Strasbourg, 24 May 2002                                                                           CDL-AD(2002)009


Opinion No. 182 / 2001                                                                                             Or. Engl.














on the basis of comments by

Mr Hjörtur TORFASON (Member, Iceland)

Mr Florian GROTZ (Expert, Germany)

Mr Richard ROSE (Expert, United Kingdom)









1.  In its Resolution 1257 (2001), the Parliamentary Assembly of the Council of Europe invited Georgia to submit for expertise the newly adopted Election Code to the European Commission for Democracy through Law (Venice Commission) in order to assess whether the current electoral legislation takes full account of recommendations made in 1999 by the Parliamentary Assembly Ad hoc Committee on the Observation of Elections and by the OSCE Office for Democratic Institutions and Human Rights (ODIHR)” (point 8.iv). This question was also mentioned in Recommendation 1533 (2001) (point 3.i).

2.  On 15 January 2002, the Georgian authorities submitted the unified electoral code of the Republic of Georgia, as adopted on 2 August 2001, to the expertise of the Venice Commission for its opinion.


3.  On 16 January 2002, the Committee of Ministers, in its reply to Recommendation 1533 (2001) (780th meeting, item 2.5b), considered that it is inter alia indispensable that the various laws or draft laws mentioned in the Secretariat’s report (in particular the Electoral Code and the Code of criminal procedure) be transmitted to the Council of Europe within the indicated time-table for expert appraisal” (see doc. 9324 of the Parliamentary Assembly, point 3).


4.  The present opinion of the Venice Commission, based on the reports by Messrs Richard Rose (CDL (2002) 10), Florian Grotz (CDL (2002) 15) and Hjörtur Torfason (CDL (2002) 53) on the unified electoral code, was drawn up by the Secretariat in conformity with the decision of the Commission at its 50th Plenary Session (Venice, 8-9 March 2002) and approved by the rapporteurs.


5.  The argumentation basically follows the structure of this law, focusing on three questions:


(a)    Which crucial provisions have been changed since the previous electoral legislation, and if such changes occurred, in which sense can the new stipulations be considered an enhancement of democratic standards?


(b)   Which items criticised by the Organisation for Security and Co-operation in Europe (OSCE) and the Council of Europe (CoE) have not been changed?


(c)    Which regulations are still missing in order to establish an adequate framework for holding truly free and fair elections in Georgia?


6.  Before turning to the detailed legal provisions, it is important to (re-)consider briefly the overall political context in which the new Code was drafted and is to be implemented:


·                     Since 1995, when its current constitution went into force, post-soviet Georgia has witnessed a series of overall competitive parliamentary and presidential elections within a politically stable environment. Nevertheless, due to unsolved ethnic-territorial conflicts, an enduring economic crisis and the structural weakness of opposition parties, the new democratic institutions are still far from being consolidated. This is confirmed, among others, by international observers noticing systematic shortcomings during the election processes, especially regarding the functioning of the electoral administration. To overcome such democratic deficits, both the OSCE and the CoE have suggested relevant legal amendments as adequate benchmarks for electoral reform.


·                     In late autumn 2001 – some weeks after the new Code had been drafted – a major political crisis occurred, as certain measures taken by the State Security against an independent television station led not only to public protests, but also caused the resignation of the parliamentary Speaker and, eventually, the dismissal of Government. Almost simultaneously, the Citizens’ Union of Georgia (CUG) – the predominant party in Parliament since 1995 – broke apart, after President Shevardnadze had resigned from its chairmanship. Both the governmental institutions and the party system are thus “in flux” again.


7.  The new Electoral Code, therefore, has to be evaluated against a “double background”: the reform suggestions of international organisations and the (most recent) changes within the political system.


General Provisions (Chapter I)


8.  Similarly to the Electoral Code of neighbouring Armenia (1999), the new electoral law of Georgia integrates the previous legal acts on presidential elections, parliamentary elections and elections for the organs of local self-government into one document. This new form – defining at first general conditions for all elections and then adding specific provisions for all relevant types of elections – has a great advantage: it basically enhances the transparency of the legal framework and, to a certain degree, enhances “democratic efficiency”, since it provides equal organisational standards for all election types. It results in a stronger emphasis on democratic equality and transparency, and one of the advantages gained is that the organisation of the system of election commissions can be provided for in a comprehensive and effective manner. The mere size of the Code, however, can make it difficult to find all details relevant for each type of election.


9.  The Chapter sets out some useful definitions and a declaration of basic principles, among which it is to be welcomed that universally recognised principles of human rights and standards of international law are expressly included as elements of the legal basis for the preparation and conduct of elections (Article 2).


10.  In accordance with this outline, Chapter I presents the “general provisions” in a very detailed way. Article 5 appropriately provides, consistently with Article 28 of the Constitution, that “elections in Georgia are universal” and that citizens of 18 years or over have the right to vote.  The Article is limited to a clear statement of this general principle, however, and leaves the problem of how to implement it in fact towards all citizens to the specific chapters of the Code.  While this approach is logical, it might be desirable to include in the Article a reference to certain major aspects of this problem, such as the matter of citizens who are residing or dwelling abroad at the time of election (see e.g. Articles 9.5.d and 10.4) and the matter that the voter group may not be entirely the same in national elections (presidential and parliamentary) and in local elections, which differ at least in that persons residing abroad are excluded in the latter (Article 110.3). Therefore, a relevant paragraph should be added under Article 5 defining which citizens (e.g. additional qualifications with regard to the duration of absence from the motherland) may participate in which types of elections (e.g. in presidential and parliamentary elections, but not in local elections).[1] Since Article 5 preferably should remain a general provision, it might be sufficient to mention there the principal or minimum effort that the State should make (through embassies and consulates and other absentee balloting arrangements) in order to enable Georgians abroad to exercise their voting right.


11.  Since the Constitution does not include Georgian residence as a condition of entitlement to active voting rights (Article 28), it is presumed that the question whether permanence of foreign residence should affect voting rights will depend on the laws concerning citizenship (under Articles 12 and 13 of the Constitution).


12.  The Code does not provide for any voting rights for residents who are non-citizens, but that presumably involves a constitutional question.


Registration of voters (Chapter II)


13.  The basic principle should be to have a single and permanent list of voters published well in advance of election day. This list should be generally accessible, updated at regular intervals (annually) and published on a preliminary basis upon each updating.  It should then be specifically updated and definitively published well in advance of any upcoming elections. More precisely, an initial list should be published giving names of  individuals compiled on the basis listed in the law. Then, individuals whose names are not on that list can request inclusion at an appropriate address under terms laid down in line with published regulations. A revised and final list can then be published and issued to the PECs for use on election day. Individuals whose names are not on the revised list for a given PEC should not be allowed to cast a ballot, for this would open the door to many types of abuse and electoral fraud. The list should be coordinated with the civil registration in the country and its compilation and maintenance should be charged to the central and municipal authorities responsible for the civil registers, in close cooperation with the Central and District Election Commissions.  There should be a clear opportunity for individuals to demand inclusion in the list according to residence and age (and also for interested persons to challenge the list) before election day.


14. The provisions of Chapter II appear to be drawn up with these principles in mind.  However, it does not seem clear that the main list of voters should be permanent and regularly updated, as the emphasis is on the compilation of the list by the District Election Commissions according to precincts and the transfer of data to the DECs at certain deadlines before election day (Article 9).  This may be reasonable according to present conditions in Georgia. If the laws on the civil registration system in the country do not include rules specifically relating to the use of civil register data for purposes of voting lists, an amendment of the Code in this respect should, however, be considered.


15.  Owing to the notorious indifference of voters in many countries to their registration until elections are imminent, the provisions of the Code relating to updating or correction at that time (Article 10 on a supplementary list of voters and Article 12 on a voting license to persons who have changed residence) are very necessary and of great importance. However, they could open the door to abuse and/or fraud, even if there is an improvement in relation to the previous situation.


16.  The establishment of supplementary voter lists (Article 10) can be considered a definite progress towards improving the administration of elections. In the Georgian context, this item is of especial importance due to the large number of Internally Displaced Persons (IDPs) from Abkhazia and other regions. In the 1999 parliamentary elections, several observers reported that IDPs could occasionally cast more votes than they would have been allowed to[2], because the administration of voter lists did not function properly[3]. Under the new regulation, IDPs can be identified more easily by the election authorities, so ‘double voting’ of IDPs should not be possible any more.


17.  For the same purpose, Article 12 provides precise regulations for including short-term changes of residency into the voter lists. Article 12 allows the voting list to be altered by license to cover changes made up to 6 p.m. on the day before election.  While it is desirable to have the possibility for correction until this time, it is questionable whether the correction ought to be made that late unless the actual change of residence has been completed earlier, e.g. more than 10 days before election (cf. Article 13) or further back in time.  However, if the provision is made more restrictive in this regard, a system of absentee balloting may be needed in order to allow persons who moved recently to vote at their prior residence.


18.  As regards persons who are not included at all on the voter list, Article 10.2 allows them to be introduced up to and including the election day itself.  While a liberal solution of this important matter has its positive side (and is known in some other countries), it also has a dangerous side, and it seems questionable to allow the Precinct Election Commission (as the clause seems to indicate) to make the introduction on its own.  The correction ought rather to be supported by the decision of a court to which the voter can have recourse, or at least the DEC.  In any case, the acceptability of the clause depends very much on the reliability of the Georgian citizen identity card and registration card which the voter is required to present.


19.  Article 11 and Article 52 et seq. The Soviet-era practice of mobile ballot box papers should be strictly limited because it is time-consuming and requires the presence of people who are expected to be in polling stations administering a ballot. Moreover, mobile votes remove the casting of the ballot from full public scrutiny and thus open the door to fraud.  Provision should be made for the issuance of absentee ballots, subject to proper scrutiny and procedures that can be modelled on those of a variety of EU states.


20.  In any case, mobile box and absentee ballot votes should be counted and published separately from the votes cast directly at the fixed polling station in the respective precincts.


21.  If a military compound has a substantial number of electors, then it could be the subject of a separate Precinct Electoral Commission (PEC). Remote areas might even be allowed to vote the day before the official polling day at a PEC established there. Most EU countries make provision for absentee voting, and a review of these provisions can develop criteria appropriate to Georgian circumstances. Initially, a more restrictive approach to issuing absentee ballots should be used, in order to make sure that the procedures are carried out in conformity with the law. Total votes should be published separately at the precinct level for absentee or mobile votes.


Election Districts and Election Precincts (Chapter III)


22.  According to Articles 49 and 4 of the Georgian Constitution, the Parliament of Georgia consists of 235 members, of which 150 members are elected by a proportional system on a nationwide basis and 85 members are elected by a plurality system.  According to Article 15.1 of the Code, these 85 are to be elected in 85 single-member constituencies or election districts, 10 of which are in the city of Tbilisi and 75 of which are established in accordance with the administrative-territorial division of the country. Regarding the delimitation of electoral districts, it is quite astonishing that Article 15 does not include any remark on the legally allowed deviation from the average ratio of registered voters per single-member constituency. This lack is not only unusual by international standards of electoral legislation. What is more, the violation of electoral equality in this sense was one of the main problems in the 1999 parliamentary elections: the average ratio of registered voters per single-member constituency varied a lot, from approx. 3,600 voters in the Lent’ekhi district or approx. 4,200 in the Kazbegi district to over 138,000 in Kutaisi City[4]. Therefore, the OSCE suggested a maximum deviation of 10% from the average ratio of voters/inhabitants or adult citizens per district (internationally such a standard is quite common)[5]. Such provision should be added to Article 15; moreover, the number of seats of the city of Tbilisi should depend on its population/number of voters or adult citizens and not be determined in the law. Since for organisational reasons it is quite sensible to retain the administrative division as a general basis for defining electoral districts, the maximum margin might be increased to 15%–20% for less accessible regions. Any bigger deviation from the average size is not acceptable from a democratic point of view, except when a very under-populated administrative district has to be represented in Parliament. Such an exception would be admissible since more than 60 % of the parliamentarians are elected on a nation-wide basis.


23.  Furthermore, electoral districts are established no later than 58 days before elections (Article 15.6). This is not advisable. The time for establishing electoral districts should not depend on the date of elections; otherwise, the door would be open to political manipulation. The reapportionment should rather take place after each regular census, that is in principle every 10 years, in order to take into account the changes in the population.


24.  The legally prescribed average size of the electoral precincts (2,000 registered voters; cf. Article 16.2), is (still) relatively large. Given the high rates of electoral participation and the organisational problems observed in previous national elections, smaller precincts (with about 1000 voters) would surely contribute to pursuing a better “formalisation” of the electoral processes.


25.  Finally, the rather inconspicuous provisions of Article 16.6 on the establishment of electoral precincts outside the national territory deserve certain attention. At first glance, this measure is a purely administrative act, and the relevant competence seems to belong rightfully to the Central Election Commission (CEC). However, the ‘counting’ of external votes is a political issue and may be – depending on the electoral system and the political distribution of external votes – severely disputed among political parties. Therefore, it would be sensible to regulate the establishment of external precincts more precisely by law (as already mentioned under Article 4), especially since citizens staying abroad are allowed to vote for candi­dates in single-member constituencies in parliamentary elections (Article 16.6).


Election Administration (Chapter IV)


26.  This chapter includes the most important changes of the new Code: the reform of the system of Election Commissions (EC). It contains extensive provisions on the organisation and functions of the commissions in charge of elections, appropriately operating at three levels, namely a Central Election Commission (CEC), District Election Commissions (DECs) and Precinct Election Commissions (PECs).  The provisions appear to constitute a successful response to the recommendations made in the reference materials and represent a significant reform enhancing the capacity for professionality and neutrality which may be expected from these essential institutions.


27.  The requirement of approval by at least 2/3 of the members of Parliament is understood as a required majority and not a quorum.  This requirement is desirable in order to ensure a multi-party influence on the composition of the CEC, while it poses the technical problem of achieving this majority.  If that problem is difficult to solve, the method of choosing the members by a proportional parliamentary vote may perhaps be considered.


28.  Though the multi-level structure of the EC remains a “centralised system of election administration” (Article 17.2), its composition is not as “state-centred” as before. Whereas the Central Election Commission (CEC) used to be completely chosen by the main state organs (President, Parliament, and regional assemblies), the Election Code has introduced a “bottom-up” nomination system modelled on the Mexican CEC (which is undoubtedly the most professional election authority throughout Latin America). According to this modus, the seven CEC-members are elected by Parliament out of a list of 14 candidates exclusively nominated by non-governmental organisations engaging in electoral observation (Article 27). Unlike previously, the Chairperson of the CEC is not elected by the President, but by the CEC among its members in a highly consensual procedure (Article 28). The organs of election authorities at lower levels (District and Precinct Election Commissions) are to be chosen in a “semi-centralised” manner, i.e. partly by the higher EC and partly by the relevant strongest parties at district and precinct levels.


29.  In sum, the modus of (s)electing EC-members according to the 2001 election code enhances the de-politicisation and, thereby, the professionalism of the key actors of the electoral administration.


30.  The possibility for political parties or blocs to withdraw (Article 21.2.i) their members is, however, completely contrary to the principle of de-politicisation and should be removed. The possibility of dismissal for violation of election legislation (Article 21.2.h) should be more precise, in order to avoid any abuse and to respect the principle of proportionality; the law should also make only one body (higher level commission or Court) responsible for such a sanction.


31.  The Commission has been informed that the composition of the CEC was once again modified by Parliament on 10 April 2002. This reform would not be applied to the next local elections to be held in June, and cannot therefore be considered as a last-minute modification of the electoral legislation. However, it must be underlined that frequent changes of electoral law, and in particular of its most sensitive features like the composition of the election commissions, will often seem to be dictated by immediate political interest and may cast doubt on the legitimacy of the democratic process itself. It is therefore advisable to adopt rules on this matter for the long run.


32.  In itself, the newly adopted composition of the CEC – according to our information, 2 members nominated by parties/coalitions/factions which passed the 4 % threshold in the 1999 parliamentary elections, 2 members nominated by the autonomous republics and 1 nominee of the President of Georgia – is not contrary to the standards of the European electoral heritage. It seems that it would actually imply a re-politicisation of this body, by eliminating the role of the NGOs. More detailed comments could be made as soon as the new text is available to the Venice Commission.


33.  Article 32 is understood as meaning that the DECs of 7 or more members are organised so that all parties who have been successful (overcome the threshold) in the nationwide elections are ensured of appointing a member – i.e., in addition to 3 appointed by the CEC (from candidacies as per sections 4 and 5) and 1 by the local governmental body.  This is highly acceptable, except that it perhaps does not cover the contingency of the successful candidate from a single-member (plurality) constituency from the last election not belonging to one of these parties.  It is assumed that the right for this member to propose candidates to the CEC is not supposed to offset this and is secondary to the right of the NGOs (i.e. will not be exercisable if they propose a sufficient number of candidates).


34.  As to Article 32, the system of appointment of the 11 members of the PECs is also highly acceptable, except that it also perhaps does not take full account of the above contingency. In Article 32.2, the expression “the relevant representative body of local self-governance” is not very clear, but this is probably a problem of translation.


35.  The following principles should be stipulated in law to apply at all three levels – the Central Commission, the District Commissions and the Precinct Commissions:


(a)  The chief officials, that is, the Chair, Deputy Chair and Secretary, should be chosen by a proportional representation system, for example, the single transferable vote. A simple majority vote is open to one-party domination and in the DECs and PECs, even a two-thirds vote is subject to domination in a similar way. This will require amendment of Article 22.2 and other similar clauses, and also of Article 27.6.


(b)  Major actions of a Commission should require the signature of the Deputy Chair as well as the Chair. As written, the law concentrates too much power in the hands of a single individual.


(c)   Where it would be inefficient to require two signatures before action is taken, explicit provision should be made for the deputy commissioner to file a written dissent if excluded from prior consultation and endorsement.


(d)  All members of the Commission should have the right to be given documents about all actions authorised by the Chair in the name of the Commission, or actions taken by the Chair and countersigned by the Deputy Chair.


(e)   Where a decision is taken by the Commission collectively in the presence of Commissioners, then any member should have the right to append a written dissent, as stipulated, for example, in Article 19.2.


36.  Article 32.5 The provision for nominating members within one week is too short; one month is much more reasonable.


37.  Article 34.2 There must be a translation problem. The expression “election district” is probably a bad translation, since it is obvious that a district election commission cannot establish election districts. Or does the text mean that the district election commission should verify whether the district boundaries are in accordance with the law?


38.  Article 36.2 PEC composition. The reference to 5 parties with the best results at the last parliamentary election is ambiguous. Does it mean the best results nationally? If so, this would exclude regional or local parties and situations in which the small size of the PEC (2000 voters) makes a party important locally that is not important nationally. Alternatively, if it refers to the five best-placed parties in the precinct, it could exclude one or more nationally prominent parties. The potential problem thus raised could be dealt with by rephrasing the law to allow for additional members to be nominated in those DECs or PECs where one of the five best-placed parties in terms of votes within that district or precinct was not one of the five best-placed nationally.


Registration of Election Subjects and Lists of Supporters (Chapter V)


39.  The provisions of this chapter (Articles 40-42) are quite acceptable by recognised standards.  In 41.5, it would be proper to require a statement of the birth date, identity number and address of the candidates in addition to their name.  As regards the information to be provided on the candidates and the supporters, it would be desirable to require the election commissions to give the persons responsible for the lists a limited opportunity to remedy defects in relation to the signatures (i.e. to correct obvious errors or omissions, but not to collect substitute signatures) without resorting to an appeal.


40.  An innovation in this chapter is the precisely defined procedure of checking the authenticity of supporting signatures (Article 42.2). The verification of a mere percentage of the signatures opens however the door to manipulations, since a big number of falsified signatures may be introduced in the sample. At any rate, every list obtaining a sufficient number of valid signatures, whatever the number of invalid ones, should be registered, which implies the verification of as many signatures as necessary to reach this number.


41.  The periods for collecting signatures and the deadlines for submitting the list of signatures should be defined in the law.


42.  It also might be provided that a supporter cannot withdraw his support after the list has been submitted to the election commission.


 Election Funding (Chapter VI)


43.  The provisions of this Chapter (Articles 43-45) on election funding and Articles 46-48 on campaign funding are acceptable by recognised standards, and the latter seem to represent significant reform. A major contribution to enhancing financial transparency is the instalment of Election Campaign Funds (Article 46ff.).


44.  Article 46.7 in fine – “free of charge (at market prices)” – This appears to be contradictory, probably due to a problem of translation.


45.  Article 47.5 The prohibition of funding by foreigners or stateless persons residing in Georgia could be reconsidered.


46.  Article 48 should be interpreted in conformity with the principle of proportionality.


Polling (Chapter VII)


47.  The regulations of this chapter provide a functional basis for securing a democratic election process. Important items include:


48.  However, a certain number of points should be revised. In various articles in this chapter, e.g. Article 50.4, reference is made to the withdrawal of candidates. As a general principle, withdrawal of candidates should not be allowed, because it opens the door to manipulation by candidates and parties, and corruption. To avoid manipulative candidate registration followed by attempts to withdraw, or public proclamations by a registered candidate that s/he would like supporters to vote for another candidate or party, the Electoral Commission should  have the power to impose a penalty, which could take one of two forms or both:


(a)   Impounding any unspent funds in the candidate's election account at the bank AND a fine equal to the amount of money spent from the fund.


(b)   Disqualification from standing as a candidate at a subsequent election, or only being allowed to stand if meeting more onerous standards, e.g. double the number of nominating signatures, a cash deposit, etc.


49.  The use of numbered ballots (as provided in Article 51.5) normally is to be discouraged, as it may weaken the secrecy of the voting process.  It may be desirable under present circumstances in Georgia, but their abolition should be a future aim.


50.  It may also be asked whether it is necessary to have the actual ballots signed by PEC members (even though this is to occur in the presence of the voter and under possibility of observation), as provided in Article 54.2.  If signature is needed, perhaps it might be sufficient to apply it to the envelopes.


51.  In the interest of secrecy, it should be forbidden to a voter to mark his ballot in any way other than as necessary to express his elective will (in order that the ballot is not identifiable after opening). Ballots found to be marked so as to indicate an intent by the voter of showing that it is his/her ballot should be discounted as invalid. This does not seem to be covered in this chapter (except perhaps at the end of 54.2.a).


52.  Article 58.4  Ballots from mobile polling boxes should definitely not be mixed with ballots cast in person.


53.  Article 62.2  The reference to "days" should be amended to refer to "working days".


54.  Article 62.3  This sentence is badly translated in English and its meaning is not clear. The clause should be scrutinised by someone familiar with the language or in receipt of a better translation.


Transparency During Preparation and Conduct of Elections (Chapter VIII)


55.  The provisions of this Chapter (Articles 65-72 on the former and 73-76 on the latter subject) represent a progressive effort to ensure free and fair elections.


56.  Article 69.9 Observers should have their passport number or Georgian ID number on their license.


57.  Article 70.1.j Observers should not have the right to appeal. This is perhaps a problem of translation.


58.  Article 72 The restrictions on activities of observers in Article 70.2 should also apply to media. There should also be rules to prevent a polling station being "packed" by media representatives, especially from a single paper/broadcasting outlet. Therefore, 72.5 should be amended to limit any media to a maximum of 2 persons present at the same time. Moreover, the PEC should be allowed to establish, prior to the opening of the poll, a maximum number of media (with a minimum of 4 or 5); if more media seek attendance, those allowed to be present should be determined by lot.


59.  Besides the structural reform of the Electoral Commissions, Chapter VIII is the second major change in the new Code. In international comparison, a separate paragraph on transparency is not very common in electoral laws. It can be considered progress to summarise preconditions for free and fair elections in “fragile” democracies. Such items include:


60.  Article 73.11 The clause about opinion polls is clear, practical and fair and should not be altered.


61.  Article 75.2 The limits to freedom of expression should be interpreted in conformity with the principle of proportionality.


62.  Article 76.4 There must be a lapsus calami in the translation (“a registered candidate who is not an employee of bodies of State authority…” instead of “who is…”). This provision should also be interpreted in conformity with the principle of proportionality.


Adjudication of Disputes (Chapter IX)


63.  This Chapter (Article 77) sets out very precise rules for the resolution of electoral disputes, stating the type of dispute and the available recourse to a higher election commission and a court of law (or the Constitutional Court), with specification of time limits.  These rules represent a reasonable response to recommendations in the reference materials.  In provisions of this kind, it should be made clear as far as possible to which court the recourse should lie, and this is generally the case.  However, it seems that there may be a choice between appeals to an election commission or to a court (Article 77.1). This is confusing and could lead to positive or negative conflicts of competence. It is then strongly suggested to abolish such a possibility of choice.


64.  The time limits given are very short, but in the case of appeals between election commissions, this is in principle appropriate, since the matters at issue tend to be ones of urgency. However, one may doubt whether the relevant provisions can be implemented as strictly as foreseen by law. For example, it is not sure that the Constitutional Court will always be able to decide upon election appeals within only five days, as requested by Article 77.4. For fully informed judgments on complex cases of (alleged) election fraud, this time span may prove to be too narrow. In such cases, the legitimacy of the constitutional review – and, eventually, of the Electoral Code – might be challenged.


65.  Furthermore, the time limits given should as far as possible be related to the time when the relevant decision, ordinance or measure becomes known to the interested persons through publication or direct notification, which may not always be the case.  The situation in this regard might perhaps be clarified by further provisions within this Chapter, but also by provisions within the Chapter on the election commissions to ensure the proper notification of their decisions (upon being made) where this is not already covered.


66.  Article 77.21 The cancelling of the registration of an electoral subject must take place only in exceptional circumstances, and in conformity with the principle of proportionality. At any rate, it is understood that a rayon (city) court cannot take such a serious decision about a list registered for national elections, and that an appeal to a higher court is open against it.


Elections of President (Chapter X–XI)


67.  The basic provisions for presidential elections have not been modified. The relevant stipulations on the regular term of office, the possibility of re-election, the candidacy and the electoral system are generally in accordance with international standards of direct elections for President.


Chapter X


68.  Article 84.4 Withdrawal of candidate for presidency at any time prior to polling day. This position is definitely not acceptable. This can only cause manipulation, confusion, speculation and suspicion. The penalty for a candidate endorsing voters casting their vote for another candidate should be disqualification from registration for a period of five years, that is, the next presidential election and all other elections for public office.


69.  One detail, however, is quite problematic from a normative point of view: Article 86.2 prescribes that the majority requirement (50%+1) is to be based on the number of votes cast and not, as commonly, on the valid votes. In other words, in this system invalid votes systematically count against the strongest candidate. In non-competitive contexts this does not pose a problem since in such “elections without choice” invalid votes may reasonably be considered a kind of “negative vote”. In multi-candidate races, on the contrary, invalid votes cannot be interpreted as clear political choice, thus they should not be taken into account. Therefore, the relevant formulations should be altered from “votes cast” to “valid votes” (see also “Elections for Parliament”).


70.  It would also be suitable, in order to avoid an indefinite repetition of the elections, to provide for the case in which the necessary participation threshold (Article 86.1) is not reached for several ballots (three times). The threshold could for example be reduced to one quarter of the electorate in that case.


Chapter XI


71.  Article 87 There should be allowance for a candidate who has finished first or second in the initial ballot to waive his/her right to contest the second round in favour of another candidate whose name is on the ballot and has finished third or fourth. This would deal with the contingency of a first or second place candidate having a low ceiling of support, and a third or a fourth place candidate having greater potential to attract a majority coalition in the second round. In no instance, however, should anyone be allowed on the second ballot who was not nominated on the first ballot and who did not poll a reasonable number of votes.


72.  Article 89 A minimum of 45 days would be appropriate for holding an extraordinary election. But this leads to a very tight timetable; in Russia, 90 days is allowed. A reasonable option would be 60 days. The law could be rewritten to provide for an election in 60 days.


73.  Article 89.2 The immediate meeting of Parliament is desirable. But Article 89.3 could lead to rancorous conflict if there is no agreement about the date of the election. It is better to fix it by statute than to make the choice of date a matter of debate – and this is true whether the election is in 45 or 60 days.


Elections for Parliament (Chapter XII–XIV)


74.  The regulations of parliamentary elections have not substantially changed either. Some provisions, however, were modified.


75.  A mere technical, but not unimportant adjustment is the stipulation in Article 96.2 that “double candidatures” (in single-member constituencies and on party lists) have to be indicated besides the relevant personal names on the party lists (according to Article 37.2 of the 1999 Election Law a list of the candidates in single-member constituencies had to be attached to each party list). The new procedure should make it easier for the voters to be fully informed in this respect, thus making the “parallel” electoral system more transparent.


76.  Article 96.1 It is assumed that the notion of “party” is defined in another piece of legislation.


Registration of election subjects taking part in elections for the Parliament of Georgia (Chapter XIII)


77.  Article 95.3 The recommendation to readdress the question whether non-parliamentary parties should need to present 50,000 supporters in the nationwide elections seems not to have occasioned a change in the law.


78.  Article 96.5,6 Provision should be made to give a person whose name has been included in two lists a choice between these two lists, in order for his/her candidature to be valid. In other words, a candidate should not be disqualified by being nominated with his/her permission on one list and then falsely nominated on another. If two parties use the same name without permission, then s/he should be allowed to withdraw.  This concern is partly but not wholly met by Article 98.3.


79.  Article 100 Cancellation of nominations for individual Members of Parliament. Here again, the clause should be withdrawn as it opens the door to many types of abuse.  In particular, the provision for possible withdrawal of candidates or parties until 2 days before polling is too liberal (Article 100.3). The first sentence of Article 100.2 is not clear, probably due to translation.


80.  A more important innovation is the stipulation of Article 102 that, unless the CEC “issues consent” upon relevant notices from the Prosecutor’s office, the immunity of parliamentary candidates must not be lifted before the electoral results have officially been published. Especially in view of the negative experiences during the 1999 elections, the legislator was right to delimit the power of the CEC in this respect. Since the formulation (“issues consent”) seems to be still too vague (at least in the English translation), it might be put even more precisely, being replaced by “unanimously decides”.


81.  The high level of the legal threshold (7%) (Article 105.7) has not been lowered despite being sharply criticised by international organisations. It goes without saying that setting a threshold of exclusion is always a political decision; therefore, legal thresholds in proportional electoral systems vary quite a lot, from 0.67% to 10% of the national vote. Within this empirical spectrum, however, Georgia has one of the highest legal hurdles world-wide.[6] Generally, it can be stated that the “mechanical” concentration effect of such a high threshold will hardly remain within the acceptable limits of “proportional representation”. In other words: It will exclude a considerable number of parties/valid votes from Parliament in favour of the strongest political forces; consequently, it tends to produce a rather majoritarian effect. This is basically confirmed by the 1999 parliamentary elections, where all in all 283,279 valid votes (14.1%) were ‘lost’ (see Table 1). Given the recent fragmentation of the Georgian party system following the break-up of the predominant CUP, the “exclusion effect” of the threshold will surely be reinforced during the next elections; it might even come to a result similar to the Russian Duma elections of 1995, when almost 50% of the valid votes were ‘filtered’ by a 5%-threshold and, due to this effect, the bigger parties could double (!) their seats (in relation to a pure proportional distribution of votes). And if the 1999 Russian Duma election had had a 7 percent threshold it would again have been very disproportional. A maximum of 5 percent is the most that can be justified. Moreover, it could be envisaged to increase the threshold by steps, e.g. 4 percent at the next election and 5 percent at the election after that (without changing the law, which would determine from the day of its adoption the date for the increase of the threshold). In sum, the 7%-threshold is definitely too high not only in normative terms, but also with regard to the actual political context. Therefore, it would be highly recommendable to lower it to 4%-5%. This would imply a revision of Article 50.2 of the Constitution.

Table 1: The 1999 Parliamentary Elections in Georgia








Total number



Registered voters



Votes cast




Invalid votes




Valid votes



















































Source: Kuchinka-Lančava, Natalie, Grotz, Florian, “Georgia”, in Nohlen, Dieter/Grotz, Florian/Hartmann, Christof (eds.) : Elections in Asia and the Pacific. A Data Handbook. Vol. I : The Middle East, Central Asia, and South Asia. Oxford : Oxford University Press, 2001.

a  The relevant figures refer to the “second votes” cast for the party lists in the nationwide constituency.

c Since the number of invalid votes is not given explicitly in the CEC protocol, this figure was calculated by the authors.

d Others include a total of 22 parties: B-RT-FG: 5,657 (0.3%); B-PF-CS: 4,339 (0.2%); B-VG-GC: 4,275 (0.2%); B-C-S: 3,778 (0.2%); B-RCPP: 3,229 (0.2%); CDUG: 2,951 (0.1%); PESDPG: 2,171 (0.1%); PDP: 1,917 (0.1%); B-XXIC-GN: 1,058 (0.1%); B-UNM: 994 (0.0%); FPG: 828 (0.0%); DAP: 758 (0.0%); B-GNUP: 733 (0.0%); PUC-LUG: 643 (0.0%); NPG: 593 (0.0%); UGN: 555 (0.0%); NIDPG: 529 (0.0%); DCG:452 (0.0%); PM-FG: 419 (0.0%); PU-S: 412 (0.0%); ILG: 344 (0.0%); PUC-AGFU: 333 (0.0%).


82.  In order to enhance the re-institutionalisation of the fluid party system, two further changes would be sensible:


(1)   A “differentiated threshold” should be introduced, i.e. a separate one for parties (e.g. 4%) and higher ones for electoral alliances (e.g. 6% for two-party alliances, 8% for coalitions of three and more parties). “Invented” in the transition processes of Central and Eastern Europe in the early 1990s,[7] differentiated thresholds had all in all positive effects on the consolidation of competitive party systems since they provided not only an incentive to build electoral coalitions, but also stimulated fusion processes among mini-parties (with similar programmes) and thus contributed to increasing intra-fractional cohesion within Parliament.


(2)   In order to guarantee a pluralist representation in the proportional part of the electoral system, the special provisions in case no party passes the legal threshold should be re-designed in a less strict manner. In such extraordinary case, holding repeat elections within a smaller sample of parties (with at least 2% of the original vote), as Article 105.17 states, is a viable regulation. However, if one party passes the threshold, no repeat elections are foreseen by law. Since the proportional part of an electoral system is not intended to produce a one-party system in Parliament, the relevant qualification on repeat elections in Article 105.16 ought to be altered from “none of the parties” into “less than two parties”. Additionally, the legislator might consider lowering the legal threshold for repeat elections, like the Polish Electoral Law of 1993 did.[8]


83.  Finally, the calculation basis of the threshold requirements should be modified as well. Like in presidential elections, the distribution of both the majoritarian and the proportional seats of the parliamentary electoral system is still based on the votes cast. As already explained above, the calculation procedure should be adapted to internationally common standards, i.e. the valid votes ought to be the calculation basis.


84.  It is difficult to understand in what cases repeating polling has to be ordered according to Article 105.13, 15, probably due to translation.


85.  Article 106.1 A second round election in single member districts with majority (that is >50%) guaranteed for the winner is acceptable. But there are both theoretical and practical reasons to allow more flexibility in deciding which two candidates go into the second round, as has been the case in France. Specifically, the lst and 2nd candidates in a district should be allowed to stand themselves or either could, if they desired, stand aside and nominate another candidate who was on the ballot and received a reasonable share of the vote. Whether this clause is often invoked depends on the situation.


86.  Article 106.9 Apparently, when a deputy elected in a single-member constituency withdraws, a person nominated by his party replaces him/her. It would be preferable to vote at the same time for a deputy and a substitute.


87.  The question of determination of the borders of single-member constituencies has already been commented upon. The establishment of an independent Boundary Commission, with provision for public hearings, and/or the CEC undertaking this task subject to special provisions for scrutiny and appeal, would be advisable. It would also be suitable to spell out the authority for reapportioning in the Constitution.


Elections of Local Representative Bodies and of Mayors (Chapter XV–XVII)


88.  The legal provisions for local elections differ from those at national level in several respects. For European democracies, this difference is a common feature, because the relationship between voters and representatives at local level is generally regarded as closer than at national level. The relevant stipulations thus follow an international trend to design specific legal provisions for the local context. These include


89.  In general, the provisions of those Chapters are consistent with recognised standards.


Local government (Chapter XV)


90.  Article 109 Multi-mandate districts elected by plurality votes (majority is an incorrect translation) can cause electoral confusion and encourage many abuses, as Japanese politics has shown. They can also produce very disproportional results.  On the other hand, in local government party lists may not be appropriate. Therefore, the Irish STV (Single Transferable Vote) form of proportional representation, in which voters state their preferences for individuals in rank order 1,2,3,... could be recommended.  This allows individuals to stand as independents. It also introduces a significant degree of proportionality. The introduction of such a change should however be preferably considered after a detailed analysis of the results of the previous elections.


91.  (And Article 115) In smaller towns and villages, a single multi-member constituency with up to 7 members would be appropriate. In large towns and cities where the council has, say, 10 or 12 members, then consideration should be given to having two districts to avoid voters having to rank up to a dozen candidates. At any rate, when the election is uninominal, there should be only one election district (cf. Article 115.2).


92.  Article 119.7 Does this provision deal with incompatibilities, or with ineligibility for election? This should be made clearer, in particular in view of the fact that the various elections are not held simultaneously.


93.  Article 120.7 It would be preferable, in conformity with the principle of proportionality, to give the party or the candidate a short deadline for submitting correct documents.


94.  Article 121 Withdrawal of candidacy. This seems even more difficult to justify in a local election context. This provision should be deleted.


95.  It could also be appropriate to consider developing absentee balloting facilities in order to enable persons temporarily outside the country to vote in their community, at least if this is not too difficult to implement.


Transitional provisions (Chapter XVIII)


96.  Article 127 The desire to maintain representation for Abkhazia is understandable, but it should be subject to a time limit. Representatives elected a decade ago have an uncertain mandate to represent the current population and their opinions - and this will be increasingly true as time passes. In a worst case analysis, the representatives could be there for life - and have a negative influence on claims to legitimacy and democracy of Parliament, as happened in Taiwan in the days of the KMT, when the Parliament was inherited from Peking.  Allowing historic representatives to remain in Parliament for one more term would be reasonable. If necessary, a compromise of two terms could be accepted - as long as their terms were not allowed to be indefinite.


Summary: Preliminary Evaluation of the Unified Election Code of Georgia


97.  In conclusion, the Unified Election Code of Georgia can be considered an important step forward in the process of securing democratic standards for representative government in the country under difficult conditions. A number of recommendations made by the international community were taken into account in the Code. In comparison with the preceding legislation, the most important innovations include:


98.  Notwithstanding this overall positive picture, some provisions remain highly problematic and should be altered before the next election. The most important points are the following.



99.  Furthermore, the decision of the Parliament to change once again the composition of the CEC is a negative signal. The stability of the most sensitive features of electoral law, including the electoral system and the composition of the election Commissions, is essential to the legitimacy of the democratic process (Chapter IV).

[1] For an extensive discussion of relevant institutional choices and their comparative evaluation cf. Nohlen, Dieter and Grotz, Florian, “External voting: Legal Framework and Overview of Electoral Legislation” in Boletín Mexicano de Derecho Comparado, Vol. 33, No. 99, pp. 1115-1145.

[2] In the ‘two-ballot system’ of Georgia, IDPs have only been allowed to vote for national party lists, but not for candidates in single-member constituencies. The problem of electoral equality connected with this stipulation and criticised by the OSCE (see fn below, at p.15) is still unsolved.

[3] Office for Democratic Institutions and Human Rights, Georgia. Parliamentary Elections. 31 October and 14 November 1999. Final Report, 7 February 2000. Warsaw: ODIHR, 2000, p. 22.

[4] Kuchinka-Lančava, Natalie and Grotz, Florian, “Georgia”, in Nohlen, Dieter/Grotz, Florian/ Hartmann, Christof (eds.): Elections in Asia and the Pacific. A Data Handbook. Vol. I: The Middle East, Central Asia, and South Asia. Oxford: Oxford University Press, 2001, pp. 371–406, at p.380.

[5] Office for Democratic Institutions and Human Rights, Georgia. Parliamentary Elections. 31 October and 14 November 1999. Final Report, 7 February 2000. Warsaw : ODIHR, 2000, p. 28.

[6] Currently, the only country with a higher legal threshold world-wide is Turkey (10%). Even Azerbaijan, which had an 8% hurdle, lowered it to 6% before the 2000 parliamentary polls. For an international overview of the relevant provisions cf. Nohlen, Dieter/Grotz, Florian/Hartmann, Christof, “Elections and Electoral Systems in Asia. The Middle East, Central Asia and South Asia”, in : idem (eds.) : Elections in Asia and the Pacific. A data Handbook. Vol. I : The Middle East, Central Asia, and South Asia. Oxford : Oxford University Press, 2001, pp. 1-46.

[7] Differentiated thresholds were introduced in the Czech Republic (1992–), Hungary (1994–), Poland (1993–), Romania (1992–), and Slovakia (1992–1998).

[8] According to Articles 6 and 7 of the 1993 Parliamentary Electoral Law of Poland, the legal threshold in repeat elections was to be lowered from 5% to 3% (for parties) and from 8% to 5% respectively.

[9] This means that every voter has as many votes as seats are to be filled within a multi-member constituency. Candidates with the highest numbers of votes are elected. In comparison with party-list electoral system, the “personal factor” tends to be more important in this less (pre-)structured form of candidacy.