19/10/2007: BELGIUM - FEDERAL ELECTIONS - 10 June 2007 : OSCE/ODIHR Election Assessment Mission Report

I. EXECUTIVE SUMMARY

In response to an invitation from the Delegation of Belgium to the Organization for Security and Co-operation in Europe, the OSCE Office for Democratic Institutions and Human Rights (ODIHR) deployed an Election Assessment Mission (EAM) for the 10 June 2007 parliamentary elections in Belgium.

Belgium has a long tradition of holding competitive elections, as attested to by these elections, and there is a high level of public confidence in the electoral system. While voting is obligatory by law, this obligation does not seem to be systematically enforced. Turn-out was approximately 91 per cent.

Belgium’s political life is distinguished by the respective Dutch-speaking and the French-speaking electorate and regions. There are no political parties with significant national support represented throughout Belgium.

The election campaign was visible, although without dominating public life. The Belgian legislation on campaigning and party financing is relatively restrictive. As a rule, political parties largely rely on distributing flyers, the presence of candidates in public places, rallies with party supporters, billboards and coverage by the media. The media are dynamic and pluralistic, and provided broad coverage on election issues.

The legal framework provides a sound basis for democratic elections. It is in some aspects advantageous to established parties, but this has not hindered new parties from emerging in the last decades, contributing to an already heterogeneous political landscape. There is no specific law on political parties, but restrictions are placed on parties which convey ideologies and platforms deemed xenophobic, racist or discriminatory. These restrictions have led to court appeals in a number of cases.

The elections were administered by the election administration in an efficient and professional manner. Most poll workers are first-time voters, positively engaging young people in the election process, but more training may be desirable.

Electronic voting has become the method of balloting for 44 per cent of the electorate. While the overall functioning of these technologies does not appear to be fundamentally questioned on a large scale, adherence to the procedures and transparency of the system could be enhanced. In this regard, some members of Parliament, and one civic group which considers the e-voting system unsafe and not transparent, have given visibility to some of these questions in Belgium.

The OSCE/ODIHR EAM also noted that the e-voting system relies heavily on the vendors, both for its instalment and maintenance, with a very limited presence and supervision of the public authorities in charge of the conduct of the electoral process. It could therefore be necessary to underscore the clear division of responsibilities between election administrators and vendors, in order to guarantee appropriate accountability. As the current e-voting system is due to become obsolete in 2008, an overall study on electronic voting is being undertaken, involving the main universities of the country.

II. INTRODUCTION

In response to an invitation from the Delegation of the Kingdom of Belgium, the Organization for Security and Co-operation in Europe’s (OSCE) Office for Democratic Institutions and Human Rights (ODIHR) deployed an Election Assessment Mission (EAM) for the 10 June 2007 elections.

The OSCE/ODIHR EAM was deployed from 3 to 13 June 2007, and was led by Ambassador Lubomir Kopaj. It consisted of ten election experts from ten different OSCE participating States. Based in Brussels, the OSCE/ODIHR EAM also deployed teams to Antwerp and Liège. The OSCE/ODIHR EAM had meetings with government representatives, election officials, political parties, and representatives of the media and civil society. In line with the standard OSCE/ODIHR approach, an EAM does not encompass systematic or comprehensive observation of election-day procedures.

Following a previous invitation from the Belgian Government to observe the October 2006 local elections, the OSCE/ODIHR had sent two experts in electronic voting, accompanied by an ODIHR election adviser for a study visit to Belgium in October 2006 to increase ODIHR’s comparative knowledge of e-voting systems. A report on this visit is available on the OSCE/ODIHR website. [1]

The OSCE/ODIHR wishes to express its gratitude to the Federal Public Service for Foreign Affairs, Foreign Trade and Development Co-operation of Belgium for the support and cooperation offered to the EAM, as well as to other state authorities and interlocutors.

III. LEGAL FRAMEWORK

A. INSTITUTIONS

Belgium is a federal State with a parliamentary system of governance under a constitutional monarch. The institutional framework has undergone significant changes in five reforms since 1970, establishing a federal system in 1993.

There are three levels of government. At the highest level are the Federal State, the three Communities (Dutch-speaking community, French-speaking community, and German-speaking community) and three Regions (Flemish region, Walloon region, and Brussels-Capital region). The boundaries of Communities and Regions do not
correspond. There are four linguistic areas : the Dutch-speaking area (the Flemish Region), the French-speaking area (most of the Walloon Region), the German-speaking area (a small part of the Walloon Region) and the Dutch/French bi-lingual area (the Brussels-Capital Region).

The Federal State, the Regions and the Communities are equal under law. They each fulfil their responsibilities in their field of competence. The Federal State is in charge of foreign relations, defence, justice, and internal affairs. The Communities are in charge of language and cultural issues and they implement social policies. The Regions have
gained increasing competence in relation to the economy, regional development, the labour market, agriculture, housing, and the environment. Each Region and Community has its own Parliament.

At the middle level there are ten provinces, [2] and only the region of Brussels-Capital is not sub-divided into provinces. At the lowest level there are 589 municipalities, which play a role in the organisation of elections. There are also 208 cantons, which are involved in the election administration.

B. OVERVIEW OF THE LEGISLATION

The main legislation pertaining to elections includes the Constitution of 1970, the Electoral Code, which dates back to 1894, the law on electronic voting of 1994, and the law related to party and campaign financing of 1989.

A significant electoral reform in 2002 led to amendments to the Electoral Code, which : reduced the number of constituencies, which now largely correspond to the provinces ; established a 5 per cent threshold at constituency level for parliamentary representation ; and introduced an obligation to gender parity on candidates’ lists.

C. STATUS OF POLITICAL PARTIES

There is no procedure for registering political parties, which do not enjoy legal personality. For financial and other activities they operate through related non-profit associations (Association sans but lucratif - Vereniging zonder winstoogmerk), nevertheless the law on party financing of 1989 contains a definition of a political party.

Several pieces of legislation include provisions aimed at curtailing the operation of persons or organisations considered to be xenophobic or denying the holocaust. As a result, in 2002 associations related to the Vlaams Blok party faced a financial penalty on these grounds. The party has since re-established itself under the name Vlaams Belang.

D. ELECTORAL SYSTEM

The legislative branch of power at Federal level is composed of the House of Representatives with 150 members elected by popular vote for four year terms and the Senate with 71 members, whose term is also four years.

The House of Representatives (Kamer van Volksvertegenwoordigers / Chambre des Représentants)

Members of the House of Representatives are elected by regional proportional representation, on the basis of constituency party lists. There are five constituencies in the Walloon region (Walloon Brabant, Hainaut, Liège, Luxemburg, Namur), five in the Flemish region (Antwerp, Flemish Brabant, West Flanders, East Flanders, Limburg) and the constituency of Brussels which includes parts of the Province of Flemish Brabant. Constituencies comprise from four (Luxembourg) to 24 seats (Antwerp) with a five per cent threshold for lists’ eligibility to seat allocations applying at constituency level. [3] According to art.63 of the Constitution, each seat should represent a similar population number. The size of the population is determined every ten years by census.

Voters can either vote for a list in its entirety by ticking the top box of the list on the ballot paper (list votes), or indicate preferences for candidates on one list by ticking boxes next to the candidates’ name (preference votes). Preference votes are not given in any order and voters are free to give preference votes to as many candidates on a list as they want. Seats are allocated to party lists according to the d’Hondt method of highest average.

Once the number of seats per party list is known, the seats are distributed to candidates whose number of preference votes reached a quotient representing the number of votes won by their list in the constituency, divided by the number of seats allocated to the list +1. If not all the seats won by the list can be filled with candidates whose number of preference votes has reached the quotient, the remaining seats are awarded as follows : List votes - which do not express preference for any of the candidates on the list - are used to ‘top-up’ the number of preference votes cast for candidates who did not reach the quotient. The allocation of these list votes is done by following the order of appearance of candidates on the list until the total number of list votes is exhausted. In order to increase the effect of preference votes, the 2002 electoral reform established that only half of the list votes could be used to top up candidates’ preference votes.

While in Brussels candidate lists are presented by Dutch-speaking and French-speaking political formations, in the other constituencies parties usually present candidate lists in their respective linguistic areas only. This is up to the choice of the respective party ; the election system allows any party to present lists anywhere in the country regardless of language.

One contentious aspect of the electoral system appears to be the delimitation and existence of the constituency of Brussels, which includes two adjacent areas, Halle/Hal and Vilvoorde/Vilvorde. It is often referred to as the “BHV constituency”. The BHV constituency existed before the 2002 electoral reform. When the legislator increased the size of electoral constituencies in 2002, which now largely follow provincial borders, it retained the constituency of BHV. Although situated in the Flemish Region, the areas of Halle/Hal and Vilvoorde/Vilvorde are populated with a significant number of French-speaking citizens, many of whom moved there from Brussels city over the last decades in a process of sub-urbanisation.

The Court of Arbitration - now Constitutional Court - decided in May 2003 that the BHV constituency raised questions of constitutionality and tasked the legislator to find a solution before the end of Parliaments’ term in 2007. However, political parties were not able to agree during the term on a solution to this issue. The Parliament was dissolved in May 2007 and elections called for 10 June, slightly before the end of the term, and thus ahead of the deadline imposed by the Court to solve the BHV issue.

Parties can declare before the elections that they ‘group’ their lists for BHV with their lists in neighbouring Flemish Brabant or Walloon Brabant respectively. Such declarations are taken into account when seats are allocated after the elections. [4]

The Senate (Sénat / Senaat)

Of the 71 Senators, 40 are directly elected by popular vote together with the elections for the House of Representatives, 21 Senators are designated by the Parliaments of the three Communities [5] and ten Senators are co-opted by the other Senators. In addition to the 71 members, the King’s children are also Senators by law, but they do not play a political role in the Senate.

From the 40 directly elected Senators, 25 are elected from Dutch-speaking lists and 15 from French-speaking lists, following a system of proportional representation. For the purpose of the election of the Senate, the Belgian electorate is composed of a ‘Dutchspeaking Electoral College’ and a ‘French-speaking Electoral College’. Voters in the Walloon Region belong to the French-speaking Electoral College and can only vote for the French-speaking lists ; voters in the BHV constituency have the choice to vote either for the Flemish-speaking or for the French-speaking lists ; voters in the Flemish Region (with the exception of Halle/Hal and Volvorde/Vilvoorde areas) belong to the Dutchspeaking Electoral College and can only vote for the Flemish-speaking lists.

Seats are allocated using the D’Hondt method to the lists obtaining at least five per cent of the votes cast within their Electoral College. Seats won by lists are then allocated to candidates according to the number of preference votes they received, in a manner similar to one used for the House of Representatives seats.

E. PARTY AND CAMPAIGN FINANCING

Following a string of party financing scandals in the 1980s, a relatively restrictive law on party campaign financing was adopted in 1989. [6] Political parties are not allowed to receive corporate donations, and private donations are limited to €500 per year, and per person, for the same party. Other sources for party funding include state funding, membership fees, income from real estate, etc. with state funding being the most significant for most parties.

Parties which have at least one representative in one of the houses of Parliament are eligible for state funding, which includes an annual support of €125,000 and a supplementary support of € 1.25 per vote received in the last general elections.

The law also provides that parties which do not respect the rights and liberties included in the European Convention on Human Rights may be deprived of state funding for a specific period. On this basis, a number of parties requested the State Council in May 2006 to withhold state funding for the Vlaams Belang. The case is still pending.

The campaign period for federal elections is three months during which parties’ expenditure for campaigning is limited to €1 Million. [7] There are specific ceilings for campaigns by individual candidates. Candidates, who are placed higher on a candidate list, often incumbent MPs, enjoy higher ceilings. While this does not affect the equality of parties competing with each other, it has an impact on the ability of candidates on one list to compete on conditions of equality. Such a provision raises questions of conformity with the OSCE commitments that the rights of citizens to seek political office shall be respected without discrimination, and that the law and public policy work to permit political campaigning to be conducted in a fair and free atmosphere. [8]

Consideration should be given to review the law so that all candidates running in the same constituency have the same ceilings for campaign expenses.

F. ELECTION OBSERVATION

The election code allows for the presence of political party ‘witnesses’ during polling, counting and tabulation of results. It appears that the larger parties in particular make use of this right. Domestic non-partisan election observation is not foreseen in the law.

The OSCE/ODIHR EAM was accredited by the Ministry of Interior and all levels of the election administration were cooperative and provided any information that was requested.

While the accreditation and co-operation received by the OSCE/ODIHR EAM was most welcome, in order to fully comply with paragraph 8 of the 1990 OSCE Copenhagen Document, the electoral code should be amended to specifically allow for international and domestic non-partisan observers.

IV. ELECTION ADMINISTRATION

A. OVERVIEW

The elections are administered by various public bodies, including the Ministry of the Interior (MoI), the judiciary and the municipal administrations. As far as electronic voting is concerned, which covers 44 per cent of the electorate, the MoI is more involved than in traditional voting, notably by issuing specific guidelines. There is no permanent election administration body, and all officials involved in the conduct of elections perform non-election duties on a day-to-day basis.

In each of the 11 constituencies there is a Constituency Main Office. They tally the constituency results for the election of the members of the House of Representatives and for the Senate. The results for the Senate are then sent to one of the two College Main Offices, one for the French-speaking electorate and one for the Dutch-speaking electorate. The 11 Constituency Main Offices and the two College Main Offices are headed by the Presidents of the respective Courts of First Instance or another magistrate.

At the next lower level, the material organization and supervision of the electoral process is in the hands of judges (Justices of Peace), who head the 208 Canton Main Offices. They nominate the presidents and members of polling stations and counting offices, have substantial responsibilities in the running of the electronic voting system, and they tally at canton level the results coming from counting offices. They are assisted in their duties by local civil servants.

At the lowest level, polling station boards administer the voting process. They are chaired by judges, attorneys or qualified civil servants selected by the Head of the respective Canton Main Office under a formula proscribed by law [9]. The chairs are complemented by four to six assessors and a secretary. For the 2007 election, the assessors were selected by the cantons from the voter registry.

Voting takes place in some 10,500 polling stations countrywide, including around 4,000 using electronic voting systems. Polling stations can have from 150 to 800 assigned voters, or in the case of electronic voting, up to 2,000 voters.

Counting of traditional paper ballots is performed in counting offices, grouping together at least 2,400 voters and three polling stations. In cantons where electronic voting is used, there are no additional counting offices. The tabulation is directly done at Canton level. [10]

While most political parties expressed general satisfaction and trust in the current system of election administration, some indicated a need for more interaction by the MoI and local election officials with the political parties and candidates regarding procedures and transparency issues, especially with regard to the electronic voting process.

B. VOTER REGISTRATION

Belgium has a passive voter registration system whereby lists are compiled by municipalities based on population registers some two months before an election. All Belgian citizens who are 18 years old and above are automatically registered in one of the 589 municipalities, as long as they are not deprived of their voting rights [11]. Voters abroad can register in a diplomatic office. Voters can complain to the authorities and appeal to courts on issues related to the voter list, e.g. if they are not included on a list (art.18-34 electoral code). Copies of voter lists are available for political parties. Some 7,720,796 citizens were registered as voters. A week before the election, canton officials, whose local database is coordinated with the MoI, sent an invitation with a reference number on it to each voter. On election day, this number is matched with the number in the voter list, facilitating quick identification of voters in the polling stations. Voters who did not receive an invitation can obtain a duplicate in municipal offices, which are open on election day.

Voting is compulsory in Belgium since 1893. Penalties are foreseen in case of breach of this obligation [12], as election administration officials will send information on voters who fail to cast their ballot to the public prosecutor for possible action. However, the OSCE/ODIHR EAM was informed that the enforcement of this provision was not systematic, and varies from constituency to constituency.

C. CANDIDATE REGISTRATION

For the election to the House of Representatives, candidates’ lists are submitted to the Main Constituency Office. Lists need to be supported by 200-500 signatures, depending on the constituency population. For the Senate election, lists must present 5,000 signatures to the relevant College Main Office. As an alternative to obtaining voter signatures, candidates’ lists may obtain the support of three MPs for election to the House, or two Senators for election to the Senate.

It appears that there was no specific procedure that established rules for checking supporting signatures and, in some cantons, the signature verification process was minimal, with just a cursory check of the signatory against the civil registry and not necessarily against the signature itself. In case of complaint from interested parties, supporting signatures were checked more thoroughly, including police verification. Some officials suggested that since the candidacy forms used by the candidates can be different, there should be standard forms/guidelines provided for consistency and ease of verification.

Consideration should be given to reviewing the legislation to provide uniform procedures for the verification of support signatures of voters for the registration of candidates, and for consistency in the forms that are used to obtain such signatures.

Candidates’ lists were required by law to have equal numbers of men and women. The 10 June election resulted in a high rate of the parliamentary seats filled by women, 36.7 per cent. [13]

Most candidates on lists for incumbent parliamentary parties obtained ballot access through MP support. The number and order of lists on the ballot was determined by lot by the MoI, with the position of those political formations providing a full list determined by lot first, and the smaller lists determined by lot separately. Paper ballots contained the logos supplied by the political formations. While the e-voting machines did not contain the logo, they did contain the party abbreviation provided by the party.

D. OUT-OF-COUNTRY VOTING

Belgian citizens residing abroad have had a possibility to vote for federal elections since 1998. Changes were made to the system in 2002 in order to provide greater access. In order to participate, citizens abroad must fulfil the ordinary eligibility rules and must have registered in the relevant consulate.

The law offers five options for casting a ballot including : (1) voting in person in Belgium, (2) voting via proxy in Belgium, (3) voting in person in a diplomatic representation abroad, (4) voting via proxy in a diplomatic representation abroad, or (5) voting by postal ballot.

With regard to the proxy voting option, it would be useful to consider a review of the regulation and practice of proxy voting, in order to further enhance consistency with the principles of the equality and secrecy of the ballot, in line with paragraph 7.4 of the 1990 OSCE Copenhagen Document.

Voters must complete a form, several months before the scheduled elections, in which they express their choice for one of the above-mentioned voting options. Voters also indicate the communes in which they will appear on voters’ lists, which is where their ballot will be added. Votes in embassies are cast one or two days before election day, and counted under the auspices of the Ministry of Foreign Affairs on election day. In practice, about 65 per cent of out-of-country voters use postal ballots. The postal votes are sent to the designated constituency and mixed with votes cast locally. Postal ballots are accepted until 13:00 on election day, when the paper ballot polling closes. Some 120,000 citizens residing abroad were registered to vote for the 10 June election.

Some out-of-country voters did not receive their ballots or received them too late. The OSCE/ODIHR EAM was informed that some 69 per cent of those who subscribed for postal voting received their ballot on schedule. This could be an indication of problems in the delivery of postal ballots. According to the law, ballots should be sent to voters abroad at least 12 days before election day (Art.180 of the electoral code). This deadline reflects the provision that candidate lists should be determined the latest 12 days before the elections. [14] Evidence suggests that overseas ballots that were received by the authorised officials may not be kept in adequately secure locations, or procedures for handling them were not fully secure.

Consideration could be given to extend the deadlines for ballot paper delivery, in order to allow for more time to address any issue that may arise with out-of-country voters. This should, however, be considered in view of other election deadlines, including the one for candidate registration.

V. ELECTRONIC VOTING

A. OVERVIEW

Electronic voting was introduced in Belgium in the beginning of the 1990s. Since 1999 it covers some 44 per cent of the electorate and has not been expanded. The contract with the companies supplying the e-voting system will expire in 2008, providing an opportunity for an overall evaluation and decision if and how to continue with e-voting. The government has tasked a research consortium to provide a report on the global state of the art of e-voting to assist in arriving at a well-informed decision. Some advocate the view that either all voters should be subject to e-voting, or it should be abolished.

While the overall technical performance of the e-voting procedures would not appear to be fundamentally questioned, some political party officials, in particular on the Frenchspeaking side, and civic group activists, have expressed concerns about e-voting. The focus of their criticism largely stems from concern with regard to the lack of effective public oversight of e-voting. At least one civic group, PourEVA (Pour une Ethique du Vote Automatisé), has been particularly active in giving visibility to these issues, and has contributed to initiating a broader public discussion.

The framework for e-voting in 2007 was similar to that in 2006, which was described in an expert report by the OSCE ODIHR [15]. This report will cover some specific issues in more detail, in particular those related to public confidence in the two e-voting platforms currently used (Digivote and Jites).

Typically, the technicalities of the e-voting system are not understandable to the average citizen ; in this regard, clear communication to the public and political parties, as well as
transparency mechanisms, can significantly contribute to voters’ confidence in the system. The introduction of a paper back-up of each vote, that is, a voter verified paper audit trail (VVPAT) [16], was demanded by some political parties as a way to enhance transparency.

Legislation and regulations for electronic voting machines should include provisions requiring Voter Verified Paper Audit Trails (VVPAT) or an equivalent verification procedure. [17] The legislation should provide for the possibility of a meaningful postelection recounts and audits of results.

B. CERTIFICATION PROCESS

The Ministry of Interior (MoI) certifies the e-voting system prior to each election. Testing of the software is carried out by private audit companies, which are chosen and contracted by the respective vendor of the e-voting software from an MoI-approved shortlist. [18] The testing firms have to fit criteria stipulated by the MoI, for example, being independent from the vendors and having the required expertise.

The main purpose of the software audit is to determine whether the e-voting solution complies with several specific requirements set out in legislation [19], and guarantees the system’s security and performance, as well as secrecy of the vote [20]. The vendors give to the auditing company the comprehensive documentation about their e-voting system. The documentation is kept with the auditing company.

The final audit reports are given simultaneously to the MoI and to the vendor. Later, the College of Experts, a control body described below also receives a copy of the software used for each of the two e-voting systems. There are no clear rules regarding publication of these audit reports, and it appears that most of the auditing companies would be reluctant to agree to their publication. They argue that it could disclose their methodology.

Once the MoI receives the audit reports, it decides whether or not to certify the implementation of the given e-voting platform for the next election. The College of Experts considers that the control carried out by the MoI over the audit reports should be more rigorous.

The transparency of the certification process should be enhanced, including access by local authorities, political parties, observers, and relevant civil society groups to certification reports. The tests should cover all aspects of the system. A detailed comprehensive list of required criteria should be developed as the basis for testing.

The practice that the auditing company is contracted by the vendor raises questions and should be reassessed.

C. THE COLLEGE OF EXPERTS AND THE ROLE OF POLITICAL PARTIES

The College of Experts is an independent consultative public body whose members are appointed before the elections by the Chambers of Parliament. It has been created in 1999 in order to provide more control on the e-voting systems. Its members are entitled to request any information, and to visit polling stations in order to carry out a series of checks, including copying the floppy discs used in these polling stations for further checking. The College starts operating 40 days before the elections.

The Assemblies are free to choose the College members ; in practice, they tend to be civil servants of the assemblies. For these elections, the College also included an external adviser. The NGO PourEVA raised concerns over the system of nomination of the College of Experts, underlining that the Chambers of Parliament are also the ultimate judges of the validity of the election of their own members.

There are no written internal procedures, and until now the members of the College have adopted decisions by consensus. The experts are neither paid for this task, nor relieved from their day-to-day duties during this period. The College has to deliver a report to the chambers of Parliament and to the MoI within 15 days after the elections. There is no legal obligation to publish it, but up to now all reports on previous elections have been published.

Political formations which have at least two representatives in one or the other chamber of Parliament can designate an IT expert who can receive the source codes of the different e-voting and counting systems and any other information that may be necessary to carry out a control. The expert has to treat these data confidentially. [21]

In order to maintain public confidence in the e-voting system, more efforts should be made to clearly explain its technical and operational elements, and ensure appropriate safeguards for transparency and accountability. The Ministry of the Interior should play a crucial role in communicating this information to the public, political parties and NGOs.

The role of the College of Experts should be enhanced. Their mandate should start earlier than 40 days before election day and their members should be available fulltime at least during the election period. In addition, they may need to employ more stafffor checking system components and controlling that operational procedures are followed. The inclusion of more external advisers and legal experts in the College should be considered.

When controlling the electronic tally, the re-calculation of the tabulation by the College of Experts should be extended to include a recount of magnetic cards at randomly selected polling stations. Finally, the electoral code should foresee that the reports of the College are published.

D. SOFTWARE / HARDWARE SECURITY AND MAINTENANCE

Measures for ensuring the security of the software seem to be adequate. However, the OSCE/ODIHR EAM saw some degree of flexibility in their implementation, at times bordering on violation of the procedures.

After certification, a master copy of the voting, counting and tabulation software is stored in a safe bank box under the responsibility of the MoI ; a second copy is given to the College of Experts, and a third one is used by the MoI to generate floppy discs to be used in polling stations and tabulation centres. During the 2007 elections, for the first time each floppy had a password under a scratch-banner [22]. Floppies and passwords were kept in envelopes stamped on the back by the MoI and secured by an encoded plastic seal. These were delivered to Heads of Canton Main offices ahead of election day.

Even though the electoral code provides that the Head of Canton Main Office should give two envelopes, one with the floppies and another one with the password, to the Chairperson of the polling station (CPS) the day before the election, [23] the OSCE/ODIHR EAM saw instances where municipal employees assumed that role on election day. This resulted in the Head of Canton Main Office having no direct and face-to-face relationship with the CPSs in this specific step. It should also be mentioned that the Heads of Canton Main Offices receive a supplementary list of the passwords required in each polling station.

Between elections, the hardware (voting and counting machines, validation devices and electronic ballot boxes) is stored by the municipalities. Ahead of these elections, the material for polling stations was delivered to the Head of the Canton Main Office in each canton. A test is done some months before the election with a non-certified software delivered by the e-voting supplier and a similar test is carried out after the deployment and setting up of e-voting equipment in the polling stations.

Finally, magnetic cards are stored in municipal warehouses with the rest of the hardware. They are transferred to the city hall some weeks before the elections in order to be included into kits prepared for each polling station. They remain in city halls until they are delivered to the relevant polling station by a city agent designated for that particular polling station.

Help-desks and emergency systems are set up by the vendors in order to address possible problems during election day. The firm Steria decided to hand over this responsibility to a subcontractor, while the firm Stesud had its own dispatching centres. In Liège and Antwerp, for instance, the vendors set up these centres with a small warehouse where the company stores spare devices and coordinates the emergency responses. The MoI also has several decentralized offices where, if necessary, a copy of the software can be generated. While help desks for both Jites (Stesud) and Digivote (Steria) were set up within the framework of terms of reference formulated by the public authorities, it must be noted that they operated without any presence or supervision of public authorities in charge of running the electoral process.

The OSCE/ODIHR EAM recommends a comprehensive vulnerability study of the entire system be undertaken by IT security experts.

The OSCE/ODIHR EAM also recommends :

• Ensure that public authorities in charge of the electoral process have adequate supervision and command of the e-voting system, and avoid excessive reliance on the vendors for running the system.
• Increase the number of persons entitled to observe the process of making floppies from the master, including, for instance, representatives of political parties.
• Reduce the number of people assuming responsibility for handling the floppy discs, ideally, only the Heads of Canton Main Offices and CPSs should handle floppy discs.
• Enhance security and integrity of the hardware while stored between elections.

VI. CAMPAIGN

The campaign was visible but remained overall low key. Most of the campaign was accomplished through distributing flyers, candidates’ presence in public places, rallies of party supporters, door-to-door canvassing and billboards. Most of the parties opened their campaign with rallies.

Most political parties expressed satisfaction with the framework for campaigning, although some asserted that campaigning could be more vivid and effective if the ceilings of campaign expenses were higher. However, smaller parties were satisfied with the limitations.

There were two main ways for the candidates and parties to convey their respective messages to voters in electronic media. TV and radio channels organized debates between the candidates during prime time and also there was free air time reserved for the political formations to campaign. The legislation did not allow paid advertisement in electronic media. In print media, however, the parties and candidates were able to place paid advertisements. Most of the political parties met by the OSCE/ODIHR EAM expressed satisfaction with the existing rules for campaigning in the media.

The OSCE/ODIHR EAM received a number of complaints from Vlaams Belang and Front National concerning their access to and coverage by the media. The Vlaams Belang complained that they were refused the possibility to place advertisements in several newspapers. Some newspapers informed the OSCE/ODIHR EAM that it was against their statutes, and two newspapers refused because the advertisements were not delivered on time. Vlaams Belang challenged these refusals before the Brussels Court of First Instance, which dismissed the claim. The two parties also complained that they had not been invited to TV debates (see Media Section below).

VII. MEDIA

A. MEDIA LANDSCAPE

The plurality of media outlets operating in the country reflects a long-standing tradition of free media, freedom of expression and an emphasis on cultural diversity. Citizens are served by 30 newspapers, more than 15 national and regional TV and radio stations, a widespread network of cable operators, and by a number of local TV and radio stations. The Belgian media are generally divided along linguistic lines and regulation of media falls largely into the competence of the three Communities.

All three Communities have a public broadcasting service financed and supervised by the Community’s institutions. The composition of the Boards of Directors is designed with a view to reflect the weight of political parties in the respective Assemblies.

The French Community is served by the public broadcasting company RTBF (Radio-Television Belge de la Communauté Française), with two television channels and two main radio stations, as well as by private media outlets. Some interlocutors complained that the representation of the Socialist Party on the board of directors of RTBF exceeds its representation in the Community Parliament and should be reviewed, as it might affect viewers’ and electoral contestants’ confidence in the impartiality of the public broadcaster.

The most popular French-speaking broadcaster is the private channel RTL – TV1 followed by the public channel La 1. TV channels from France are also popular, particularly TF1 and France 2, which enjoy a 27 per cent market share. As far as French-language newspapers are concerned Le Soir, La Libre Belgique, L’Echo and La Dernière Heure are the dailies with the widest circulation.

In the Flemish audiovisual market the public broadcasting company VRT (Vlaamse Radio- en Televisieomroep) is the most important with its two channels Eén and Canvas reaching an audience of around 35 per cent. The private VTM is the second most popular channel. The audience for TV channels broadcast from the Netherlands is limited. There are a variety of dailies and periodicals, including Het Laatste Nieuws, Het Nieuwsblad, Het Volk, De Morgen, De Tijd and De Standaard.

The German-speaking community has one public radio station BRF - Belgischer Rundfunk der Deutschsprachigen Gemeinschaft. Some ten other private radio stations operate at community level. The German newspaper Grenz-Echo is the main daily newspaper of this Community.

B. REGULATORY FRAMEWORK

1. Federal level

Article 19 of the Constitution protects freedom of expression and article 25, more specifically, freedom of the press. Limitations to freedom of expression are included in a law on racism and xenophobia [24] which sanctions any publicity given to the intention of resorting to discrimination, hate or violence against groups and individuals on the basis of their race, colour, national or ethnical origin. Similar provisions are included in the law against the negation of the genocide committed by the German national-socialist regime [25]. Many broadcasters have directly incorporated these provisions in their statutes and editorial codes.

The law on the limitation of campaign expenses also impacts on media obligations in the three Communities [26]. During a three-month period before polling day, any public information campaign by authorities must receive, prior to their publication, an advice by the ‘Council of control’ created within each Communities’ Parliament. This procedure aims at verifying whether public information campaigns appear to promote, fully or in part, parties or candidates. In addition, paid advertisement for candidates and political parties is forbidden on radio, on television and in the cinemas [27].

2. The Communities

There is no further national legislation on media conduct during election campaigns and each linguistic Community has its own regulatory framework. Nevertheless, regulation in all three Communities shares the same features, in particular a high level of selfregulation for election coverage ; allocation of free airtime for contestants in the public channels (the so-called “Tribunes”) ; a tight control to exclude or reduce the possible advantages of incumbents in terms of media access, and a complaint mechanism for candidates.

The French-speaking Community’s regulator CSA (Conseil supérieur de l’audivisuel) issued recommendations to all audiovisual media for the three months campaign period, including a number of general principles and journalistic best practices. The CSA recommended refraining from giving access to political groups or parties whose manifestos and ideologies are in breach of the federal laws against racism and xenophobia [28]. In line with CSA recommendations the majority of broadcasters published plans for election coverage [29].

It could be useful if all broadcasters would publish their internal rules for coverage of the elections.

The CSA also recommended that balanced and pluralistic debates between candidates of the leading parties be broadcast and that information be provided on parties, which for legal reasons were excluded from debates [30]. Beyond making recommendations, the CSA can sanction violations of laws.

The public service broadcaster RTBF adopted a detailed election plan which foresaw the allocation of free airtime to political parties on the basis of their parliamentary representation [31] ; parties with no parliamentary representation were given some limited access on the basis of the number of candidacies filed. For the coverage of candidates in news and debates, the RTBF decided that it should be proportional to the representation of parties in the Parliament of the French Community.

The framework for election coverage in the Dutch-speaking Community combines legal rules and self-regulation under the aegis of the regulator VRM (Vlaamse Regulator voor de Media). A decree on radio and television broadcasting [32] obliges broadcasters to adopt a number of principles in their general editorial coverage, namely an obligation of non-discrimination and impartiality in programmes for all broadcasters. The public VRT has the obligation to allocate free television broadcasting time to the political parties which are represented in the Flemish Parliament according to a proportionality formula [33]. Article 111bis §1 of the decree imposes an obligation of non-discrimination and impartiality on all broadcasters. VRT’s election coverage was further guided by their ethical code [34] and by their internal directives for federal elections [35]. The directives set a number of good practices for election coverage, including the obligation to impartiality, pluralism and objectivity.

The regulator VRM has effective law enforcement powers. When a violation of the provisions of the media legal framework is established, it can decide among a range of sanctions, which include : a warning with an order to cease the violation ; a compulsory communication, relevant to the case, at the expense of the party which committed the violation ; an administrative fine up to 125,000 euros ; the suspension or withdrawal of the broadcasting license ; the suspension or withdrawal of the broadcasting company as a legal entity.

During the campaign, the leader of Vlaams Belang filed a complaint with the Flemish media regulator VRM against the VRT, alleging that the decision of the public television not to invite him to the two debates "Het Groot Debat" (“The Great Debate”) with three candidates for prime minister (Guy Verhofstadt, Yves Leterme et Johan Van de Lanotte) violated the obligation of impartiality and non-discrimination as set out in the decrees. The complaint is still pending.

Complaints on media conduct during election campaigns should be addressed swiftly by media bodies, in order to allow for the remedy to be effective.

The media authority of the German-speaking Community, the Medienrat, has less authority than its counterparts in the other Communities. It has no right to sanction media for violations of laws, and while it is in charge of receiving complaints and giving advice, it has no power to impose obligations related to media coverage. The public broadcasting service, the BRF, adopted an internal self-regulatory code for election coverage. During the period of the campaign BRF broadcast debates with candidates and allotted free airtime to political parties.

The main national newspapers covered the election campaign and candidates extensively. Contestants are allowed to purchase advertising space in the print media, but some newspapers refused to publish advertisements put forward by the Vlaams Belang. The Commercial Section of the Brussels Court of First Instance ruled before the elections that newspapers had the right to refuse advertisements by political parties, whose aims run counter to the editorial principals and statutes of a newspaper, but that no practical or business reason could justify such rejections. Consequently only two papers (Gazet van Antwerpen and Metro) were obliged to run advertisements by the party.

Newspapers should make a special effort to publicise their statutes at the outset of election campaigns, in order to increase transparency on how they will deal with party advertising.

VIII. VOTING AND COUNTING

A. POLLING

In line with standard OSCE/ODIHR practice, the EAM did not observe election-day proceedings in a systematic or comprehensive manner. Nonetheless, the OSCE/ODIHR EAM visited polling stations in Brussels, Antwerp and Liège, on a limited basis.

Out of some 7.7 million voters registered in Belgium, some 7 million cast their ballot in some 10,500 polling stations across the country [36], a turn-out of approximately 91 per cent. Some 44 per cent of the electorate used one of the two types of e-voting systems available in Belgium, while the remaining ones used paper ballots.

Two ballot boxes, one for the House and one for the Senate, were provided in polling stations using traditional paper ballot. Six polling booths were usually set up in each station. Posters providing information on parties and their lists of candidates, as well as other information on voting procedures were displayed inside polling stations.

The voting process started at 8:00 and finished at 13:00 in polling stations using paper ballot and at 15:00 in polling stations using the electronic voting. The early closing time is a feature of the Belgian election system, and takes into account the duration of counting preference votes. Short delays were observed in the opening of some of the polling stations visited, mainly in those using electronic voting. Delays were also observed after polls were concluded, in the phase of preparation for transportation of ballots and other relevant election material to the respective counting centres. These delays appeared to be primarily due to insufficient knowledge of procedures and training of poll workers (assessors) of the respective polling stations.

Overall, assessors tend to be young voters. Once selected as assessors for a polling station, service is mandatory, unless proper justification is provided. In past elections, fines have been imposed on those not providing a justification for their absence.

While training is foreseen for presidents and secretaries of polling stations, it is not mandatory, and, according to interlocutors, a significant number do not attend the 20-30 minute training sessions, although all are required to pick up instructions and materials prior to election day.

The OSCE/ODIHR EAM was told that for any single polling station, up to 24 voters would receive notification to serve as an assessor. They are then required to pick up instructions prior to election day, and to report for duty at 7.15 on election morning. According to election officials, usually around ten assessors report for duty, some just before 8.00, when the polls should open.

In many cases, the presidents of polling stations are forced to make decisions at the last minute on who will serve, resulting in little time for training and observation of some of the crucial aspects of the opening and closing of the polls, as required by MoI regulations. This lack of training was evident in some polling stations visited by some OSCE/ODIHR EAM team members on election day. Several interlocutors suggested that pre-election training should be mandatory for all polling station officials, especially in e-voting stations.

The involvement of first-time voters in administering elections is commendable, but in order to ensure consistent and efficient implementation of rules regarding opening and closure of voting procedures, consideration should be given to make training mandatory for poll workers ahead of election day.

Apart from some delays in opening and closing in some polling stations, the voting process seemed to unfold smoothly with voters being familiar with voting procedures. However, a number of elderly voters and voters with an immigrant background requested assistance from poll workers or their relatives inside polling booths. This phenomenon was more evident in polling stations using electronic voting.

Voters showed to the polling station staff the invitation letter and their ID, and received it back after casting the magnetic ballot card or the paper ballot into the ballot box. The polling station staff checked consistently voters’ identity and stamped their invitation cards.

The same was done for proxy voting. Voters received one magnetic card in polling stations using electronic voting or two paper ballots, one for the House of Representative and one for the Senate elections. Voters voting with proxy received ballots also for the voters on behalf of whom they were voting. A voter could vote by proxy only for one other voter. [37]

Political parties are entitled to have witnesses in each polling station and counting centres. However, in the polling stations visited, only a few party witnesses were seen. Party witnesses were also present in the counting and tabulation centres at Canton level.

B. COUNTING AND TABULATION

In polling stations where ballot papers are used, the ballots are transferred to the counting centres. Counting centres for paper ballots count a minimum of three polling stations. Ballots for the House of Representative are counted separately from those of the Senate elections. In a first step, votes for party lists are counted, followed by a counting of preference votes for each list. The results are tabulated on paper protocols and publicly announced. They are then transferred to canton level where data are entered in computers and aggregated results sent electronically to Province level and the MoI.

Results for the House of Representatives are aggregated and declared at constituency level by the Constituency Main Office. For the Senate elections, results from electronic voting and from paper ballot counting stations are sent from the Canton Main Offices to the respective Electoral College Main Office and the MoI. Results are aggregated and declared by the main offices of the two electoral colleges.

Out-of-country paper ballots delivered by mail were counted in a specific canton of each constituency. Ballots were randomly distributed over a number of counting centres, in order to preserve the secrecy of out of country votes. The tabulation procedure was the same as for paper ballots cast in the country.

The election administration appears to enjoy the trust of the electorate and other stakeholders. Nevertheless, as a matter of best practice and transparency, it is recommended that detailed results of all polling stations (for e-voting) and counting centres (for paper ballots)are publicly disclosed when counting is completed.

C. PROCEDURES FOR ELECTRONIC VOTING

1. Opening of Polling Stations

E-voting polling stations must follow specific procedures concerning the use of passwords, handling of floppies and starting the computers. Polling station chairpersons are usually lawyers, while other Polling Station Board members or ‘assessors’ tend to be young first-time voters, without any previous training before the election day. As a result, assessors might find it difficult to follow the procedures performed by the Chair in a meaningful way. In polling stations observed by the OSCE/ODIHR EAM, it appeared on occasion that chairpersons carried out the first sensitive steps (i.e. sealing of the ballot box, opening the envelopes with the password and the floppy discs) before the formal constitution of the board, and without the presence of the assessors. In addition, the OSCE/ODIHR EAM also noticed that municipal staff was often involved in the technical preparations beyond the role foreseen in the law.

In order to ensure that procedures are followed and verified accurately, the time for training of poll workers should be extended.

2. Voting

After more than a decade of e-voting with the same devices, voters overall appeared to be familiar with them, and most voters seemed to make their choices in the booth in less than one minute. Nevertheless, the OSCE/ODIHR EAM also saw a significant number of cases where voters needed assistance.

After inserting their electronic ballot card into the voting computer, voters touch the screen with a pen to make their choices : they first choose the language to be used, then the party they want to vote for, and finally there is another screen where they can indicate preference votes. Voting is concluded by confirming the choice on the last screen. The system offers possibilities to cast a “blank” vote on the screen. In order to help the voter, the polling stations display paper ballots on the walls, as well as explanations on e-voting procedures.

The e-voting technology used in Belgium does not foresee that voting machines produce a paper back-up of each vote. This feature is being criticised for undermining citizen’s control of the process, and limiting transparency. The introduction of a voter verifiable paper audit trail (VVPAT) has been requested by some political party officials as a way to enhance transparency.

The system could be reviewed to :

• include a separate final confirmation screen where the voter could verify his/her choice of political party and selected candidates.
• introduce a voter verified paper trail, in order to increase transparency.

3. Closing and Counting

Closing in e-voting polling stations usually goes quickly. It can be carried out in 20 minutes. A copy of the ballot box floppy disc is generated and introduced, along with the original, in envelopes that the Polling Station Chairperson should bring to the counting centre. There is also a protocol to be signed by the polling board and the witnesses of the political parties. Since this document can be filled before making a copy of the ballot box floppy disc, in some cases the assessors left the polling station before this last step was carried out by the Polling Station Chairperson alone. Therefore, a number of envelopes with the floppies had no signatures in the back, in some cases even without that of the Polling Station Chairperson, [38] and the transfer to the counting centres also had no witnesses.

In Liège, there were instances where polling station chairpersons left the ballot boxes in the polling stations, while the city staff is supposed to bring them to a secure place. [39] There were no hand-over of documents to guarantee a proper custody chain of this equipment.

Consideration should be given to introducing the following procedures :

• Include a hand-over document for ballot boxes after the closing, in order to facilitate accountability of their transfer to the counting centres (Jites).
• Change closing procedures so that the members of the polling board cannot leave without attending the copying procedure of the ballot boxes’ floppy discs and signing the envelopes when they are enclosed.
• Guarantee the presence of at least one assessor during the transfer of the documentation to the counting centre.

Counting is carried out under the responsibility of the Head of the Canton Main Offices. The floppies were tabulated together in several computers running a tabulation software, and the results were printed out.

Afterwards, and for the first time in federal elections, an electronic copy of the results —a txt file— was uploaded with another computer to the MoI’s central server that generated a pdf file. This file was then sent back to the counting centre, signed electronically by the counting centre and sent again back to the MoI. In some places where OSCE/ODIHR EAM members attended the tabulation, the locally produced initial results and those received from the MoI were not checked against each other systematically and in detail.

Afterwards, and for the first time in federal elections, an electronic copy of the results —a txt file— was uploaded with another computer to the MoI’s central server that generated a pdf file. This file was then sent back to the counting centre, signed electronically by the counting centre and sent again back to the MoI. In some places where OSCE/ODIHR EAM members attended the tabulation, the locally produced initial results and those received from the MoI were not checked against each other systematically and in detail.

In this election, the College of Experts carried out for the first time a separate tabulation based on polling stations’ floppies with a specific software developed by the College itself.

Consideration could be given to :

• guarantee a hand over of the floppies from the Polling Station Chairperson to the Head of the Canton Main Office in person.
• ensure that a printout of the initial local results is given to political parties.

IX. COMPLAINTS AND APPEALS

Complaints concerning voter lists can be lodged with the municipal administration which is in charge of the compilation of the voter lists. The decision on a complaint by the municipal administration can be appealed to the Appeals’ Court.

Refusal to register a candidate because of ineligibility or of lack of proper
documentation can also be appealed to the Appeals’ Court [40], while other decisions of election offices cannot be appealed to courts.

Instead, the Constitution vests in the Chambers of Parliament the final say in adjudicating election disputes when validating their respective election results. [41] Complaints in regard to election results or distribution of seats, as well as election day complaints, are adjudicated by the newly elected Houses of Parliament. No possibility of appealing the election results or judicial review of such a decision is envisaged. The jurisdiction of the Houses of Parliament to validate the election of their own members, with no involvement of the Judiciary, proceeds from the Belgian interpretation of the separation of powers.

Procedurally, the two Houses of Parliament establish Committees on the Verification of Powers (Commissie voor het Onderzoek van de Geloofsbrieven / Commission de vérification des pouvoirs) on the day of their first plenary session. The House of Representatives established six committees of seven members designated by drawing lots among the newly elected members. Each committee validates the elections within one or two of the 11 constituencies. For the Senate, only one verification committee is established, comprising the seven oldest members of the newly elected Senate. On 28 June 2007, both the House of Representatives and the Senate held their first session.

After their designation, the Senate’s verification committee members withdrew to consider the elections’ results and complaints received by the Senate Office. [42] They had to adjudicate on a number of complaints requesting the invalidation of the elections, most of them arguing that the existence of the above-mentioned BHV constituency created discrimination among voters and was unconstitutional. The Committee argued that it was not within the jurisdiction of the Senate to decide on the constitutionality of the laws organising the elections, and dismissed the claims.

The Committee also had to decide on a case concerning the implementation of the rules for the designation of co-opted and Community senators. The Constitution (Art.68) provides that the total number of senators’ seats distributed to parties - which includes the 40 directly elected, the 10 co-opted and the 21 Community senator seats - must be proportional to their results in the direct election. Technically, community and co-opted senate seats are distributed among lists by following the order of the D’Hondt quotients obtained by party lists in the direct election contest. The 40 seats of directly elected senators are first allocated, then Community seats, and finally co-opted seats. For Community senators, Article 211 of the Code, provides that only political parties which have at least one directly elected senator, and have a group in the relevant Community Parliament, are entitled to receive Community senator seats.

Based on its results in the direct election among the Flemish electorate, the Lijst Dedecker [43] held one D’Hondt quotient which would have entitled it to obtain one of the ten seats of Flemish Community senators. However, since the Lijst Dedecker is not represented in the Flemish Parliament, the Senate decided on 5 July that the distribution of the Community seats would leave out the Lijst Dedecker and give the seat to the party who won the next D’Hondt quotient, in this case the OpenVLD.

There were discussions in the Senate as to whether the Lijst Dedecker should obtain a seat of co-opted senator in compensation. This raised another legal question : pursuant to the Constitution (Article 68 §1) the requirement to be represented in the relevant Community parliament does not apply to co-opted senators. Yet, Article 220 §2 of the code provides that the distribution of seats of co-opted senators starts after the distribution of the Community seats with the first D’Hondt quotient that follows the last quotient used for the designation of Community senators. In this particular case, the distribution of co-opted senator seats had to start lower in the order of D’Hondt quotients than the quotient that could have entitled the Lijst Dedecker a seat of Community senator. As a result, the Senate Verification Committee decided that Lijst Dedecker would not obtain either a Community seat or a co-opted seat.

This came after substantial debates in the Senate on whether or not the constitutional principle of proportionality should prevail over the letter of the Electoral Code. The Senate’s Department for the Evaluation of the Legislation issued a document in which it recommended that the constitutional principle of proportionality should take precedence over the provisions of the Electoral Code, which would imply that the Lijst Dedecker would get one more seat. Some elected candidates complained that such an opinion was beyond the jurisdiction of that department and rather up to the Constitutional Court or the Council of State to decide. Others objected that resorting to the opinion of any of the courts would be a breach of the separation of powers.

The Lijst Dedecker brought an action before the Brussels Court of first instance with a request for the court to issue a stay order on the list of co-opted senators and oblige the head of the Senate Office to change its content under threat of incremental penalties for non-compliance. During the debates surrounding the adoption of the report of the Verification Committee, several Senators protested what they considered an intolerable attempt to draw the Judiciary into interfering with the work of the Parliament, which they argue is a breach of the separation of powers.

On 28 June 2007, the House of Representatives held its first plenary session. [44] The Verification Committee in charge of the BHV constituency had to adjudicate on several complaints. One of the complaints argued that the e-voting system puts the control of the elections in the hands of a restricted number of computer specialists, and that it lacked citizens’ control. The Verification Committee dismissed the claim, arguing that it had an unspecified character, and that the relevant laws had been respected.

Six hundred and twenty-four complaints challenged the constitutionality of the BHV constituency, arguing that it created discrimination among voters and that, following the 2003 ruling of the Court of Arbitration, the constituency should have been changed. The Verification Committee submitted a report in which it suggested dismissing the claim on the basis that the elections had been called for 10 June, slightly before the end of the term, and thus ahead of the deadline imposed by the Court to solve the BHV issue. There were debates in which some newly elected MPs criticised the functioning of the Verification Committee as being problematic and undemocratic. The House approved the report and validated the election of the members from the BHV constituency.

Both the case of the Lijst Dedecker, and the case of the constitutionality of the BHV constituency, generated lively political debate within the two Houses of Parliament. Some members, including a representative of Lijst Dedecker, complained that the work of the Committees lacked transparency and was undemocratic.

Notwithstanding any possible considerations on the substance of these cases and their handling by the Verification Committees, the principle according to which it is up the winning parties in an election to act as the ultimate judges on election disputes is unusual and potentially problematic. The system could certainly call into question the impartiality of the adjudicating body and the effectiveness of the remedy available to complainants. In this respect, some interlocutors suggested that the Constitutional Court should be entitled to review the decisions of the Chambers of Parliament.

The OSCE/ODIHR inventory of commitments and other principles for democratic elections states that election contestants must have the opportunity to submit complaints on all aspects of election operations to a relevant court. [45] The Council of Europe’s Commission for Democracy through Law, Code of Good Practice in Electoral Matters 2002, states : “the appeal body in electoral matters should either be an electoral commission or a court. For elections to parliament, an appeal to a parliament may be provided for the first instance. In any case, final appeal to a court must be possible.” [46]

Notwithstanding the established legal basis for the existing complaint procedure, the new Parliament should consider measures to provide for impartial resolution of electoral disputes, including the possibility of an appeal to a court. [47]

ANNEX : ELECTION RESULTS [48]

HOUSE OF REPRESENTATIVES

Registered voters	7.720.796
Ballots cast	7.032.077
Invalid and blank	360.717
Valid votes	6.671.360

List	Votes	Percentage	Seats
FN	131.385	1.97 %	1
CDH	404.077	6.06 %	10
CD&V NVA	1.234.950	18.51 %	30
VLAAMS BELANG	799.844	11.99 %	17
MR	835.073	12.52 %	23
open vld	789.445	11.83 %	18
PS	724.787	10.86 %	20
GROEN !	265.828	3.98 %	4
sp.a-spirit	684.390	10.26 %	14
ecolo	340.378	5.10 %	8
VIVANT	5.742	0.09 %
BELG.UNIE-BUB	8.607	0.13 %
VITAL	1.780	0.03 %
FORCE NATIONALE	6.660	0.10 %
Wallon	8.688	0.13 %
PJM	4.373	0.07 %
R.W.F.	26.240	0.39 %
PC	19.329	0.29 %
FN B	9.010	0.14 %
CAP	20.083	0.30 %
NP-FN	1.605	0.02 %
VÉLORUTION	1.453	0.02 %
Belgique Positif	880	0.01 %
GSCD	170	0.00 %
UMP-B	1.408	0.02 %
MP Education	1.362	0.02 %
PLURALIS	757	0.01 %
CDF	11.961	0.18 %
PTB+	14.931	0.22 %
PVDA+	37.758	0.57 %
PTB+PVDA+	3.478	0.05 %
Parti Wallon	3.139	0.05 %
Lijst Dedecker	268.648	4.03 %	5
UNIE	856	0.01 %
DLC	464	0.01 %
FDB	901	0.01 %
TREFLE	920	0.01 %

SENATE

Registered voters	7.720.796
Ballots cast	7.032.384
Invalid and blank	404.257
Valid votes	6.628.127

Dutch-Speaking Lists

Valid votes	4,098,554
Threshold	204,927
Quotient	143,043

List	Votes	Percentage*	Seats
CD&V NVA	1.287.389	19.42 %	9
VLAAMS BELANG	787.782	11.89 %	5
open vld	821.980	15.38 %	5
GROEN !	241.151	3.64 %	1
sp.a-spirit	665.342	10.04 %	4
stijn	11.097	0.17 %
CAP	12.938	0.20 %
PVDA+	34.768	0.52 %
Lijst Dedecker	223.992	3.38 %	1
NEE	12.115	0.18 %

French-Speaking Lists

Valid votes	2,529,573
Threshold	126,478
Quotient	135,959

List	Votes	Percentage*	Seats
FN	150.461	2.27 %	1
CDH	390.852	5.90 %	2
MR	815.755	12.31 %	6
PS	678.812	10.24 %	4
ecolo	385.466	5.82 %	2
FORCE NATIONALE	14.866	0.22 %
R.W.F.	32.094	0.48 %
PC	19.632	0.30 %
CAP	8.277	0.12 %
CDF	13.319	0.20 %
PTB+	20.039	0.30 %

* Percentages are expressed in relation to overall nation-wide valid votes

ABOUT THE OSCE/ODIHR

The Office for Democratic Institutions and Human Rights (OSCE/ODIHR) is the OSCE’s principal institution to assist participating States “to ensure full respect for human rights and fundamental freedoms, to abide by the rule of law, to promote principles of democracy and (…) to build, strengthen and protect democratic institutions, as well as promote tolerance throughout society” (1992 Helsinki Summit Document). This is referred to as the OSCE human dimension.

The OSCE/ODIHR, based in Warsaw (Poland) was created as the Office for Free Elections at the 1990 Paris Summit and started operating in May 1991. One year later, the name of the Office was changed to reflect an expanded mandate to include human rights and democratization. Today it employs over 130 staff.

The OSCE/ODIHR is the lead agency in Europe in the field of election observation. Every year, it co-ordinates and organizes the deployment of thousands of observers to assess whether elections in the OSCE region are conducted in line with OSCE Commitments, other international standards for democratic elections and national legislation. Its unique methodology provides an in-depth insight into the electoral process in its entirety. Through assistance projects, the OSCE/ODIHR helps participating States to improve their electoral framework.

The Office’s democratization activities include : rule of law, legislative support, democratic governance, migration and freedom of movement, and gender equality. The OSCE/ODIHR implements a number of targeted assistance programs annually, seeking to develop democratic structures.

The OSCE/ODIHR also assists participating States’ in fulfilling their obligations to promote and protect human rights and fundamental freedoms consistent with OSCE human dimension commitments. This is achieved by working with a variety of partners to foster collaboration, build capacity and provide expertise in thematic areas including human rights in the fight against terrorism, enhancing the human rights protection of trafficked persons, human rights education and training, human rights monitoring and reporting, and women’s human rights and security.

Within the field of tolerance and non-discrimination, the OSCE/ODIHR provides support to the participating States in strengthening their response to hate crimes and incidents of racism, xenophobia, anti-Semitism and other forms of intolerance. The OSCE/ODIHR’s activities related to tolerance and non-discrimination are focused on the following areas : legislation ; law enforcement training ; monitoring, reporting on, and following up on responses to hatemotivated crimes and incidents ; as well as educational activities to promote tolerance, respect, and mutual understanding.

The OSCE/ODIHR provides advice to participating States on their policies on Roma and Sinti. It promotes capacity-building and networking among Roma and Sinti communities, and encourages the participation of Roma and Sinti representatives in policy-making bodies.

All ODIHR activities are carried out in close co-ordination and co-operation with OSCE participating States, OSCE institutions and field operations, as well as with other international organizations.

More information is available on the ODIHR website (www.osce.org/odihr).

+++OSCE+++