OSCE/ODIHR Election Observation Mission Final Report on the
Republic of Armenia Parliamentary Elections 25 May 2003*

EXECUTIVE SUMMARY

The 25 May 2003 parliamentary elections in the Republic of Armenia marked an improvement over the 2003 presidential election in the campaign and media coverage, but fell short of international standards for democratic elections in a number of key respects, in particular the counting and tabulation of votes.

The improvements mirrored a similar development during the 1998 presidential and 1999 parliamentary elections. The political leadership of the country made efforts to discourage violations, but perpetrators were again not held accountable. The authorities must end this atmosphere of impunity to give a clear sign of determination to meet their international commitments for democratic elections.

The Electoral Code, amended in 2002, generally provided a basis for the conduct of democratic elections. However, implementation, in particular in regard to the formula for the composition of election commissions, did not met expectations for proper administration of the elections. Genuine political dialogue is needed to identify a viable formula that ensures confidence of election stakeholders and professional and impartial performance of election commissions.

The election campaign was generally calm and quiet, with only isolated incidents of violence. A diverse choice was offered in the proportional contest, although there was no competition in a number of majoritarian races. There were equitable opportunities for campaigning and public television complied with its legal obligation to provide equal conditions to contestants. However, private television stations showed bias. Women continued to be seriously underrepresented as candidates and political party activists.

The effectiveness of the election administration was hampered by a general lack of consistency, transparency and professionalism. Important decisions were taken late and without sufficient clarity. Legislation on the registration of candidates, political parties and blocs was applied inconsistently and selectively. Attempts to implement a transparent process of tabulating results, a key recommendation of previous OSCE/ODIHR reports, was obstructed by the failure of Territorial Election Commissions (TECs) to act according to law. While voter lists were improved in some communities, further work is required to increase their accuracy.

Voting was assessed positively by observers in most polling stations visited. However, problems continued to be observed, including the presence of unauthorised persons in polling stations, undue restrictions on party and candidate representatives (proxies) and open voting by the military. Domestic observers were again present in large numbers and their legitimacy was more widely recognized than in the presidential election. Regrettably, a fatal shooting took place at a polling station, although this appears to have been unrelated to the elections.

Serious irregularities were noted in a third of all counts monitored by international observers. Violations included ballot box stuffing, falsification of results and intimidation of observers and proxies. To the credit of the authorities, these violations were generally acknowledged and some corrective steps were promptly taken, including re-runs of majoritarian elections in three particularly troubled constituencies. However, it remains to be seen whether violators will be held accountable. Re-run elections on 14-15 June generally showed improvement over the 25 May elections, although some shortcomings remained.

The complaints and appeals process remained inadequate and confusing, leading to inconsistent interpretation and application of legislation. However, greater willingness to file formal complaints was noted and the courts showed greater independence in striking down erroneous decisions of election commissions. In a welcome development, the Constitutional Court continued to demonstrate objectivity and independence, annulling the results in two constituencies and ordering re-runs because electoral violations influenced the outcomes.

Despite improvements, the significant drop in voter turnout illustrated that political parties and the wider public lacked confidence in the electoral process. In a follow up visit by the ODIHR and the OSCE Parliamentary Assembly to Armenia in July 2003, the authorities indicated a desire to improve the legal framework and conduct of elections. The OSCE/ODIHR stands ready to assist and offers recommendations which should be addressed swiftly if confidence is to be restored.

POLITICAL BACKGROUND

The 25 May 2003 parliamentary elections were the third parliamentary elections to be held since the declaration of independence in September 1991, and the first since Armenia joined the Council of Europe in January 2001. Previous elections in Armenia, including the two-round presidential election in February and March 2003, fell short of international standards for democratic elections.

Armenia's political landscape has undergone significant realignment since the last parliamentary elections in May 1999, and this was reflected in the configuration of the parties and blocs contesting the proportional list election in May 2003.

A key factor in that realignment was the 27 October 1999 attack on the National Assembly, in which the Prime Minister, the Speaker of Parliament and six other parliamentarians were assassinated. Issues of determining accountability for the assassinations contributed to a split between partners in the "Unity" bloc, the Republican Party and the Armenian People's Party, which had entered the National Assembly as the largest faction after the May 1999 elections. The Armenian People's Party went into opposition. Later, the same issues played a role in a split from the Republican Party by figures forming the opposition Republic Party. The Republican Party continued as the ruling party, with party leader Andranik Margaryan serving as Prime Minister since May 2000. The Armenian Revolutionary Federation (ARF) Dashnaktsutiun was also represented in the government, but was not in formal coalition with the Republican Party.

In a separate development, the two constituent parties of the "Right and Accord" faction, the other electoral bloc which entered the National Assembly after the 1999 parliamentary elections, ended their co-operation. Further, two of the four individual parties that won proportional list parliamentary representation in 1999 underwent splits - the Communist Party into two, and the National Democratic Union into four. Therefore, of the six parliamentary factions elected in the proportional list election in 1999, only two, the ARF Dashnaktsutiun and the Orinats Yerkir (Law-Based State) Party, remained intact throughout the mandate of the outgoing parliament.

Another realignment of parties affecting the 25 May elections was a direct consequence of the 2003 presidential election. The parties that supported presidential candidate Stepan Demirchyan in the second round run-off against the incumbent and ultimate victor Robert Kocharyan joined forces as the Justice Alliance bloc. Nine parties were represented in the bloc, including the Armenian People's Party and the Republic Party.


LEGISLATIVE FRAMEWORK

A. GENERAL OUTLINE
The legislative framework for parliamentary elections consists of the 1995 Constitution, the Electoral Code (adopted 1999 and amended 2002), the Civil Procedural Code, the Criminal Code, decisions of the Central Election Commission (CEC) and other laws. In addition, decisions of the Constitutional Court provide binding interpretations to the legislation.

The Constitution provides for the election of 131 deputies to the National Assembly every four years. Following the 2002 amendments, the Electoral Code provides for the deputies to be elected according to a mixed election system, with seventy-five seats allocated on a proportional basis to parties or blocs that gain more than five percent of the votes cast in a single national constituency and fifty-six deputies elected from single-mandate constituencies through a one-round majoritarian system. This is a reverse from the previous system, which provided for 75 majoritarian and 56 proportional seats. Governmental figures and political parties from across the spectrum have criticized the majoritarian system for serving candidates' individual economic interests and contributing to a weak and fragmented parliament. Most interlocutors have called for the majoritarian system to be modified to reduce the number of constituencies, or be abolished altogether.

B. ASSESSMENT OF THE LEGISLATIVE FRAMEWORK
In general, the legislative framework provides a substantive basis for the conduct of elections in compliance with international standards. However, practical implementation of the legislation during the 2003 parliamentary elections, and other recent elections, has revealed numerous shortcomings in the Electoral Code, which contributed towards the holding of elections that failed to meet international standards. Most of these shortcomings were previously identified in the OSCE/ODIHR final report on the 2003 presidential election1.

Members of the CEC and TECs were appointed according to a formula in the Electoral Code by which three members were nominated by the President and one member by each of the six parliamentary factions established by the 1999 elections. As with the 2003 presidential election, this formula, which purports to provide an effective balance of political interests, failed in this respect in both the proportional and majoritarian contests. Under a related formula, members of Precinct Election Commissions (PECs) were appointed to represent individual TEC members. This formula also failed to ensure political balance in polling stations.

The Electoral Code does not require adequate levels of independence and professionalism in the work of election commissions. Commission members are only under an obligation to operate "based on the principle of partnership" (Article 39.1), with insufficient legal requirement to work in an impartial, accountable and transparent manner. Election commission meetings may be held with a 50 percent quorum, currently five members, with decisions adopted if half of those present are in favour. Thus, key decisions can be adopted by only three members, which is not an adequate safeguard for a decision making process based on broad agreement.

The Electoral Code does not provide appropriate procedures for fair and transparent resolution of election disputes. The relevant provisions, contained in Article 40, are inadequately drafted and unclear in specifying the procedures to be used by electoral stakeholders in challenging violations or the decisions of election commissions. There is no requirement for an election commission to consider a complaint in quorum or in the presence of a representative of the complainant. In practice, this has allowed complaints to be decided by individual election commission members or administrative staff without a formal vote of the commission.

Under Articles 41(11) and 42(7) of the Electoral Code, superior election commissions have the power "to eliminate" the decision of a lower commission. However, there is no provision on the steps that then would follow, i.e. whether the lower commission is required to take its decision again, or whether the superior commission imposes its own decision. A key concern in this regard is the absence of any power for a TEC or the CEC to declare invalid the results from individual polling stations where violations have occurred. In such cases, the current practice of election commissions is simply to order the "verification" (i.e. a recount of) the results, which is an inadequate remedy to complaints alleging voting and counting violations, such as ballot-stuffing.
A decision of an election commission can be appealed to a Court of First Instance but no further. Article 40 provides for appeals to the Courts of Appeal and Cassation in cases relating to the registration of candidates, but this is contrary to Article 155(2) of the Civil Procedure Code, which was considered binding2. The lack of the right to appeal decisions of the Courts of First Instance across Armenia led to inconsistent interpretations and applications of the law during the parliamentary elections, for example, in appeals related to candidate registration.

The Electoral Code does not provide adequate sanctions against election violations. While Article 139 renders persons criminally liable for a number of actions, the Code does not address the specific steps that can be taken by election commissions against candidates or parties/blocs that have committed electoral violations ahead of election day. There are exceptional powers allowing election commissions to invalidate the registration of a candidate or party/bloc that prevents them from taking further part in the election, but such a remedy is not always appropriate3.

The Electoral Code lacks sufficient requirements for accountable and transparent tabulation and publication of preliminary and final results. There is no legal obligation for the CEC to publish either preliminary or final results broken down by polling station. Moreover, while Article 42(8) requires TECs to publicly post the preliminary results in this manner, there is no deadline for them to do so. In order to make the count and vote tabulation transparent, this data can and should be made public immediately and forwarded in this format to the CEC.
The rights and responsibilities of observers and proxies (i.e. the authorised representatives of political parties/blocs or candidates) are clearly expressed in the Electoral Code, with Article 30(5) stating "No restriction of the rights of proxies [and] observers…is allowed". However, a CEC decision of August 2002 restricted the rights of proxies through limiting their movement inside polling stations. The decision, which allows only one proxy at a time to move around a polling station, compromised the ability of proxies to observe the voting and counting procedures, and thus reduced transparency of the process.

The number of registered voters in each constituency cannot differ by more than 15 percent. However, Article 17(1.1) of the Electoral Code does not specify the procedure to be used to calculate the differences. It is not possible to confirm whether the sizeable factual differences that exist between constituencies are within the limits permitted by law.

Voters are allowed to vote only in the polling station where their names are included in the voter list. This prevents voters who are hospitalized, housebound, temporarily away from their local area or otherwise unable to attend their polling station (such as police on duty) from voting. In contrast, registered voters who are resident overseas and members of the military are given opportunities to vote, although only in the proportional election.

A strength of the Electoral Code is the guarantee of equal opportunities to persons with disabilities as voters. In addition, a decision of the CEC required polling stations to be located in premises that allow easy access by all voters4. However, more than half of the polling stations visited by IEOM observers on election day were not accessible by disabled persons.

The Electoral Code requires updating to incorporate decisions of the Constitutional Court that have supplemented the election legislation. Further, there are numerous instances where the text requires stylistic re-arrangement. A number of legal provisions lack technical details, in particular in relation to the production of election materials and the exact procedures to be followed when separate elections are held on the same day, specifically on the use of ballot boxes. Comprehensive procedures to account for the political choice of those who voted "against all" are also required.

C. ACCOUNTABILITY FOR VIOLATIONS OF THE LEGAL FRAMEWORK
The OSCE/ODIHR has previously identified the most serious problem with the legal framework to be in the failure of the authorities to enforce its provisions and to hold accountable those persons responsible for violations of the law. During the 2003 presidential election, widespread electoral violations are acknowledged to have taken place, many of which were recorded by international observers. Furthermore, around 30 complaints relating to alleged criminal violations of the law were made to the CEC and/or the Office of the Prosecutor-General. In a ruling on disputed election results, the Constitutional Court also held that violations had occurred in 40 polling stations and ordered their investigation by the Office of the Prosecutor-General. As a follow-up action, the EOM for the parliamentary elections observed closely the procedures by which the authorities would hold accountable those responsible for previously committing violations.

No election official has been held accountable, either criminally or administratively, for their role in violating the Electoral Code during the presidential election. Indeed, senior commission members of some TECs where numerous violations occurred, such as the Chairman of TEC 29, remained in office with full responsibility to administer the parliamentary elections. Although EOM observers reported widespread replacement of individual PEC members between the two elections, there was no indication that the replacement was limited to those polling stations where violations occurred. Indeed, observers noted cases where PEC members responsible for violations were simply moved to other polling stations in the same or neighbouring TECs. The only recorded instance of persons being held accountable related to eight police officers disciplined in relation to their failure to prevent an assault at a campaign rally.

Of particular concern is the failure of the Office of the Prosecutor General to take effective steps to investigate the complaints referred to it. While suspects were arrested in relation to breaches of the Criminal Code, investigations of violations of the Electoral Code, such as cases of ballot stuffing, either remained in their preliminary stages or had been closed on the basis of no available evidence.

Failure by the relevant authorities to take measures against election violations in the presidential election has undermined the credibility of, and public confidence in, elections in Armenia.

Moreover, such inaction contributed towards a sense of impunity by election officials and led to further violations of the Electoral Code in the parliamentary elections. To allow more effective and prompt action to be taken against violators, the Electoral Code should be amended to allow for the dismissal of those election officials found by a superior election commission to have been responsible for an electoral violation, or to have permitted it to have taken place. More positive, however, was the role of the Constitutional Court in protecting the constitutional rights of voters and candidates following the presidential election. The Court approached disputes relating to the results of the election in an active and independent manner. Further, during the parliamentary election campaign, genuine efforts were taken by the highest levels of authorities to prevent a repetition of some of the shortcomings that marred the presidential election. In particular, officials supported the election administration to adopt a number of the recommendations contained in previous OSCE/ODIHR reports. These steps contributed to the improvements noted in this report.


ELECTION ADMINISTRATION

The election was administered by a three-tier commission structure: the CEC, 56 TECs and approximately 1,885 PECs. Around 35 polling stations were also established in diplomatic missions of Armenia. While PECs are temporary bodies for the duration of the elections, the CEC and TECs function on a permanent, four-year basis. New commission members took office on 10 July 2003.

A. COMPOSITION OF ELECTION COMMISSIONS
All election commissions were composed of nine members, appointed in accordance with the formula described above (see Section IV: Legislative Framework). Of the six factions established following the 1999 elections, only two remained intact throughout the mandate of the parliament. However, parliamentary rules of procedure, which forbid the formation of new parliamentary factions, deemed the six factions to continue to exist in their original configuration for the purposes of election commission membership.

As a consequence, all "Unity" representatives on election commissions came from the governing Republican Party, which had a majority of deputies from the former bloc. The People's Party of Armenia was thus without formal representation on any commission. Similarly, all appointees from two other factions, the Communist Party and the National Democratic Union, came from the rump leadership of the original parties and did not include representatives of the five new parties that were created from their respective break-ups. However, the Right and Accord bloc had proportionally allocated the appointment of commission members between its two constituent parties prior to their break-up, which were retained by the individual parties.

In contrast to the second round of the 2003 presidential election, the formula was able to achieve a limited level of representation on election commissions by the leading opposition political force. The Justice Alliance bloc, which included the People's Party of Armenia, was represented by appointees from the National Democratic Union. Moreover, in around a third of TECs, the Justice Alliance bloc was also represented by appointees from another of its member parties, the Constitutional Right Union, which held some of the nominations from the former Right and Accord faction. Nevertheless, there remained on the CEC and each TEC a de facto majority of members representing the presidential administration and the two parties of government.

The formation of PECs was generally completed within the legal deadline. The formula for the appointment of PEC members is also intended to achieve, but fails to guarantee, representation of the different parliamentary factions in every polling station. In practice, there were widespread instances where there was no attempt for those appointed to PECs to reflect the political representation envisaged by the law. Many PEC members appeared to be unaware of the faction they were supposed to represent. Indeed, the EOM is aware of many claims of alleged corruption in the appointment of PEC positions. These could not be substantiated.

Most political parties and many candidates expressed a significant lack of confidence in the objectivity, impartiality and transparency of the election administration. The dominant position of appointees of the President and the two parties represented in government, who held the chairs of the CEC and all but one TEC, did not foster trust in the work of the commissions. The chairs of over 90 percent of PECs visited by international observers were presidential appointees as well.

B. ADMINISTRATION OF THE ELECTIONS
The CEC held overall supervisory jurisdiction over the parliamentary elections and specific responsibility for the administration of the proportional list contest. The 56 TECs were directly responsible for the administration of the single-seat majoritarian contest in their constituencies. A training programme, implemented by the United Nations Development Programme (UNDP), was provided to most TEC and PEC members.

In general, election commissions completed preparations for the elections on time despite pressure experienced by a number of TECs to meet deadlines for appointment of PECs and delivery of election materials. In order to implement the provisions of the Electoral Code, the CEC took over a hundred decisions, largely in a timely manner and containing sufficient detail. However, the publication and dissemination of CEC decisions to lower election commissions and other election stakeholders, including candidates, was not systematic.

As in the 2003 presidential election, the CEC took the positive decision to use transparent ballot boxes. However, this decision was adopted very late (12 May). The CEC had earlier (30 April) decided that only one large ballot box could be used in each polling station. This was based on the CEC's interpretation that the law prevented the use of more than one ballot box, although it is in fact ambiguous on the issue. As there were no technical facilities to produce large transparent ballot boxes, the 30 April decision meant that ballot boxes would not be transparent. On 12 May, the CEC reversed its opinion and ordered the use of separate transparent ballot boxes for the parliamentary elections and the referendum vote. A number of political parties and the EOM had commented publicly that the use of non-transparent ballot boxes would have been a backward step5.

Several key decisions taken by the CEC in the later preparatory stages of the election were insufficiently detailed. For example, the decision to use more than one ballot box in polling stations did not specify the voting and counting procedures to take account of the new ballot box arrangements or the steps to be taken to ensure the integrity of uncounted ballot papers. This led to non-uniform implementation on election day.

In another positive development, the CEC formally instructed TECs to comply with the legal requirement, under Article 42(8) of the Electoral Code, to publish the preliminary results of both the proportional and majoritarian contests broken down by polling station. In previous elections, the provision had been widely ignored. However, this instruction was again given late (22 May) and was delivered orally rather than in writing.

The Chairman of the CEC made a personal commitment to publish the preliminary results of the proportional election broken down by polling station, to comply with important recommendations made by previous IEOMs. However, he was not able to fully meet his commitment.

Provisions of the Electoral Code were applied inconsistently by TECs. In particular, the registration of candidates was undertaken using different criteria for declarations on property ownership, citizenship and financial circumstances. The CEC, which is responsible to "oversee the equal application of [the Electoral] Code", failed to provide sufficient guidelines on candidate registration to ensure consistency amongst TECs.

Following decisions of the Constitutional Court in April 2003 on disputes relating to the results of the presidential election, the CEC was required to adopt new procedural regulations on the complaints and appeals process for election commissions. However, the regulations only formalised the previously existing ad hoc procedures that were criticised by the Court. A key concern with the procedures is that they allow a complaint to be decided upon by only one commission member without referral to the commission as a whole.

There was a general lack of transparency in the manner in which the superior election commissions operated. Most decisions of the CEC were determined by its executive officers and secretariat outside of formal sessions, while the CEC Chairman and Deputy Chairman regularly held informal meetings with individual TEC Chairpersons that were not public. Although opposition representatives on election commissions were less marginalized than in the 2003 presidential election, CEC and TEC sessions continued to be short, infrequent and not conducive to debate or discussion. All formal sessions of the CEC were open to proxies, accredited observers and the media. The EOM was informed of a few instances where proxies were obstructed from attending TEC and PEC meetings because of a failure to properly announce the sessions.

In general, the EOM was concerned that security measures to protect the integrity of election materials were insufficient to prevent potential abuse, especially in relation to the printing and distribution of ballot papers and other sensitive materials, although no lapses were confirmed by the EOM. In contrast to the presidential election, the EOM observed fewer instances of ballot papers being handled outside of polling stations.

C. VOTER LISTS
Under the Electoral Code, responsibility for the compilation and maintenance of voter lists rests with the bodies of local self-government known as communities6, under the "control" of the CEC and TECs. Each voter list must also be signed by a representative of the Passport Department of the Police Service of Armenia7. However, the Code is insufficiently detailed on the roles of the different agencies and the level of control by the CEC and TECs. Moreover, it is silent on the de facto use of Police passport records, which provide crucial data for the updating of voter lists. Armenia has no centralized civil or voter register to safeguard against potential multiple entries across community borders.

As in previous elections, the accuracy of voter lists was of concern and most political parties have little confidence in them. The EOM received reports that co-operation between the various Armenian institutions responsible for civil registration matters was imperfect and often failed to delete the records of deceased persons.

Significant improvements have recently been made in the quality of voter lists in a number of communities following co-operative efforts between community leaders and civil society organizations, with the support of the international community.

Voter lists were displayed in most polling stations for 40 days ahead of the elections enabling voters to check and amend data8. The final voter lists were posted at polling stations two days ahead of the election. Eligible voters omitted from the voter lists could apply to a Court of First Instance, including on election day, to obtain a court certificate entitling them to vote9.

Citizens of Armenia residing abroad have the right to vote in parliamentary elections and to be included in the voter lists of the diplomatic mission near their current residence or the polling station of their permanent residence. Although there is no accurate record, it is generally acknowledged that a large number of Armenian citizens reside out-of-country10. Again, there were widespread allegations of abuse of these circumstances. Voters registered abroad, as well as military personnel and detained persons have the right to vote only in the proportional election, and not in the majoritarian contest.

On 22 May, the CEC announced that there was a total of 2,317,945 registered voters entitled to vote in the proportional election and 2,280,557 voters for the majoritarian contests. According to information supplied by the CEC, around 4,050 voters, who were not included in voter lists, were issued with court certificates permitting them to vote. However, in its protocol of final results issued on 31 May, the CEC recorded the total number of registered voters as 2,340,744. This discrepancy has not been explained.

REGISTRATION OF PARTIES, BLOCS AND CANDIDATES

A. LEGAL FRAMEWORK
In order for parties or blocs11 to nominate a list of candidates in the proportional election, their lists must be supported by the signatures of 30,00012 registered voters and a financial deposit of 2,500 times the minimum salary. Each candidate on the list must supply the CEC with proof of five-year citizenship and permanent residence in Armenia, and a declaration of their private property and family members' income. Five percent of candidates on all proportional lists must be women, although there is no rule on their placement within the list. Candidates nominated in a list do not need to be members of the party submitting their list; for example, the second-placed candidate in the Republican Party list, the Minister of Defence Serge Sargsyan, was not a member of the party.

The Electoral Code does not allow blocs to nominate candidates for majoritarian contests, a restriction that has no rationale. In addition to those from political parties, non-partisan majoritarian candidates can be nominated upon the "citizens' initiative" of at least 50 voters. To be registered by the TEC, nominations must be supported by the signatures of at least 500 voters registered within the boundaries of the TEC and a financial deposit of 100 times the minimum salary. Once again, a candidate must provide proof of their five-year citizenship and permanent residence in Armenia, and financial information on their property and family's income.

The Electoral Code prescribes a formula for verifying signatures that only requires two percent to be checked; if a number of invalid signatures are found, the number is multiplied by 50 to assess proportionally how many invalid signatures are present in the total signatures submitted. This is an inadequate formula for verifying signatures and is open to abuse, especially given the relatively small number of signatures needed by majoritarian candidates. The requirement for candidates to prove five-year "permanent residence" is inconsistent with the law regulating residency issues, the Civil Code of Armenia, which has no such concept13.

Article 97(2) of the Electoral Code prevents persons holding public office from standing as candidates in majoritarian constituencies, unless they resign ahead of registration. No such restriction applies to the proportional election. Thus, two nominees were refused registration as majoritarian candidates (TECs 30 and 51) for having failed to resign as Community Heads, while the Mayor of Yerevan and the Minister of Defence were able to register as proportional list candidates, and continue in their official positions throughout the campaign. The restriction is not consistent, and alternative safeguards against potential abuse of public office should be identified.

B. REGISTERED PARTIES BLOCS AND CANDIDATES
Seventeen political parties and four electoral blocs, involving over 1,000 candidates, were registered by the CEC to take part in the proportional election, offering voters diverse choice. Of the two parties in government, the Republican Party fielded 120 candidates and the ARF "Dashnaktsutiun" 101 candidates. The Justice Alliance bloc fielded the longest list, with 136 candidates14. Members of some minor parties were included in the lists submitted by larger parties to which they were affiliated but not in formal alliance.

At the close of the candidate registration, 10 April, TECs had registered some 380 candidates for the 56 majoritarian contests. Subsequent withdrawals and appeals left a total of 272 majoritarian candidates on election day. The significant number of withdrawals, almost a third, came from across the political spectrum. Many of those who withdrew gave only general reasons for doing so, although some confirmed political deals between parties and candidates. Allegations that some were due to bribery or intimidation were not confirmed.

No constituency race reflected the proportional list contest in terms of the range of competing political interests: the highest number of candidates in a single constituency was 11. Nine had limited contests between two candidates, while in four constituencies there was just one candidate. Many candidates were locally-renowned figures, often from the business community, and some contests reflected local rivalries.

Well over half of majoritarian candidates were "non-partisan", although it was clear that many had connections to government or political parties. No political party attempted to cover more than a third of the 56 constituencies. Of the 17 parties contesting the proportional list election, 13 also fielded majoritarian candidates, as did two parties that for the proportional election had entered candidates on another party's list. Several member parties of the four blocs contesting the proportional election fielded candidates in their own right. In two Yerevan constituencies, parties in the Justice Alliance bloc ran against each other. Two minor parties not involved in the proportional election each fielded a single candidate in a majoritarian constituency.

A total of 37 political parties competed in one way or another in the elections. A number of candidates were nominated in both the proportional and majoritarian contests.

C. REFUSAL OF REGISTRATION OF CANDIDATES
Based on appeals filed with Courts of First Instance, the EOM learned of 22 instances where potential majoritarian candidates were denied registration by TECs, or were subsequently de-registered for failing to meet certain legal requirements.

In at least six cases, candidates were refused registration on the basis of a failure to declare fully their personal property. In these cases, the legal interpretation by TECs of the definition of "personal property" was often spurious as it included circumstances where candidates were deemed to possess an equity interest in property belonging to other family members. Neither the CEC nor any TEC gave guidelines to candidates on what property needed to be declared, leading to inconsistent application of the criteria. EOM observers also confirmed that property checks with the local authority were made selectively against certain candidates, mostly those linked to opposition political forces. At least seven nominations were rejected on the grounds of invalid or false signatures, with no opportunity provided for the potential candidates to remedy the deficiency. In one case where the boundaries of the constituency administered by TEC 13 had recently been changed, a nominee was rejected for obtaining signatures of voters who were previously, but no longer, resident within this constituency.

In seven of the 22 cases, the decision of the TEC was overturned by a Court of First Instance, indicating a welcome independence of the judiciary and enabling candidates to be registered or reinstated. In two cases involving TEC 30, the Court held that valid signatures had been wrongly declared to be false.

The CEC decided to de-register one candidate from the National Unity Party list on the grounds of failure to declare fully his property. The decision was appealed and overturned by the court, which reinstated the candidate. The case reflected the lack of consistency in the registration process as the same person had been denied registration as a majoritarian candidate on the same grounds by TEC 8, in a decision that was upheld by another court.

One of the leaders of the Justice Alliance bloc, Aram Karapetyan, was removed from the list of registered candidates in both contests, following decisions taken by TEC 17 and the CEC that he did not meet the five-year permanent residency requirement. The decisions were taken following information supplied to TEC 17 by the Armenian Police, ostensibly on the grounds of "media speculation". According to the Police, Mr. Karapetyan had been registered as a permanent resident in the Russian Federation during the past five years. Mr. Karapetyan stated that he had remained registered as a permanent resident in Armenia throughout the period, as was his legal right. Incongruously, Mr. Karapetyan had been registered as a presidential candidate in January 2003, which required that he meet a ten-year residency requirement. While details remained unclear, the EOM is concerned that the decisions lacked transparency and were taken despite unclear legal definitions of permanent residency.

ELECTION CAMPAIGN

The election campaign took place in a generally calm and quiet environment. In a positive development from the presidential election, there were no reports of significantly unequal opportunities for campaigning, particularly in the proportional contest. Active campaigning took place throughout the country, featuring posters and public meetings. Interest in the election was relatively low, reflecting voter fatigue as well as some disillusionment resulting from the conduct of the recent presidential election. Of note was the tendency of parties which supported President Kocharyan, particularly the two parties in government, to openly campaign against each other.

Campaigning was generally more active in the majoritarian contests. Pre-election tensions were noticeably higher in some constituencies, which reflected local rivalries. The limited contests in constituencies with one or two candidates appeared to derive from a reluctance to challenge the interests of dominant community figures, especially economically powerful ones. Persistent allegations of coercion of voters to support particular candidates in Alaverdi (Lori region) and in Syunik were made to EOM observers, although no formal complaints were filed. Few allegations of intimidation were made. There were attacks on candidates' campaign offices in three constituencies, Ararat, Abovyan and Sevan. In the last of these, rival candidate campaign offices were both attacked at different times. There was also an assault on a TEC member in Syunik region. In a welcome step, the CEC issued a public statement strongly condemning all acts of election-related violence and called for participants to show tolerance.

Some parties violated provisions in the Electoral Code which prohibit providing or promising free goods or services to citizens during the pre-election period. Complaints received by the EOM included the distribution of "gifts" of fuel, food and computers, while the EOM confirmed instances of the asphalting of roads and repairs to apartment buildings. While few such complaints were made formally to the electoral authorities, the CEC and TECs generally failed to take steps to fully investigate compliance with the law. Allegations of vote buying before election day were widespread but not substantiated.

There were isolated instances of public resources being used and public employees, especially teachers, inappropriately involving themselves in campaigning to the advantage of certain parties or candidates. Although the Electoral Code prohibits the involvement of charities in the electoral campaign, there was an overlap between the activities of a charitable organization in Hrazdan and a candidate's campaign.

Provisions for the supervision of campaign spending by candidates and parties/blocs in the Electoral Code15 are ineffective because the limits allowed are unrealistically low. The provisions also lack regulatory sanction for exceeding the limits, except for de-registration. Prior to an election, candidates and parties/blocs must open bank accounts through which all their income and expenditure for pre-election campaign activities must be channelled. Bank statements must be submitted to the CEC or TECs every three days. Final declarations must be submitted to the CEC within 15 days of the election day. A limited enquiry into pre-election funds by the EOM revealed that in most cases the accounts were "shells" that did not reflect the real incomes/expenditures, and in a number of instances were not even opened by candidates.

MEDIA

A. MEDIA LANDSCAPE
Television is by far the dominant source of information about politics in Armenia. Public TV, H1, is among the few broadcasters with republic-wide coverage and is the most influential media outlet. Apart from public TV, approximately 45 TV stations operate in the country, more than 20 of which are based in Yerevan. The failure to renew licences to two established independent broadcasters, A1+ and Noyan Tapan, remains a contentious issue. Two State-established bodies supervise and regulate the electronic media: the National Council on Radio and Television (NCRT) licenses and oversees the private broadcasters, and the Council of Public Television and Radio governs the State-funded media.

Two State-funded and several private newspapers offer diverse but polarised opinions. However, due to their small circulation and generally poor economic conditions, most newspapers are financially dependent on sponsors and have a limited outreach.

While freedom of the media generally exists in Armenia, the information provided lacks sufficient variety, quality and independence to allow voters to make a well-informed choice. No incidents of election-related intimidation or harassment were reported against journalists during the parliamentary elections, though previous incidents of violence were considered by some journalists as continuing to negatively affect their work, leading to self-censorship.

B. LEGAL FRAMEWORK
The legislation regulating media behaviour during an election campaign generally meets international standards. The Constitution guarantees freedom of speech and the right to receive and disseminate information and ideas through any medium of information. The Law on Television and Radio Broadcasting prohibits State censorship and obliges State-funded television and radio to provide objective and unbiased information. However, the Law does not contain similar provisions to cover the private electronic media. The Law on Mass Media requires the print media to be guided by the principles of equality and diversity of opinion.

The Electoral Code stipulates that the State-funded media must provide equal conditions to parties and blocs taking part in the proportional election, including access to free and paid air-time and print-space. On 28 March, the CEC adopted a resolution that provided specific details on the procedures for allocating airtime and print-space equally. The decision also obliged the private broadcasters to provide parties/blocs and candidates in the proportional and majoritarian contests constituencies with airtime on equal basis.

An amendment to the Criminal Code signed by the President on 29 April, which protects State officials from defamation through mass media, could compromise the freedom of journalists in their coverage of future elections.

C. MEDIA MONITORING
The EOM conducted quantitative and qualitative analyses of media coverage of the campaign by selected media outlets from 25 April until 23 May. The monitoring included five TV channels (H1, Prometevs TV, TV Alm, Shant TV and Armenia TV) and six newspapers (Hayastani Hanrapetutyun, AZG, Aravot, Golos Armenii, Hayots Ashkharh and Orran).

In an important and positive development in comparison with the presidential election, H1 fulfilled its legal obligation to grant free time to parties and blocs under equal conditions in accordance with the schedule determined by the CEC decision of 28 March. Moreover, it generally provided unbiased coverage of the campaign in its prime-time news and analytical programs, both in quantity and quality. Overall, H1 offered a comparable portion of its coverage to both the governmental parties, the Republican Party (9.6%) and ARF Dashnaktsutiun (9.5%), and major parties and blocs not in government - National Unity Party (11.4%), Justice Alliance bloc (11.1%) and Orinats Yerkir (10.4%). While the portrayal of the Justice Alliance bloc was mainly neutral throughout the campaign, coverage of the governing Republican Party became progressively more positive nearer to election day.

In contrast, private broadcasters monitored by the EOM generally showed bias. Prometevs TV and TV Armenia, the only two private channels with republic-wide outreach, allocated up to 60% of their prime-time news to pro-governmental parties. The coverage was significantly positive. In contrast, the Justice Alliance bloc and National Unity Party were largely ignored, accounting for less than 1% of news coverage. In a particularly blatant manner, ALM TV showed strong preferential treatment, both in quantity (33%) and quality, to the People's Party, a minor party taking part in the proportional election, which is headed by the owner of the channel. While Shant TV was more balanced, its coverage of the Republican Party was noticeably positive (20.4%). The EOM also noted several instances where paid political programs on ALM TV were not identified as such, contrary to the CEC decision of March 2003. The NCRT reported receiving no complaint related to private television coverage of the election.

Following the first TV debate between the two main contestants in the presidential election earlier in the year, public TV and many local private broadcasters organized numerous debates between candidates and offered various formats of discussion and analytical programmes on election issues. TV spots providing voter education information were supported by the international community.
The print media again provided a plurality of views, but all newspapers showed bias. More positively, the print media paid greater attention to important issues generally not covered by television, such as the platforms of smaller parties and non-partisan majoritarian candidates.

The main State-funded newspaper, Hayastani Hanrapetutyun, showed clear support to the governing parties, allocating 20.5% of space to the Republican Party with an overwhelmingly positive slant. In contrast, the newspaper allocated just 5.6% of print space to the Justice Alliance bloc, with a neutral or negative tone, and 1.8% to the National Unity Party. Thus, the paper violated its legal requirement to provide equal conditions for parties16.

Of the private newspapers monitored by the EOM, Orran showed clear support to the Justice Alliance bloc, allocating it 23.6% of space with positive coverage. While Aravot was critical of the pro-government parties, Hayots Askharh and Golos Armenii showed support to these parties. AZG, allocated 62% of overwhelmingly positive coverage to its founder, the Ramkavar Liberal Party. The 24-hour campaign silence period was breached by Hayots Askharh, which ran interviews with two non-partisan candidates.

In contrast to the presidential election, the media accurately covered the release of the IEOM Statement of Preliminary Findings and Conclusions and related press conference on 26 May. Hayastani Hanrapetutyun carried the statement as a paid advertisement, which it had refused to do during the presidential election.

COMPLAINTS AND APPEALS

Appeals against the decision of an electoral commission may be lodged with a superior election commission or a Court of First Instance. In practice, however, TEC decisions related to the majoritarian contest were not considered by the CEC to be within their jurisdiction except in an advisory role. There is no appeal permitted from the Court of First Instance.

As in previous elections, the complaints and appeals procedures were inadequate. The procedural framework was confusing and led to inconsistent interpretations and application of legislation. While political parties and the wider public continued to lack confidence in the independence of the judiciary and the quasi-judicial role of the CEC, there was an improved willingness to file formal complaints by aggrieved persons and a more active role by the courts to quash administrative errors by election commissions.

Other than on issues related to candidate registration, few substantive complaints were filed with the CEC and TECs during the pre-election period. However, from election day to the close of the 48-hour deadline for the receipt of complaints (28 May), the CEC received 60 complaints. The majority of these related to events in polling stations that the CEC considered outside of its jurisdiction, and referred to TECs to investigate. As already noted, most decisions of the CEC were taken in closed sessions, often by individual commission members. No information was available on the number of complaints received by individual TECs.

The Office of the Prosecutor-General received 17 complaints of violations of the Criminal Code in relation to the campaign and election day. A number of these cases referred to incidents of violence and other breaches of the Criminal Code. As in the presidential election, the Office of the Prosecutor-General appeared reluctant to investigate reports of serious election violations that had been brought to its attention, including those observed by the IEOM on election day.

PARTICIPATION OF WOMEN

The elections did little to improve the low representation of women in the National Assembly. Only seven (5.3%) women were elected, a marginal increase from the four (3%) in the outgoing parliament17. Women were also under-represented as candidates - only 4% of majoritarian candidates and 15% of proportional lists were women. All political parties and blocs met their obligation to include 5% of women in their candidate list but most were placed in positions that were effectively unwinnable. Only three political parties or blocs had women among their first ten named candidates, and only one list had women placed evenly throughout. Most political party leaders, particularly outside Yerevan, expressed no interest in encouraging greater participation of women in political life or in addressing issues of particular concern to women.

There remains serious under-representation of women in government, elected bodies and political party leaderships18. Some political parties have women's sections and during the campaign a number of parties made efforts to address women's concerns, including through the organization of women's forums. However many women object to the concept of targets or quotas for women's participation as a reversion to the Communist period when women were represented more widely, albeit symbolically, in the legislature and executive.

DOMESTIC OBSERVERS

Twenty-eight domestic observer groups were accredited by the CEC to observe the elections. The most substantial domestic observation effort was carried out by "Its Your Choice" (IYC), which observed the pre-election period and deployed observers to around 700 polling stations throughout the country on election day. Other groups focused on specific elements of the election, such as media monitoring or disabled persons' access. Public statements by domestic observers were generally well-covered by the media.

OBSERVATION OF VOTING AND COUNTING19

A. VOTING
Election day was generally calm, but was marred by a fatal shooting outside a polling station in Shahumian (Ararat region), which also resulted in the injury of others inside the polling station. IEOM observers were present at the incident, but were unharmed. The person alleged by police to be responsible for the shooting later died from injuries sustained. There was no evidence to contradict police reports that the shooting was unrelated to the elections. Isolated incidents of violence were also reported in a number of other polling stations, including a fight in PEC 0272.

The Chairman of the Parliament's Foreign Affairs Committee withdrew as a majority candidate in TEC 45 on election day alleging intimidation. IEOM observers reported other acts of intimidation in three percent of polling stations visited.
In around 90% of the polling stations visited by IEOM observers, voting was assessed as either positive or with only minor problems. Nevertheless, problems were observed, including one instance of vote-buying at PEC 1287; carousel voting at 0851 and 1447; unused ballots being removed from 0376 and distributed to persons outside; persons in possession of multiple passports at 0029, 0413 and 0622; and two voters allowed to vote twice at 0947. In at least 13 polling stations, more than one person was observed in the voting booth in conditions that did not require assisted voting. Open military voting was observed at PECs 0955 and 0256 taking place in a systematic manner for over one hour. Observers reported widespread allegations of double-voting on election day, although very few instances were confirmed.

The presence of unauthorized persons was observed in 10% of polling stations, and in at least 11 instances such persons were seen to be directing the work of the PECs. Armed police officers were seen in 20 polling stations, half of whom had not been formally invited by the PEC Chair. Private persons with firearms were observed in 0013 and 0028.

There were a number of polling stations where security of voting materials appeared compromised. For example, full ballot boxes did not have their slots sealed when taken out of use in five polling stations, and ballot boxes were not sealed tightly in 11 polling stations. Moreover, seals (stamps) when unpacked at the opening of 0124, 0623, 1650, 1696 were observed to have apparently been used very recently beforehand, suggesting preparation of fraudulent ballots. In 0442, the seal (stamp) was temporarily removed from the polling station. Fewer ballots were delivered to 0668 and 1650 than the number recorded by their TECs. Ballots that appeared to be photocopies were observed being used at 0620 and 0621.

On average, around ten proxies were present in each polling station visited by IEOM observers and domestic observers were present in more than half of polling stations visited. Intimidation of proxies was observed in a small number of polling stations, including a serious incident in 0813, where IEOM observers were required to escort the proxy to a safe place. There were reports that proxies were not allowed to register complaints because of the restriction (under Article 57(2) of the Electoral Code) that one proxy cannot do so and a small number of instances where complaints by more than one proxy were not registered.

B. COUNTING
Significant problems were observed during the counting process in one third of polling stations where the vote count was observed in its entirety (28 out of 84 counts). In 17 cases, polling station officials were seen to act in deliberate violation of the Electoral and Criminal Codes.

There was falsification of result protocols in nine PECs (0347, 0378, 0408, 0553, 0800, 0814, 1086, 1236 and 1335). In 1236 this included the destruction of valid ballots to make figures tally. Additionally, in 0385 a PEC member falsified signatures on the voter list. In five polling stations (0239, 0378, 0347, 0408 and 0379) the counting of ballot papers for each candidate or party was done in silence, with results communicated among PEC members in whispers or by note, to deliberately compromise the transparency of the process.

Interference with ballot materials in conditions that enabled fraud was observed in five PECs: in 0272 an open ballot box was overturned; in 0408 and 0553 sealed envelopes of counted ballots were reopened; in 0711 ballot papers were temporarily removed from the room where counting was taking place; and in 1236 unused ballot papers were not invalidated. Illegal possession of ballot papers by a PEC Chairman and a proxy was seen at two polling stations (0347 and 0272). The theft of counted or uncounted ballot papers took place in three PECs (0272, 0266 and 0408).

Evidence of ballot stuffing was seen in nine polling stations and stuffing took place during counting in 0090 and 0347. Unsealed ballot papers found in the ballot boxes at 0090 and 0814 were sealed (stamped) during counting, and unsealed (unstamped) ballots found in the box at 1236 were declared valid. Two different colours of seals (stamps) were seen on ballot papers in 1335 with both accepted as valid, as was an obviously photocopied ballot paper. There was a selective approach to the invalidation of ballots in 0385 and 1236.

Other irregularities observed included serious procedural problems, such as errors in the process of completing protocols. A number of such problems indicated the absence of formal instructions from the CEC on the procedures to be followed during counting. Observers also noted unexplained delays before the start of counts. In some polling stations, counting procedures were observed to be confused, even chaotic. PEC Chairpersons unlawfully refused to provide result protocols to IEOM observers in PECs 0329, 0385, 0495, 0800 and 1510.

International observers experienced serious intimidation in their work and threats to their security in TEC 29, despite warnings given to the authorities by the IEOM that violence and intimidation could be expected in the area. Elsewhere, an IEOM observer was refused entry to PEC 0212, observers had their movements restricted or were prohibited by PEC chairpersons from speaking with other PEC members or proxies in 0112, 0294, 0328 and 0626, and observers encountered confrontational or abusive behaviour from PEC members in 0112, 0219 and 0415.

IEOM observers generally considered the established procedures for counting to be inadequate, cumbersome and slow. A number of observer teams attended counts that lasted for over ten hours. The absence of guidelines on the counting of ballot papers for multiple elections led to a non-uniform process.

C. TABULATION OF RESULTS
IEOM observers attended 39 TECs to observe the tabulation of results from polling stations. In four TECs, results were being tabulated manually rather than with the use of the available computers. Problems or irregularities were noted in six TECs, mostly where the tabulation process was considered to be taking place in a non-systematic manner and in cramped and chaotic conditions. IEOM observers were intimidated by election officials and other persons inside the premises of TEC 29.

D. PUBLICATION OF PRELIMINARY RESULTS
The Electoral Code requires TECs to aggregate and publish the preliminary results of the proportional and majoritarian elections within 18 hours of the close of voting (i.e. Monday, 26 May at 14:00 hours). TECs are specifically obliged to publish preliminary results broken down by polling station. This is an important legislative safeguard to ensure transparency in the tabulation process and allow opportunities to confirm or challenge the correctness of the recorded results of individual polling stations and the accuracy of the aggregation of results by TECs.

In violation of the Electoral Code and an explicit instruction from the CEC, most TECs did not publish tabulation of the preliminary results of either the proportional or majoritarian contests broken down by polling station, and instead provided only summarised preliminary results. Of 41 TECs visited by the EOM in the period following the publication deadline, 35 had not published the results in the required format.22 In advance of election day, all TEC Chairpersons confirmed to EOM observers that they were aware of their obligation and their TECs had the technical capacity to publish preliminary results in the required manner. None of the TEC Chairpersons who violated the law provided a credible reason for having to do so.

The CEC must publish preliminary results of the proportional election within 28 hours of the close of voting. In a positive attempt to implement a recommendation of previous OSCE/ODIHR and IEOM reports, and to improve the transparency of its tabulation process, the CEC took steps to publish the preliminary results of the proportional contest broken down by polling station. This included televised announcements and updated internet pages. However, the official preliminary results could only be published in a summarised format because of the failure by TECs to provide full information.

E. RECOUNTS
Article 62(10) of the Electoral Code allows proxies and PEC members to request for TECs to verify the results of a polling station by way of a recount. Requests must be submitted in writing before the deadline of 14:00 hours on the day after the election. In a decision related to the presidential election, the Constitutional Court confirmed that TECs were obliged to undertake a recount if a request was made; previously, a number of TECs had voted to refuse requests. However, IEOM observers confirmed that TEC 30 was not open to receive requests for recounts and TEC 29 refused to accept requests for recounts.

Recounts of the results from around 80 polling stations in 26 TECs took place. Over half of these were monitored by EOM observers. While most recounts confirmed the results contained in the PEC protocol, a number revealed major discrepancies from the PEC protocol, especially in TEC 12. In TECs 17 and 18, recounts in the proportional contest showed similar discrepancies between protocols and actual votes cast and many signatures on the voter lists raised serious concerns of fraud.
EOM observers also noted wider problems surrounding some recounts. One TEC Chairman resigned after allegedly receiving threats, while other TEC members and proxies claimed to have been assaulted. In TEC 16, an EOM observer was obstructed by TEC members. In TEC 56 proxies for a losing candidate withdrew requests for recounts, apparently after coming under pressure to do so. In TEC 12, serious irregularities were noted in the conduct of the recount.

COMPLAINTS TO THE CONSTITUTIONAL COURT AND FURTHER RE-RUN ELECTIONS

The Constitutional Court received 19 applications regarding the results of the majoritarian elections before the close of the seven-day deadline on 6 June. The Court joined two pairs of applications concerning the same TECs and declared two applications inadmissible, while one applicant withdrew the complaint, resulting in 14 formal cases proceeding to a hearing20. The two applications declared inadmissible related to challenges against the results of specific polling stations in TECs 4 and 9. The Constitutional Court considered that as such cases related to "voting results", rather than the "election results", they were outside of the Court's jurisdiction. One application regarding the proportional election results was submitted by the Justice Alliance bloc and declared admissible. An additional application regarding the results of the re-run election in TEC 12 was lodged and heard.

Article 102 of the Constitution requires the Constitutional Court to complete its decision on a case within 30 days. This meant that the Court operated under considerable time pressure to hear the 16 cases, potentially compromising its ability to fully investigate all issues and causing problems in the disclosure of materials and the enforcement of disclosure orders.

Of the total of 15 challenges to majoritarian results, two, relating to constituencies 16 and 50, were allowed, leading to the annulment of the official results and the holding of repeat elections on 29 June (TEC 50) and 20 July (TEC 16). The 13 other cases were refused. The challenge to the results of the proportional election was also refused.

The re-run elections ordered by the Constitutional Court were not observed by the EOM as it had closed. In TEC 50 (Artik), and in TEC 16 (Yerevan), the re-run elections produced the same winner as on 25 May.

ELECTION RESULTS

The final results of the proportional election were announced on 31 May. The majoritarian constituency results were announced a day earlier.

Irregularities observed by the IEOM during vote counting, and analysis of the published results by individual PECs, raised concerns that the announced turnout of 1,234,925 voters was inflated. Even so, this represented 52.71% and marked a significant drop in turnout from the 68.44% of voters who, according to official results, participated in the second round of the 2003 presidential election.

A. PROPORTIONAL LIST
In the national proportional list election, five parties and one electoral bloc passed the threshold of five percent of votes cast, to gain representation in the National Assembly21. The distribution of the 75 seats filled by the proportional system is as follows22:

Republican Party 23 seats
Justice Alliance (bloc) 14 seats
Orinats Yerkir 12 seats
Armenian Revolutionary
Federation (ARF) Dashnaktsutiun 11 seats
National Unity Party 9 seats
United Labour Party 6 seats

Five of these six parties/blocs won representation through the proportional system in the 1999 parliamentary elections. The newcomer was the United Labour Party, established only in September 2002. For the first time since Armenia's independence in 1991, the Communist Party failed to win representation.

The Armenian Democratic Liberal Union (HZhAM), with 4.66 percent in the proportional list election, was highest placed of the parties and blocs that failed to pass the 5 percent threshold to enter the National Assembly. The party, which missed the threshold by under 4,000 votes, complained to the EOM that it had been prevented from gaining a seat by a falsely inflated turnout, but chose not to pursue this in the Constitutional Court.

B. MAJORITARIAN SYSTEM
Of the 56 majoritarian constituencies, 37 were won by candidates nominated by citizens' initiatives rather than by parties.

C. FINAL COMPOSITION OF THE NATIONAL ASSEMBLY
Under the National Assembly's rules of procedure, the parties and blocs that won representation by the proportional list system constitute parliamentary "factions"; no other factions may be formed. Deputies elected by the majoritarian system may affiliate to one of the factions by the latter's consent. As there are six factions, election commissions will again consist of nine members.

The composition of the factions formed after the National Assembly, which met for the first time on 12 June, confirmed that many non-partisan candidates who won in majoritarian constituencies had a political party affiliation.

Composition of the National Assembly, by faction/ deputies' group

Republican Party faction 40 seats
Orinats Yerkir faction 19 seats
Justice Alliance faction 15 seats
Orinats Yerkir 12 seats
Armenian Revolutionary Federation
(ARF) Dashnaktsutiun faction 11 seats
National Unity Party faction 9 seats
United Labour Party faction 6 seats
"People's Deputy" group 17 seats
Unaffiliated 14 seats

Total 131 seats

Of the seven women elected, six were from the proportional list (two each from the ARF Dashnakstutiun and Orinats Yerkir, and one each from the Republican Party and the United Labour Party). One of the women elected on the Orinats Yerkir list was appointed as Minister of Culture and Youth. The other won in majoritarian constituency 32, where she was a non-partisan candidate who was included on the Republican Party proportional list and joined its faction in the new National Assembly.

RECOMMENDATIONS

The following recommendations are offered for consideration to the Armenian authorities:

A. GENERAL RECOMMENDATIONS

1. The recommendations contained in the OSCE/ODIHR final report on the 2003 presidential election in Armenia should be implemented. Key recommendations that have not yet been implemented are:
(a) holding to account persons responsible for violations of the Electoral and Criminal Codes;
(b) imposing sanctions on State-funded media for violating Electoral Code requirements to provide equal coverage of presidential candidates;
(c) repealing provisions of the Administrative Code concerning administrative detentions;
(d) ensuring that use of public resources during an election campaign is clearly defined in law.

2. Bodies appointing election commission members should hold administratively and politically accountable their representatives on TECs and PECs, who violated the law. Such persons should not be re-appointed to election commissions.

3. Every effort must be made to ensure full transparency during each phase of the election process. A zero-tolerance approach should be taken by election stakeholders against attempts to limit transparency through legislative or administrative measures.

B. LEGISLATIVE FRAMEWORK

4. The formula for appointing members to election commissions should be reviewed to ensure confidence of election stakeholders and professional and impartial performance of the election administration.

5. Article 40 of the Electoral Code, relating to complaints and appeals, should be revised to ensure an unambiguous, effective and transparent appeals process. It should include appropriate sanctions against election violations and remedies for all classes of aggrieved persons, and require commissions and courts to provide reasons for their decisions.

6. The Electoral Code and the Civil Procedure Code should be amended to allow appeals on election issues from the Court of First Instance to a higher court in an adequate timeframe.

7. The Electoral Code should be amended to:
(a) stipulate the method to be used for calculating the difference in size between constituencies;
(b) ensure consistency on the issue of permanent residency;
(c) allow blocs to nominate candidates for majoritarian contests;
(d) reduce the number of signatures required for registration of lists of candidates to a maximum of one percent of the registered voters in the republic-wide constituency;
(e) loosen the requirement to submit both petitions of signatures and financial deposits, to require only one of the two safeguards;
(f) include clear and reasonable compatibility conditions for simultaneous holding of public office and running for political office;
(g) subject candidates for majoritarian and party list seats to equal restrictions related to compatibility with holding public office;
(h) require all signatures provided in support of a candidate or party list to be checked, unless the number of correct signatures has reached the legal threshold;
(i) state explicitly the investigatory powers of the CEC and TECs in relation to candidate registration;
(j) improve the supervision of campaign funding and spending by election commissions;
(k) enable voting by hospitalized voters or those otherwise unable to attend their polling station on election day, with appropriate safeguards against election fraud;
(l) require ballot boxes to be transparent and to clarify that more than one ballot box can be used in a polling station;
(m) allow a violation to be entered into a PEC record upon the request of one PEC member or one proxy;
(n) require greater safeguards to prevent double-voting, including marking voters' fingers with invisible ink;
(o) provide full details regarding how votes "against all" are accounted for or to remove this option from the ballot;
(p) require the CEC to publish preliminary and final election results broken down by polling station and require TECs to publicise their preliminary results broken down by polling station at the same time as they are forwarded to the CEC;
(q) allow TEC and PEC members to be dismissed upon the decision of a superior election commission if found to be responsible for committing an electoral violation or permitting it to take place; and
(r) reflect the decisions of the Constitutional Court taken on election issues since October 2002.

C. ELECTION ADMINISTRATION

8. The Armenian authorities should undertake a thorough review of the system for compiling and maintaining voter registers. The role and powers of all bodies involved in the process should be clarified in detail. There should be a sustained and systematic effort to improve the quality and accuracy of the voter registers across the Republic, and create guarantees against potential multiple entries.

9. As required by the Electoral Code, the CEC and TECs should take formal decisions on all relevant complaints and appeals and make publicly available, in a timely manner, full details of all complaints received and decisions taken.

10. The CEC should:
(a) immediately undertake measures to hold accountable the Chairpersons of TECs which violated the Electoral Code and the CEC instruction by failing to publish preliminary results in the required format;
(b) ensure a systematic method of publishing and disseminating its decisions and instructions to TECs, PECs, political parties and the wider public. Binding instructions should be made in writing;
(c) repeal decisions which restrict the movement of proxies in polling stations and limit the number of requests for recounts;
(d) produce detailed guidelines on the criteria for the registration of candidates which should be publicly available ahead of the registration process;
(e) review procedures for counting to ensure full transparency and efficiency during the count; and
(f) seek to improve the efforts of TECs to ensure polling stations are accessible to disabled voters.

D. MEDIA

11. Private electronic media should meet basic requirements of objective and impartial reporting during an election campaign period.

12. The Law on Television and Radio Broadcasting should be amended in a manner which will ensure clearer criteria for the conduct of license tenders and a more independent composition of the National Council on Radio and Television.

E. PARTICIPATION OF WOMEN

13. Significant and sustainable steps should be taken to increase the participation of women in the electoral process and especially to improve the representation of women as candidates and in parliament.