-Surveys (regularly conducted after each federal voting) show hat only 3-4% of all voters use the Internet for acquiring knowledge about the issues in question. [2]

Thus, we face the remarkable paradox that the country most known for its traditions of direct democracy does make little use of many obvious democratic potentials opened up by online communication: its capacity for more intensive "bottom up politics" and for a more extensive plebiscitarian participation. Instead, the Internet has mainly given rise to rather centralized "top down initiatives" which tend to collide heavily with the decentralized, fragmented power structures as well as with a conservative minded population.

We may speculate that exactly these highly established democratic traditions are the reason why the need for online channels has remained low, because they have given rise to a dense network of offline communication channels: ranging from informal pub and tavern gatherings ("Stammtischrunden") to a wide range of assemblies within the formal framework of local and supralocal party sections, citizen groups, movement activist groupings and voluntary associations. Adding the tiny geographical areas and the fine-grained settlement patterns which make that anybody has easy access to some of these structures (without extensive travelling), it is evident that virtual communication channels will also in the future less be needed than in countries like Australia, Canada or Russia where million of miles distance (and several time zones) have to be bridged.

Of course, the same factors have been responsible for a fine-grained allocation of ballot places, so that most citizens find a voting opportunity within walking distance.

Who will find electronic voting attractive (or even indispensable) under such favourable conditions? And: how could significant numbers of citizens be brought to use their PC for executing the solemn act of voting, when they have never acquired the habit of using the Internet for any more trivial purposes of political information or communication?


2. The postal vote as a major step toward "virtual voting"

Nevertheless, as voting rate have become lower and lower since decades, it has become folk wisdom to believe that this trend could be weakened or even broken by making voting more easily accessible in a technical way. Since the late 70ies, the Swiss government thus undertook initiatives to use new ballot procedures for increasing participation levels, especially among the younger and more mobile segments of population.

Thus, as early as 1979, a working group instituted by the Federal Department of Justice had come to the conclusion "dass alles getan werden muss, um materielle oder psychologische Hemmnisse für die Ausübung des Stimm- und Wahlrechts abzubauen. Dazu gehört, dass der Gang zur Urne erleichtert oder durch die briefliche Stimmabgabe sogar unnötig gemacht wird." [3]

In the following, the postal vote procedure was introduced by the Federal Government and most cantons in the early nineties: a step which went along with many deep changes in voting behaviour, changes usually associated with the electronic vote.

The major modification consists certainly in the extended time span available for depositing the individual vote. In most of the cantons, citizens receive the voting materials about 25 to 28 days before the official polling date, and on the average, about 60% of the voters make use of this new procedure [4]. In urban settings like Basel, or Geneva, the rate of postal voters usually exceeds 80%, while in the suburban and rural towns and villages, the traditional walk to the ballot urn (open at specific hours from Friday evening to Sunday morning) is still preferred by the majority.

But even in some cities, postal rates have remained low when a dense network or polling stations has been maintained. [5]

At least in the case of Geneva, empirical findings allow the conclusion that the postal vote has been effective in raising the turnout rates. Since about 1997, Geneva now ranges above the Swiss average, while in the past, it was notoriously far below. [6]

A the moment, we see in many places the initial phases of a "vicious cycle“: insofar as many polling stations are closed because usage have fallen too low - causing again many voters to send votes per mail because walking distances have become intolerably long.

Two highly problematic disfunctions of postal vote procedures have to be noticed, which are both aggravated by peculiarities of the Swiss political system:

1) Almost half of all postal voters deposit their vote before the governmental agencies have sent out their informational material, and before the public campaigns and discussions focussing on the voting issue have gained momentum. This of course is highly inconsistent with a concept of deliberative democracy in which voting decisions should only be made after all organized groups have been able to propagate their views and information, and opinions have been shaped within widespread processes of discursive communication. In fact, such extensions of voting periods privilege larger and richer groupings and associations which have enough means for letting their campaigns start weeks or even months before polling dates, and it hurts weaker groupings which have to concentrate their activities on the final days.

2) A second disfunction concerns the exercise of two other political rights provided by the Federal constitution: the signing of initiatives and referenda.

In the past, groups organizing such campaigns had an easy and cheap option of collecting signatures: by just catching citizens at the exit gates of the polling places. Today where less and less citizens still make use of physical ballots, other, more costly procedures have to be used. Again, this change favours the more potent groupings and organizations, which can pay solicitors who walk the streets or engage in door-to door canvassing. [7]


3. The initiating and facilitating role of Federal government

In February 1998, the Swiss government ("Bundesrat") has committed itself to a long-term e-government strategy aiming "to make Switzerland one of the leading countries in this realm." [8] Since then, this initiative has given rise to a multitude of rather loosely connected specific projects: carried on by specialized agencies and offices on the one hand and by lower-level governmental units (cantons and communities) on the other.

They can easily be subsumed under two headings:

  1. "Guichet virtuel" projects: promoting the use of online channels for administrative procedures (e.g. tax procedures, military service, renewal of passports etc.).

  2. "E-voting" projects: aiming at the development of secure online methods for depositing yes-no votes in issue-related pools as well as lists of candidates in various elections. In addition, future citizens should also be able to exercise all other political rights by online channels: e.g. by signing petitions, initiatives and referenda in a legally binding way.

These aims have been clearly stated by the Swiss "Bundesrat" in responding to a parliamentary interpellation:

„Die Unterschrift zu eidgenössischen Volksinitiativen und Referenden oder zu Nationalratswahlvorschlägen sowie die Stimmabgabe bei eidgenössischen Volksabstimmungen und bei Nationalratswahlen werden elektronisch erteilt, übermittelt, geprüft, gezählt und weitergeleitet.“ [9]

As a major consideration, the executive hopes that electronic voting will increase voting turnout rates, particularly among the younger cohorts who are known to be regular Internet users, but to have below-average participation rates at conventional polls. In addition, it is envisaged that population segments could be activated which were hitherto not able to vote for various physical reasons: e. g. Tourists or business-men absent on polling days, physically handicapped individuals [10], and particularly the more then 500 000 emigrants ("Auslandschweizer") on all continents who have retained their citizenship as well as their voting rights (even in the second and third generation). [11]

More than that, it is envisaged that the government should use the new online channels for making their citizens express more differentiated opinions than just “yes or no”. By inviting them to comment their votings by remarks and arguments, it is hoped that government will be better able to interpret the results and to guess what kind of options citizens “really prefer”.

„In elektronischen Abstimmungssystemen kann die Möglichkeit vorgesehen werden, dass Stimmende gleichzeitig mit der Stimmabgabe zur Vorlage fakultativ Fragen beantworten und Meinungen äussern. Dies schafft die Möglichkeit, dass Abstimmungsresultate elektronisch interpretiert werden können. Damit könnte den Haltungen der Stimmberechtigten vermehrt Rechnung getragen werden. Unter Vorbehalt einer strikten Anonymisierung der Stimmabgabe wären Auswertungen von Abstimmungen möglich, die sehr genaue Daten liefern können. Daraus könnte eine Zunahme des Einflusses der Bürgerschaft auf die Entscheidungen in der Politik entstehen.“ [12]

It is evident that the Swiss political system is especially prone for e-voting for the following reasons:

  1. An extremely large number polling procedures (elections as well as issue votings) on the federal, cantonal and communal level take place every year, so that considerable economic and organizational advantages can be gained by making use of the new technological means. In addition, the pressure for change is aggravated by the fact that it becomes increasingly difficult to recruit enough volunteers for organizing and supervising the polls and counting the ballot papers.

  2. Voter registration procedures are on a very high level, because registration is compulsory and registries are permanently updated (e.g. when people move from one community to another).

  3. By adopting postal vote laws in the 90ies, Switzerland has gone already a long way toward "distant-polling" procedures no longer based on personal appearance at the ballot stations, so that e-voting will not appear as a major revolution. Such far-reaching voting facilitations are themselves an indicator that polling in Switzerland is more "desacralized" than in many younger democracies where it is a ceremonial act not easily amenable to any innovative change. [13]

  4. Finally, the introduction of e-voting is facilitated because norms concerning the secrecy of votes are less pronounced than in many other Western countries. This is due to a long tradition of public votings (in open assemblies like "Landgemeinden" or town hall meetings) where this secrecy can evidently not be secured at all. [14]

Within Swiss society, these e-voting endeavours can currently rely on considerable support:

  1. A representative survey conducted in Nov. 2000 has shown that about 66% of the Swiss citizens are much in favour of electronic voting. Thus, it is certainly one of the most preferred Internet project undertaken by Government: ranging closely behind online communication with administrative offices (74%) or electronic passport prolongation (72%).

  2. On the level of organized groupings, the political parties (especially the Radical Democrats) have expressed much support for e-voting procedures, as well as the association of cantonal administrators (Staatsschreiberkonferenz), which will have a major role in the process of implementation (see below).

Finally, in a survey conducted in Spring 2001, all cantonal governments have expressed their willingness to engage in e-voting projects - at least under the condition that the Federal government provides expertise and pays the major costs.

Despite this broad support, the latest documents indicate that the Federal government has recently become more cautious (if not pessimistic) about the whole endeavour. Thus, official agencies are eager to underline that 2010 is the earliest possible date where an official start of e-voting (on the federal level) could be envisaged, and that costs between 400 and 600 Mio Swiss Franks have to be expected, 50% of which would have to be carried by the communities. [15]

In addition, some statements indicate that the government has lost enthusiasm by taking into account several critical arguments brought forward by experts and organized groupings. In particular, the Bundesrat seems to share the concern that

  • the territorial basis of the Swiss voting system (organized into cantons, districts etc.) could become eroded when votes could be delivered from any location in space,

  • the use of online technologies could result in an acceleration of voting processes, which would have a negative impact on the time-consuming processes of political campaigning and deliberation. [16]


4. The ambiguous implications of decentralized Federalism

Due to its historic formation (as a secondary confederation of pre-existing political units) and to its multitude of locally based languages, traditions, and political cultures, Switzerland has conserved a rather decentralized system of political power.

First of all, the 26 cantons have maintained considerable legal, financial and organizational autonomy in many policy fields, particularly in the realm of political and administrative organization. Secondly, most of them provide rather extended autonomy rights to their communities, especially in the German speaking regions.

From the perspective of political innovations and reform, this decentralized structure is certainly propitious insofar as it provides a rich “variety pool” of smaller units trying out their own ways and disposed for trial-and error experiments which involve little costs and risks because they remain bound to smaller parts of the territory and population.

In the realm of election and voting procedures, this functionality is explicitly taken into account by the Federal government. Thus, while the Federal Law of political rights (BPR) obliges all cantons to observe traditional voting procedures (e. g. implying that votes have to hand-written and voting sheets have to be personally signed), Par. 84 provides space for exemptions by authorizing the Federal governments to grant specific exceptions:

„Der Bundesrat kann die Kantonsregierungen ermächtigen, für die Ermittlung der Wahl- und Abstimmungsergebnisse mit technischen Mitteln von diesem Gesetz abweichende Bestimmungen zu erlassen. Wahl- und Abstimmungsverfahren mit technischen Mitteln bedürfen der Genehmigung des Bundesrates.“ [17]

This Paragraph provides a sufficient legal basis for testing out electronic voting procedures (without even specifying to what extent traditional rules may be violated), so that no additional legal (or even constitutional) rules had to be created. [18]

On the other hand, the same Federalist structures hamper the implementation of political reforms insofar as

  • various cantons and communities have established their own traditional procedures and institutions since decades (or even centuries)

  • Federal agencies always remain dependent on the voluntary cooperation of subnational units, and nationwide reforms are highly dependent on reaching a consensus, which can be promoted, but not authoritatively enforced.

As a result of this basic need for voluntary multi-level accordance, all Federal commissions dedicated to e-governments have to be recruited according to highly complex criteria: so that widespread representation of different cantons and language regions is secured. [19]

In summer 2001, the Federal working group for the promotion of E-voting has conducted a survey in order to assess the current state and the future plans of e-voting projects in all 26 cantons.

The results indicate that while everybody is generally interested in implementing e-voting procedures (by collaborating with the Federal government), very few cantons have started concrete projects, and even less have already created the necessary legal provisions.

Most of the Cantons are very far away from fulfilling one important precondition: the establishment of a central electronic voting registry. Even worse: they have no means for realizing this aim, insofar as constitutional norms don’t allow them to enforce homogeneous rules and procedures on all their communities. Given the tiny population and administrative apparatus of most municipalities, they often lack the financial means and expertise to make the transition from physical to electronic registries on their own. And in several cases, the cantonal administrations themselves lack the means to provide professional help. [20]

State of e-voting endeavors in 26 Swiss Cantons (survey from July 2001) [21]




Interested to undertake an e-voting project in collaboration with the Federal government



E-voting project currently underway



Legal provisions for e-voting have already been enacted



Endeavours to create such provisions are underway



Voting registries are administrated by the Canton (not by the Communities)



A standardized electronic voting registry has been established



The communities have changed to electronic voting registries

all=7 (some=12)


Thus, we may conclude that the implementation of national e-voting will critically depend on major reforms

  1. on the political level: by creating the legal basis for homogeneous electronic registries

  2. on the organizational level: by providing the personnel, expertise and technology necessary to implement such homogeneous procedures on the level of all (about 2990) communes.

It is to be expected that centralized top down endeavours by the Federal governments to implement homogenous procedures on all subunits will never be accepted. Instead, only highly time-consuming bottom-up solutions will be viable: e.g. in the sense that intercommunal and intercantonal networking of voter registries occur, so that the Federal government will only possess a “virtual registry” by having access to intercantonal servers. [22]

Given these contingencies, the question arises whether the interest in e-voting will sustain on sufficiently high level to motivate such far-reaching changes.

As to be expected, the four cantons actively engaged in projects have not coordinated their endeavours (despite the fact that they have invited each other to participate in the specialized commissions), and the goals they envisaged are rather different.

The most potent Canton (Zürich) has currently the most modest aim of just coordinating and homogenizing the communal registries, so that the transition to electronic formats will be possible in a later step. Neuchâtel is currently also still occupied by tasks of administrative standardization and centralization, but with the more ambitious aim of developing a format of electronic signatures which can be applied to e-voting as well as to initiatives and referenda.

At the moment, only Geneva is seriously committed to the concrete establishment of a fully functional official e-voting procedure: confronting itself deeply with all the legal, technical, administrative, psychological and cultural implications associated with such an unprecedented step. The project is strictly focussing on yes-no issue votings, so that the complications associated with electoral procedures (e.g. name cancelling or cumulations on lists) have not to be dealt with.

Given the need for centralized and standardized registries and administrative procedures, it is highly significant that the more ambitious projects are exclusively found in the French-speaking region of Switzerland where traditions of direct democracy and communal autonomy are rather weak. This situation favours e-voting for two very different reasons:

1) Implementation is easier because standardization has already been made, and because Cantonal agencies have more authority to enforce common rules and practices

2) E-voting is more likely to be accepted by the citizens because for many of them, the community is not a relevant level of political participation, so that they have never become familiar with traditional ballot locations.

Geneva is particularly standing out as a canton where the political relevance of communities is minimal, and where rather anonymous relationships between citizens (and between citizens and governmental agencies) prevail. This is at least partly due to the high demographic fluctuation, which implies that many inhabitants have just recently immigrated.



5. The three pilot projects in the stage of experimentation

5.1 Neuchâtel: a comprehensive strategy encompassing all plebiscitarian rights

Based on an unusually tight collaboration between communities and cantonal agencies, the executive of Neuchâtel has decided in favor of a comprehensive one-step concept, which should almost simultaneously establis e-voting at all plebiscitarian issue votings (on the communal, cantonal and federal level)

  1. e-voting at all elections for the legislature and the executive (on the same three levels)

  2. online subscriptions of petitions, initiatives and referenda.

All three aims should be realized at the end of 2002.

Contrasting with Geneva where the authorization code has only be submitted after the vote has been sent, the Neuchâtel strategy relies on an anticipatory authorization as it is used by most telebanking procedures After personal authentication, the system checks whether the individual has not yet voted before. When this check has been successfully passed, the voter receives an access code as well as a password, which are necessary in order to get access to the official voting site. When delivering the vote, an additional secret number has to be submitted which citizens have received by postal mail. Like in Geneva, Online votes are stored in an encrypted “electronic urn”. When counting begins, electronic votes are added to those delivered by postal letter or at the ballot stations, and a check with the central voting registry ensures that no citizen has submitted more than one vote.

Contrary to the Geneva system, which treats individuals exclusively under the aspect whether, they are (or are not) entitled to vote, the Neuchâtel solution has the advantage of embedding e-voting processes in a larger universe of citizen-government relationships. This e-guichet concept implies that whenever an individual has completed successfully its authentification, he or she gets access to a multidimensional universe of official relationships based on a manifold of salient relationships (e.g. tax payer, enrolment in education, welfare receiver, public employee or member of the catholic church).

Another characteristic is that in contrast to Geneva, the Cantonal administration plays a lesser role because many operative tasks (distribution of voting material, administration of voting registries, vote counting etc) are taken care of by the communities.

Instead of constituting a centralized physical registry of its own, the Canton is satisfied to establish a “virtual central registry” ad hoc at each polling date: by simply creating access to the 62 communal registries (which are mirroring current conditions because they regularly updated each day or week).

Thus, the Neuchâtel solution relies much more than the Geneva concept on the practices of “cooperative Federalism” between communal and cantonal agencies as autonomous actors of equal status) which are characteristic for most Swiss Cantons.


5.2 Zürich: Overcoming the obstacles of decentralized administrative structures

In contrast to the two aforementioned French cantons which can both rely on highly centralized cantonal administrations, the canton of Zürich displays many the features which are typical for German-speaking regions: a highly heterogeneous set of rather autonomous communities (ranging from about 200 to 350 000 inhabitants) with highly divergent Computer and Software systems as well as rather heterogeneous voter registration systems.

Given these inhospitable premises, the establishment of a unified population and voter registry would already be a laudable pioneer task.

The cantonal executive has chosen the only viable compromise: to establish a secondary (“virtual”) voting registry based on communal registries, which have to be maintained and updated by communal administrations.

For the purpose of every polling procedure, a “virtual voter registry” is created by aggregating the current communal registries. Each citizen receives an automatically generated numeric code, which can be used as identification key for getting access to the electronic polling system. On the polling paper, the same code is printed as a bar code. Whenever a citizen votes at the ballot station or by mail, this barcode is checked in order to verify whether the same individual has not yet submitted his vote online (in which case the vote is invalid). Given that the counting is administered by the communities, each commune has to have access to the central “virtual registry” and to be equipped with the machinery for deciphering these codes.

While Geneva and Neuchâtel tailor their systems exclusively to the Personal Computer, only Zürich takes into account that future citizens may more rely on mobile phones, PDA’s or other lean and easily mobile devices. As a consequence, a polling system is envisaged where all problems related to reliability, security, encryption, privacy etc. are solved for a very wide range of different hardware configurations, operating platforms, software applications and transmission protocols.


5.3 Geneva: the most comprehensive and pioneering endeavour

5.3.1 Introduction

In a comparative international perspective, the current e-voting project in the Canton of Geneva stands out as one of the very few really serious attempts to implement formally binding governmental voting procedures on the WWW. This implies that all the major factors on the cultural, psychological, political, administrative and legal level have come into play, and water-tight solutions for all preconditions necessary for legitimate democratic voting procedures (security, anonymity, reliability, controllability etc.) have to be addressed and solved on an operational level. As a consequence, the Geneva pilot project is a breeding ground for fruitful discussions, innovations and experiences, which may be relevant for similar future projects within communities, governments, voluntary associations or any organized democratic systems.

5.3.2 Favourable preconditions

For five different reasons, Geneva is better predisposed for implementing E-voting than most other Swiss cantons:

  1. A centralized electronic voting registry exists, so that no obstacles related to divergences among different communities have to be overcome.

  2. Contrary to all other Cantons, the voting law (instituted in 1982) authorizes the Cantonal executive "to collaborate with the communities in trying out new voting methods at variance with the present law, in order to bring voting procedures in line with new technological conditions." [23]

  3. As the rate of votes delivered by postal correspondence has recently risen to about 90%, the large majority of citizens has already turned away from traditional ballot urns to more anonymous voting procedures. As a consequence, the additional change to electronic voting is not likely to be felt as a dramatic change.

  4. An unusually high percentage of all citizens (about 5.9%) are currently living abroad. Thus, there are high incentives to enlarge the effective electorate by providing a low-threshold access channel for every entitled citizen all over the globe.

  5. Geneva possesses a rich of supply of highly competent experts and institutions (e.g. the CERN for expertise related to Net technologies, and the Cantonal University for legal questions and issues related to political science). [24]

5.3.3 Goals and driving motivations

Apart from more specific goals (e. g. to increase turnout rates by mobilizing additional younger voters or emigrated citizens), the Geneva authorities seem to be driven by more general ideological motivations and by their ambitious endeavour to make Geneva the "E-capital" of Europe. These broader goals are also manifested in several other electronic innovations implemented within the last years: e g. the rapid transmission of voting results on WAP-enabled mobile phone. Various statements in official documents also corroborate the impression that the e-voting project is primarily nourished by a high-flying technological enthusiasm, not by concretely defined expectations:

"Pourquoi introduire le vote par Internet ? Les avancées scientifiques contemporaines vont au-delà de simples progrès technologiques. De l’avis de nombreux observateurs, nous vivons une période charnière, un basculement d’époque, comme a pu l’être il y a cinq siècles la période séparant l’invention de l’imprimerie (1457) de la Renaissance.
Plutôt que de subir ces bouleversements, l’Etat doit les anticiper et les accompagner, notamment en veillant à ce qu’à terme personne n’en soit exclu."

5.3.4 On the framework of planning and implementation

In March 2000, the Cantonal executive (Conseil d'Etat) has agreed on a proposal to implement e-voting within the subsequent three subsequent three years. In the first stage, it should be completely restricted to issues votings (where only In the first stage, it should be completely restricted to issues votings (where only yes-no statements have to be processes), while applications for the election process should be delayed to a later period.

In the course of a public subscription in fall 2000, two private enterprises (Hewlett-Packard and Wisekey SA) were selected for working out the design of the system and the adequate software requirements. At the same time, various experts from reputable local institutions (CERN, University, central hospital etc.) were chosen for studying various critical issues of e-voting procedures (e.g. problems of voter identification, security and confidentiality of votings etc.).

In the following, several smaller test runs were made in connections with Federal votings in June, September and December 2001 as well as in March 2002.

The total costs running as high as SFR 1.25 Mio (of which 1Mio was carried by the Federal government) were distributed as follows: [26]

Hardware equipment

180 000

Upgrading of telephone systems

288 000

System Software

40 000

Development of procedural applications

210 000

Security testing (incl. mandated hacking)

80 000

Legal and political science study

30 000

Public Communication

315 000

Public Hotline

50 000

Diverse costs

50 000

In the longer run, it is estimated that e-voting will cause a cost increase of about 8% for each individual voting procedure in the Canton (from 500 000 to 540 000).

5.3.5 Voting procedure

All citizens receive the official documents for voting per mail three weeks in advance. The enclosed personal voting card entitles everybody to deliver the vote online or by mail at any moment until Saturday at noon before the official election day, or by traditional ballot voting at this official day.

The access to Internet voting is made very simple by using methods similar to most e-commerce applications: Without having to register in any ways, Individuals can point their browser to a portal site which leads them to all relevant information and form pages. They even can transmit their vote freely and experiment themselves "whether it really works". Formal registration is only necessary in the last phases in order to validate the submitted vote. This is also the only stage where communication with the voting server is encrypted (on the basis of SSL 3.0).

This is achieved by filling out the form below: inserting

a) an individual identification number (PIN code)

b) a personal password (which can both be made visible on the voting card by scratching it at specific locations),

c) the date of birth.

By scratching the paper for reading the password, the voting card is visibly invalidated, so that it can no longer be used for mailed or ballot procedures.

After personal authentication, voters get an immediate confirmation that their vote has been accepted, and they can optionally receive an additional confirmation by email (without the content of the vote) to an indicated address.

The servers dedicated to identification, registration and voting shall have the capacity of processing up to 2000 voting requests per hour, while the servers used for presenting informational material shall be able to absorb much large numbers of visitors.

This procedure (used similarly in the realm of lottery tokens) has the advantage of being functional without a digital signature, but on the other hand, it cannot guarantee that the vote really originates from the entitled person. [27]

During the polling periods, the encrypted votes are accumulated in an "electronic urn" to which nobody (especially no citizens who are entitled to vote) has access.

The encryption procedure is still under scrutiny. It is envisaged to use two keys each of which is known to only one representative of the voting committee (which is again steered by representative of the political parties). The two individuals are requested to define the two codes at the beginning of the polling period, so that it can be used as a basis for encryption. During this period, the keys are conserved in a closed enveloped guarded by the president of the cantonal "chamber of solicitors". When the counting of votes begins, the chamber president opens the envelope and communicates the keys, so that they can be used for decryption.

5.3.6 Future perspectives of amelioration

On Jan 28th 2002, the committee dedicated to the study of security problems has published an extensive report in which many still unsolved problems plaguing the current e-voting project are addressed. [28]

The main argument is that within the sectors controllable by government and administration (e. g. receiving servers, counting procedures etc.), security deficiencies are rather easy to identify and can be readily solved on a reliable basis. On the other hand, this is not true for the user-controlled sectors which are characterized by a gamut of complex, unpredictable and volatile problems related to bugs in various operation systems, to viruses and trojans as well as to intentional attempts of hackers to observe or even modify delivered votes on their transit, or to redirect them to other servers.

In order to reduce such risks and dangers, the committee proposes a multitude of possible measures: many of them apt to generate considerable additional costs and organizational overhead on the one hand and incalculable unintentional disfunctions on the other.

For instance, the establishment of a "test urn" is suggested which would enable controllers to verify at least indirectly whether the electronic voting procedure is operational without and distortions. This could be realized with the following procedures: certain segments of the electorate are selected for delivering two parallel votes (one online and the other in a conventional ballot urn), so that it could be checked whether identical results are achieved on both channels.

Secondly, it is suggested that the official voting software should be made available to all citizens on a CD-ROM, in order to create identical security standards on all peripheral computers. Evidently, this measure would engender many additional difficulties resulting from deficient hardware equipment, driving software, damaged diskettes and other problem sources: so that equal chances for online voting participation could no longer be secured.

Finally, the commission suggests to open additional gateways for e-voting (e. g. vie mobile phone) in order to make voting procedures more independent from the PC and the WWW. Again, such measures would currently imply that additional sources of insecurity would be built into the systems: problems even less thoroughly known than the rather well studied deficiencies of the PC operating systems and the World Wide Web.

The authors at least acknowledge the basic dilemma associated with the fact that e-voting procedures have to fulfil two equally important, but highly contradictory goals: (1) providing a low-threshold access to everybody - even users with rather simple and antiquated computer equipment - and (2) following the highest technological standards in shielding the voting process from any security loopholes and any possible frauds.

In fact, the first of these considerations has been predominant during these beginning phases of implementation. [29]


6. Some longer-run concerns

6.1 The collisions with norms of transparency and lay control

Swiss political culture is highly democratic not only in the sense that most important issues are submitted to plebiscitarian voting procedures, but also in it insistence that all important governmental and even administrative processes (votings included) should be subject to lay control, which could potentially be exercised by any citizens, without specialized technical, administrative or legal expertise.

In the realm of votings and elections, this implies that all ongoing procedures should be transparent in a way hat anybody can observe them and give a judgment, whether they accord with the required rules.

Evidently, the exercise of such controls is highly facilitated when low-tech offline procedures are in use, which can be understood by everybody without acquiring particular knowledge or investing costly time and labour. In such traditional systems - which certainly include votings by mail - all the documents and devices, which could potentially be subject to manipulations (voter registries, voting papers, ballot urns handwritten signatures etc.) exist in physical form, which makes them amenable to objective visibility and unimpeded examination. Such controls can easily be applied during the poll procedures as well as anytime after it has finished (e. g. when results are so close that a careful second or even third counting is necessary in order to eliminate all possible doubts).

As major disfunction of e-voting is that such physical media are eliminated, so that all processes of data generation, transformation and storage occur in "black boxes" often not fully transparent even to the technical experts (e. g. to the programmers who discover weekly new surprising bugs in rather well-tested software applications).

And to the degree that sophisticated and effective measures against frauds and errors are implemented, only tiny groups of experts know how they really work": groups which exclude most citizens entitled for voting on the one hand and include untitled specialists (e. g. from foreign countries) on the other.

As the „Groupe des Utilisateurs Linux du Léman" (Gull) [30] has articulated in an open letter directed at the Cantonal executive, such problems are aggravated in the case of Geneva because of the highly decisive role given to private corporations (HP and Wisekey). This cooperation implies that the software developed is of a proprietary nature, so that in contrast to open source applications, only the specialists involved in creating the software (and employed by the firm owning the property rights) are able to make checks because all others have no access to the code.

"Under such conditions, the transparency essentially needed for democratic procedures can never be secured. To make an analogy: this resembles a society where only layers and judges have access to the text of legal statutes." [31]

The major consequence of all this, is that the legitimacy and acceptance of democratic voting procedures becomes critically dependent on the degree of confidence accredited to the relevant experts by the general populations. Given the daily news about unexpected flaws in even widely established programs developed by the most sophisticated firms (like Microsoft), this seems a rather shaky foundation for future e-democracy indeed.


6.2 The precarious role of government in the web of online information channels

A major consequence of e-voting is that acts of political decision making are transported into the same media setting where almost all preceding political activities (campaigning, discussion and deliberation, information getting etc.) can take place.

It is highly reasonable to assume that e-voters will have the tendency to use online sources for informing themselves about candidates or about the pro’s and con’s of current voting issues, and that many of them will cast their vote immediately after gathering the necessary information.

Thus, there seem to be many sound reasons for governmental agencies to enhance their e-voting sites into prolific “one-stop pages” which provide a manifold of links not only to official statements and recommendations, but also to various media, discussion fora, information sources and to the propaganda sites of candidates, committees, citizen groupings, associations and political parties. [32]

By reducing information costs, such sites may well be functional

  1. to increase the quality of votes by inducing more active citizens to gather more diverse and detailed opinions and information:

  2. to increase the quantity of votes by mobilize hitherto passive voters.

On the other hand, such expansions of governmental media activities would be highly problematic for the following reasons:

  1. The “digital divide” between e-voters and traditional offline voters would be amplified, because the latter would still need to consult newspapers and direct mail material, to watch radio and TV broadcasts (or even attend public assemblies) in order to get the same quantum of relevant information.

  2. By following hyperlinks, citizens would be less able to differentiate clearly between “official” documents and documents stemming from nongovernmental sources.

  3. Governmental agencies would have to be selective in their external links: e.g. explicitly neglecting sites considered to contain false information, extremist positions or even slandering and openly racist statements. Of course, such governmental filtering would be considered as “censorship” highly incompatible with freedom of speech, or even worse: as an attempt of government to manipulate public opinion according to its own preferences. [33]

  4. By inducing citizens to cast “just-in-time votes” immediately following the consultation of the e-voting sites (and some of its neighbouring first-order links), voting decisions may easily become influenced by spontaneous impressions and emotional reactions. In other words: the “cooling-out” period between information intake and voting output is eliminated, so that less opportunity for intellectual operations (taking critical distance, reflections, synthesizing operations etc.) is left. [34]


6.3 Questionable potential impacts on political participation

The major driving force for spending so much money, brains and administrative efforts for electronic voting procedures stems from the hope that it will increase political participation by offering a low-threshold access to vote deliverance which should particularly appeal to rather inactive citizens which don’t find it worthwhile to walk to a polling station at a prescribed stretch of time.

In March 2001, the IPSO research institute has conducted a survey encompassing 1000 citizens from Geneva, in order to gain insight into the prospective impact of e-voting on voting participation. In this survey, 35% of all informants expressed the opinion that e-voting would motivate them to participate “more regularly” at the polls. Nevertheless, this effect is much more pronounced for males (43%) than for women (27%). This would imply that an increase in voting turnout up to about 9% could realistically be expected. [35] Such conclusions, however, are heavily contested in an expert study conducted by Prof. Wolf Linder (University of Berne) who concludes that this mobilization effect will remain below 2% because most of the current Internet users are already active voters. [36]

The most significant quantitative effect of e-voting may well be a rising political participation rate among the 580 000 Swiss citizens living abroad. While this rate has risen fourfold since 1992 (from 17000 to 70 000), it is still low because of false addresses, inefficient international mail services and other handicaps associated with postal voting procedures. [37]

On the other hand, including more emigrants may well doing harm to the rationality of democratic decisions, because voting outcomes become more influenced by individuals who are rather incompletely informed and who have no stake in the domestic political system.

For similar reasons, it might be questioned whether its a good thing to use e-voting for increasing the participation rates among rather passive citizens who have not voted so far: because this would lead to a increased weight of low-rationality voters who are rather undecided and poorly informed.



[1] Pressemitteilung des Bundesrates vom 5. 4. 2001.

[2] vgl. Auer, Andreas / Trechsel, Alexander H: Voter par Internet? Helbing & Lichtenhahn, Genève, Bâle, Münich 2001: 17ff.

[3] Amstutz, Peter: Der Souverän entscheidet aus Distanz. Basler Zeitung vom 5. August 1999.

[4] "Zwischen 97 und 2 Prozent" Neue Zürcher Zeitung, 15, 5. 2002: 17.

[5] Interestingly enough, the costs of postal fees seem to be irrelevant, because rates of postal votes are similar in settings where community pays these fees (e.g. Zürich) as in the settings where they are paid by the voters. (See. Baumann, L.: "Gang zur Urne immer seltener /Briefliche Stimmabgabe erreicht annähernd 60 Prozent" (Neue Zürcher Zeitung, 16. 03. 1999: 106.))

[6] Amstutz, Peter: Der Souverän entscheidet aus Distanz. Basler Zeitung vom 5. August 1999.

[7] Amstutz Peter: Der Souverän entscheidet aus Distanz. Basler Zeitung vom 5. August 1999.

[8] Antwort des Bundesrates auf eine Interpellation des Ständerats Peter Briner (Freisinig, Schaffhausen) vom 6. 9.2000.

[9] Antwort des Bundesrates auf eine Interpellation des Ständerats Peter Briner (Freisinig, Schaffhausen) vom 6. 9.2000.

[10] Bericht über den Vote électronique Chancen, Risiken und Machbarkeit elektronischer Ausübung politischer Rechte vom 9. Januar 2002: 9.

[11] Présentation de M. Robert Hensler au 2e Forum Mondial de la Démocratie Electronique (17 Mai 2001) 
Möglicherweise versprechen sich die bürgerlichen Parteien eine verstärkte Partizipation relativ wohlhabender Wählersegmente, weil marginale Bevölkerungskreise oft (noch) nicht über den erforderlichen Netzzugang verfügen. De facto ist damit zu rechnen, dass E-voting – insofern es überhaupt die Stimmbeteiligung erhöht – das Gewicht jüngerer, männlicher und gebildeter Wählerschichten verstärkt.

[12] Bericht über den Vote électronique Chancen, Risiken und Machbarkeit elektronischer Ausübung politischer Rechte vom 9. Januar 2002: 10.

[13] Thus far, only half of the EU member states have introduced postal voting procedures (Denmark, Germany, UK, Netherlands, Portugal and Sweden), and all of them practice it more restrictively than Switzerland.

[14] Stimmgeheimnis, Stimmzwang und Volksrechte in den souveränen Staaten der Welt, Bundeskanzlei. Bern, 2001.

[15] Bericht über den Vote électronique Chancen, Risiken und Machbarkeit elektronischer Ausübung politischer Rechte vom 9. Januar 2002: 10.

[16] Antwort des Bundesrates auf eine Interpellation des Ständerats Peter Briner (Freisinnig, Schaffhausen) vom 6. 9.2000.

[17] Par 84 BRP

[18] Of course, this no not mean that extralegal procedures would be allowed, because it is the responsibility of the Cantons to secure the legitimacy of such innovations on the basis of democratic laws. 

[19] Antwort des Bundesrates auf eine Interpellation des Ständerats Peter Briner (Freisinnig, Schaffhausen) vom 6. 9.2000.

[20] This is particularly true for a tiny Canton like Appenzell-Innerrhoden with about 15 000 inh.

[21] Arbeitsgruppe "Vorprojekt E-voting“: "Umfrage bei den Kantonen", Bern 2001.

[22] Bericht über den Vote électronique Chancen, Risiken und Machbarkeit elektronischer Ausübung politischer Rechte vom 9. Januar 2002: 29.

[23] Art. 188: Loi sur l'exercice des droits politiques du 15. octobre 1982. Nevertheless, this competence covers only non-binding trial procedures (on the communal and cantonal level) practiced during limited periods of time.

[24] Particularly relevant is the "Centre d'études et de documentation sur la démocratie directe" (C2D) which is officially engaged to follow and evaluate the whole process of testing and implementation. (See:

[25] Vote par Internet. Frequently asked questions.

[26] Bericht über den Vote électronique Chancen, Risiken und Machbarkeit elektronischer Ausübung politischer Rechte vom 9. Januar 2002: 29Beilage 10: Pilotprojekte, Vote éléctronique. Elektronische Ausübung politischer Rechte. Chancen, Risiken Machbarkeit, Bundeskanzlei Bern, 9. 1. 2002.

[27] This deficiency also hold in the case of traditional voting procedures which also cannot exclude that voting sheets are filled out by any non-entitled individuals.

[28] Rapport du Comité Sécurité sur l’application de vote par Internet Etat de Genève, CERN. Hôpital Cantonal Universitaire de Genève, 28 janvier 2002

[29] Hensler, Robert : Chances et défis du vote par Internet.


[31] Gilbert Robert / José Manuel Nunes: Swiss Open Systems User Group.

[32] Auer, Andreas / Trechsel, Alexander H. op. cit 2001: 58ff.

[33] Niedermann, Dieter Arbeitspapier E-voting. Bericht des Bundesrates an die eidgenössischen Räte. St. Gallen, 27. Juni 2001.

[34] Linder, Wolf: Gutachten zum E-voting. Bern, 2001

[35] IPSO 2001, zitiert nach: Auer, Andreas / Trechsel, Alexander H.: Voter par Internet? Helbing & Lichtenhahn, Genève, Bâle, Munich, 2001: 52ff.)

[36] Linder, Wolf: op.cit. Bern, 2001.

[37] Auslandschweizerorganisation

Last update: 01. Feb 15

Home: Cybersociety and Vireal Social Relations Home: Sociology in Switzerland


  Prof. Hans Geser
Soziologisches Institut

der Universität Zürich