|
Strasbourg, 22 February 2002
|
Restricted
CDL (2002) 16 rev
Or. eng.
|
|
|
|
EUROPEAN COMMISSION
FOR DEMOCRACY THROUGH LAW
(VENICE COMMISSION)
COMMENTS ON THE law
on the election OF MEMBERS
of the representative bodies
of local and regional self-government units
of croatia
by:
Mr Christos
HARTMANN
(Expert,
Germany)
1.
Introduction
1. The
importance of sub-national elections for the protection of minorities is evident.
In a democratic system many policies that are relevant to minorities are devolved
to sub-national levels; in countries that have emerged from violent civil wars
(like Croatia) the sub-national level presents an arena for the reintegration
of former enemies into the political process, their access to resources, and
their experimentation with different and democratic modes of decision-making.
2. Institutional
strategies that aim to protect minorities rarely rely on the electoral system
alone, but are complex packages with executive, legislative and judicial
dimensions that are often closely interrelated. The Croatian example clearly
demonstrates that the regulation of citizenship, to give an example, is central
to the working of the electoral system. The conscious engineering of electoral
provisions should keep in mind these limitations and side-effects.
3. The comment
of the Croatian Law on the election of
members of the representative bodies of local and regional self-government
units deliberately concentrates on the question of electoral rights and the
electoral systems, as these two elements are most relevant in the context of
minority representation. Electoral rights define who may participate in
elections (2), electoral systems are methods whereby votes are translated into
parliamentary seats or into governmental offices (3). The analytic perspective
draws from political science, especially from the concepts and insights of
international comparative research on electoral politics.
2. The Right to Vote
2.1. General Remarks
4. In Europe
the principles of universal, equal, direct and secret suffrage are generally
accepted and applied. Any deviation from these international standards does not
longer occur, neither at national nor at local elections. The detailed
provisions for suffrage are generally determined by three factors: voting age,
residence, and nationality.
5. Whereas voting age stands at 18 years
in nearly all European countries, there is less homogeneity with regard to the
traditional provision that registered voters must have a permanent residency
within the relevant electoral constituency. Since the past few years there has
been a worldwide tendency to expand the democratic rights of the people by
weakening this residency requirement and permitting citizens living abroad to
participate in national elections from their foreign place of residence
(so-called external voting). Such external voting has been applied for
elections to national offices, but also criticised due to problems of
transparency (organizational aspects), of judicial review (legal aspect), and
questions of political representation. From this perspective the right to vote
is not simply an individual right, but it contributes to the creation of public
institutions. Why should citizens living outside the country keep the right to
judge about the composition of representative organs whose decisions are only
binding for citizens residing inside the state territory? Only those citizens
who bear the consequences of their electoral decisions should be entitled to
vote. This problem of representation is particularly important in those
countries with a considerable number of citizens living abroad, as external
voters are likely to become crucial for the overall electoral result. It
follows that external voting is even more problematic for local elections,
where a local demos should vote for their representatives who will thus be
entitled to decide on matters of crucial importance for the local community.
6. With regard to the third element, the
question of nationality, we have, on the contrary, seen an expansion of voting
rights that is restricted to local level elections. EU citizens are indeed
entitled to register for local elections in other EU member states under the
same conditions as its nationals, subject to detailed arrangements which may
provide for derogations where warranted by problems specific to that member
state. A 1992 Convention of the Council of Europe on the Participation of
Foreigners in Public Life at Local Level recommends granting foreigners the
right to vote and stand for local elections provided they have been lawfully
and regularly resident in the host country during the five years preceding the
election.
7. In sum, provisions about electoral
rights should take into account different functional demands of national,
regional, and local levels. We might add two additional caveats. First, it must
be possible to exercise voting rights effectively. Without procedural rules
ensuring a fair, legal and transparent organizational context formerly disenfranchised
groups will still be hindered from effectively casting their votes or having
them counted properly. Second, the political importance of granting voting
rights to contested communities is directly linked to the relevance of the
decisions that the elected bodies might effectively be able to take. Although
the symbolic value of participating in elections should not be underrated,
turn-out will always depend on the expectation that representative organs
matter for the lives of the voters and replace established modes of decision-making
in the local arena (centralist, informal (military), or according to
traditional social norms). It is thus certainly much more important to struggle
for the suffrage of disenfranchised groups in the context of national
elections.
2.2. The Croatian Provisions
8. The Croatian
Local Elections Law provides a conventional definition of voting rights.
Suffrage is granted to all Croatian
citizens who have turned 18 years of age, and with the permanent residence in the area of the unit for the
representative body of which the elections are conducted (Art. 2, Para 1).
No voting rights are granted to resident non-citizens. The analysis of the
legal document meant to regulate local and regional elections reveals a high
conformity with international practice and no specific discrimination of any
minorities.
9. What seems to be problematic is less
the wording of the law than (a) the specific political context of the recent
elections and the (b) application and interpretation of the terms ‘permanent
residence in the area’ within the actual political context. Six years after the
end of violent conflict in Croatia the return of Serb refugees, equal
opportunity for citizenship rights regardless of ‘nationality’ and the full
restoration of property rights continue to be unresolved or only partially resolved
issues.
10. The decision of the Constitutional Court
from 1999 that allowed ethnic Croats living in Bosnia and Herzegovina and
holding dual citizenship to vote for local government elections in Croatia, may
be based on relevant provisions in the Constitution but certainly not in the
Local Elections Law (that was not even enacted by that time). While there are
good reasons to relax the residency requirements for national polls (especially
with regard to the importance of minority-sensitive political decisions taken
at this level), such arguments hardly exist for local and regional elections.
With freedom of movement guaranteed all citizens living currently in or outside
the national territory might decide to return to their place of permanent residence
until the next local and regional elections of 2005. All others should be
barred from influencing the outcome of local contests. The difficulties
of former Croatian citizens of Serb origin to renew their citizenship and thus
their voting rights from their current place of residence, is a problem of
citizenship laws, and certainly not an issue to be regulated by electoral legislation.
11. Similarly, organizational obstacles
witnessed during the 2001 elections in some polling stations for ‘displaced’
Serb voters represent a problem in the application of the relevant legislation,
not of the legal document itself. As is described in the electoral reports, the
specific voting arrangements for displaced persons were not covered in the
Election Law. To sum up, the responsibility for these problems and their
negative implications for minorities cannot be attributed to the Electoral Law
but to ad hoc executive or judicial decisions that deal with the transitory context
in 2001.
3. The Electoral System
3.1. Types of Electoral Systems and the
Protection of Minorities
12. Any
assessment of electoral systems relies on some assumptions that need to be made
explicit. The two most fundamental of these assumptions that underlie the
following analysis is that a) electoral systems are often powerful levers for
shaping the content and practice of political processes, and b) that there is
no single electoral system that is likely to be best for the protection of
minorities in all countries. In our discussion we also follow the mainstream of
international electoral research in arguing that the working of electoral
systems at the local level has no inherent logic different from the working of
such systems at the national level. The concepts and insights from comparative
research on electoral systems that are mainly drawn from national elections are
thus relevant for an analysis and engineering of electoral systems at the local
level.
13. Most of the thinking about the effects of
electoral systems is not particularly interested in the protection of
minorities. We should, however, not conclude, that the effects of electoral
systems are therefore indifferent to the protection of minorities. Opting for
majoritarian, combined or proportional systems (and their sub-types) might have
important consequences for the representation of minorities. We distinguish two
types of strategies of minority protection through electoral systems design: We
might speak of indirect strategies,
insofar as the electoral system is mainly concerned with satisfying other
functional demands (governmental stability, accountability, proportionality,
simplicity) and does not directly address the question of minorities. Direct strategies, on the contrary, consciously
tackle the question of minority rights within the electoral system.
a)
Proportional representation systems
(PR) normally facilitate a representative legislature and include all
significant social groups in the parliament regardless of the extent or distribution
of their support base (which is of central importance within majoritarian
systems). The establishment of closed party lists allows like-minded parties to
place representatives of minorities high on their party lists. But the huge
differences within PR systems should not be neglected. We may have different
sizes of constituencies, different forms of lists (blocked or open), and different
formulas for converting votes into seats (d’Hondt versus Hare quota) that all
have an impact on the chances of minor parties and groups to gain seats. There
can be a multiplicity of ways in which these elements are combined. Especially
the size of constituencies is an important variable. Small multi-member
constituencies create de facto thresholds of representation that reach
sometimes much beyond the legal thresholds applied via electoral laws. The
smaller the constituency is, the smaller is the probability that a
(proportional) representation of minorities will be achieved. In elections to
small local councils with only seven to ten seats, minor parties need to obtain
between eight and 12 percent of valid votes in order to get elected.
b)
Research on electoral systems in divided societies has identified the so-called
Alternative Vote (AV) electoral
system as a promising strategy to counter extremism and conflict behaviour. AV
is a majority system where voters are required to declare not only their first
choice of candidate on a ballot, but also their second, third, and subsequent
choices amongst all candidates standing. Candidates who wish to maximise their
electoral prospects have thus a strong incentive to garner for the second
preferences of voters from other (minority) groups. AV is a personality-centred
system, needs single or very small multi-member constituencies, and has never
been applied in local elections, to the knowledge of this author.
14. Direct strategies,
on the contrary, explicitly recognize the presence of contending ethnic,
religious groups or races within the electoral system. In most of these cases
the representation and the ratio of different groups in the representative
bodies is fixed before the elections. We might distinguish two basic options:
a)
Separate electoral rolls: Each defined group/minority has its own
electoral roll, and elects only members of its own group to parliament. The
entire system of parliamentary representation might thus be divided on a
communal basis (like in the Fiji Islands), or, in the case of national
minorities, separate rolls are established exclusively for the members of
specific ethnic or religious groups. One example is the optional separate roll
for Maori voters in New Zealand. Maori electors can choose to be on either the
national electoral roll or a specific Maori roll, which elects five Maori MPs
to Parliament.
b)
Quota systems: Different models operate at the level of candidature,
with one single electoral roll at the voter level. Here the number of seats
attributed to each group or minority might be fixed before the election, but
members for parliament are still elected by all voters. This system operates in
Lebanon at the national level, but countries as diverse as Jordan, India,
Colombia or Taiwan provide for reserved seats as a way of ensuring the
representation of selected minority groups. In the specific model of Mauritius,
quotas are not fixed, and best loser seats are allocated to the highest polling
candidates of under-represented ethnic groups in order to balance ethnic
under-representation resulting from an unrestricted electoral competition.
3.2. Mechanical and Psychological
Effects of Electoral Systems
15. There are two
types of effects a electoral system may have: a mechanical (technical) effect
(which is inherent in the procedure of translating votes into seats) and a psychological
effect (which regards the specific incentives a electoral system may create for
candidates and electors).
16. The mechanical
effect is easily described: With exactly the same number of votes for parties,
one system might lead to the representation of minority parties in councils and
parliaments and another system to a single party assuming majority control of
parliament / government. It is possibly to re-run elections under different
electoral formulas (plurality versus proportional representation) in order to verify
these mechanical effects.
17. The psychological
effect is related to the impact of the system on the decisions of candidates
and voters. It is much more difficult to describe or to evaluate, as we cannot
measure the psychological effect or the incentives that a specific system
provides. Electoral systems may reward particular types of behaviour and place
constraints on others. A minor party which runs the risk of not being able to obtain
the share of votes required by a legal threshold is impeded by the latter not
only de facto but also psychologically: the voters may be afraid of losing
their votes and consequently they vote for another party.
18. Both the mechanical and the psychological
effects of electoral systems are interrelated with two main contextual factors
specific to the country concerned, i.e. (a) the nature of group identity and
the demographic and geographic distribution of minorities, and (b) the
intensity of conflict and the stage of post-conflict democratisation. The first
variable asks for the foundations and the degree of malleability of ethnic
identities (i.e. how rigid and fixed). It also takes into account the spatial
distribution, the relative size, number, and degree of geographic concentration
or dispersion of minorities. The second factor regards simply the intensity and
depth of hostility between the competing groups, and the presence and
acceptance of accompanying accommodative institutions within the political
process (outside the electoral system). The design
of electoral systems, i.e. the policy of conscious manipulation of rules in
order to produce specific outcomes (say proportional representation of minorities),
should without any doubt be highly sensitive to these contexts.
19. As a preliminary conclusion we might thus
consider likely impacts of the aforementioned electoral systems on minority
protection. PR list systems tend to
be strong on the mechanical side, i.e. in delivering a highly accurate representation
of all groups and minorities (at least in so far as they organize politically).
Of course, the inclusionary character of the system hinges upon the size of the
constituencies. On the negative side, as it is easy to mobilize support by
playing the ethnic card, major parties may have every incentive to emphasize ethnic
issues and appeals. PR is thus weak on the incentive structure. It needs
accommodative elites, because voters are not induced to look across the ethnic
boundaries. Alternative vote is a
system which is less concerned with proportional results and places more
emphasis on the need to force different groups to work together. The single
most important condition for a success of this electoral system is that there
are several ethnic groups competing in a constituency. Where a candidate is
confident of achieving an absolute majority of first preferences due to the
domination of his or her own ethnic group in an area (say over 50%), they need
look no further to win a seat.
20. Direct strategies such
as quota systems or separate rolls are certainly the strongest on the
mechanical side, insofar as the numerical representation of minorities is
assured a priori, i.e. before the elections. Such systems have, however, no
incentive to ethnic accommodation, tend to reinforce the formation of
minoritarian political parties. Candidates have no incentive to attract also
votes from other ethnic or religious communities. Quota systems are not aimed
at overcoming ethnic differences but at reflecting openly and assuring the representation
of minorities. Such electoral formulas should thus be preferred only when
neither elites nor the voters are likely to display moderation on their own,
i.e. whenever political mistrust in the aftermath of violent conflict is high,
or religious cleavages politicised and deep-seated.
3.3. Analysis of the Croatian
provisions
22. As argued above, PR systems are generally
considered to be favourable to the interests of politically organized
minorities. While the option of the Croatian legislators for such a system is
thus generally positive from the perspective of minority protection, it is less
with regard to the 5% threshold. The intention of the threshold is always that
of excluding small political parties from the representative body and promoting
the concentration of the local party system. It never serves the interests of minorities.
While the small size of many constituencies renders superfluous the application
of the legal threshold (de facto thresholds caused by the magnitude of the
electoral districts being of greater significance than the legal hurdle), there
is no reason to introduce such thresholds in the larger constituencies when the
format of local party systems and the number of contenders is unknown (and the
Electoral Law is concerned with proportional representation of minorities in
other provisions).
23. Two articles in the Electoral Law deal
explicitly with the representation of minorities. According to Art. 9 the statutes of local and regional
self-government units shall determine the number of members of representative
bodies from among the Croatian citizens, members of ethnic and national
communities or minorities, in accordance with the proportional share of their
members in the total population of the unit. When proposing their slates of
candidates, parties are obliged according to Art. 15 to give the names, ethnicity, addresses and personal
identification numbers of candidates.
In combination with the PR electoral system in a single constituency this provision
signifies that all parties need to take into account the ethnic quotas
determined by their self-government unit when compiling their slate. How
exactly this could be made is not explained in the Electoral Law. Art. 11 para
3 mentions only the obligation to take care of the principle of gender
equality, but is silent on ethnic proportionality. Even those parties that
might be seriously committed in presenting such proportional slates would not
able to guarantee any proportionality in the outcome, as the overall composition
of the council is the result of the winning candidates of different party lists
(and not different ethnic lists).
24. Art. 23 points out: All voters having permanent residence in the area of that unit who come
to the polls, shall elect, on the basis of the slates of candidates, all
members of the unit’s representative body. This provision excludes any
separate voter rolls or ethnic representatives elected by the voters of their
group exclusively. If we take the letter of the legal provision we have thus an
electoral system limited at using direct and explicit strategies to protect
minorities by reserving quotas at the level of candidature without however
assuring proportional outcomes. At the same time, all parties proposing
candidates are obliged to present mixed slates of candidates and the system
thus includes a major incentive for inter-ethnic coalition building. Without
any clear provision about the ranking of candidates within the slates, the
necessity to include proportionally representatives of minorities does indeed
not guarantee their election to the local and regional councils.
25. There exist two solutions to address this
problem. The first solution would remain within the overall logic of the
electoral system by providing for best loser seats according to the
Mauritius model. If the electoral outcome did not reflect the established
quotas, the lowest-ranked winners according to the d’Hondt formula were to be
substituted by the best-placed minority candidates from the same lists. The
second option would be to shift to a majoritarian electoral system with open
party lists (and possibly multiple voting) where voters cast their votes
for single candidates on the lists, and the seats would be distributed to the
candidates obtaining the highest number of votes taking into account the agreed
seat ratio between the different groups (Lebanese model).
26. The application of all such systems,
however, has inevitably two drawbacks, as first someone has to decide on the
quotas, and second due to migration and demographic change the population
ratios of different groups may shift and the quotas need to be revised
regularly. The Croatian Local Electoral Law has opted here for a less fortunate
solution in giving the local and regional self-government units the autonomy to
fix the quotas. It is true that the interim provision of Art. 61 (that will be
analysed in detail below) with its mentioning of the census results indicates
that the local and regional bodies will have a limited autonomy in interpreting
the proportional shares of their minorities. But it would seem much wiser to
confer these competencies to an independent commission that is established
before each local/ regional election, composed of parliamentarians and/or representatives
of minorities, that decides on the basis of the census results and, if necessary,
additional evidence.
27. The difficulties encountered during the
2001 elections are a perfect illustration of these dilemmas. As the Electoral
Law was prepared just in time before the elections, the electoral system did
not work as described above. Art. 61 states that at the regular elections 2001, the proponents of slates shall, while
compiling the slates, acknowledge the principle of the adequate representation
of the minority population, taking into account the local circumstances.
The Law additionally provides for by-elections to be held within 90 days after
the results of the 2001 census are published in order to correct eventual
under-representation of minorities. It has been pointed out in detail in the
OSCE report that these provisions lack specific operational rules how exactly
to organize such by-elections. And it remains to be seen to what extent such
by-elections will ever be held, if the results of the census are published.
28. The Croatian legislator has made it clear
that the census should serve as a basis for the definition of quotas and
reserved seats. At the same time the census has reserved to citizens the right
to declare or not to declare their allegiance to a ‘nationality’. The
proportional quotas will therefore reflect only the figures of the declared
adherents of such minorities. But there are certainly no better definitional criteria
than self-ascription. Such dilemmas are aggravated by the specific refugee context
of Croatia with a number of would-be citizens of Serb origin that may be
hindered from applying for citizenship (and consequently to increase the minority
quotas within their original places of residence), and data on Croatian
citizens who have never lived in Croatia that may artificially blow up quotas
of ethnic Croatians within contested constituencies. Electoral provisions can’t
offer any short-cut to the resolution of these problems. If the decision is
taken to explicitly reserve quotas within the electoral system there is also a
need for definitional criteria, and the census is certainly the most objective
one available. As minority representatives have already raised scepticism with
regard to the reliability of the census results, the definition of quotas for
the 2005 elections should take into account (as proposed above) additional
evidence or data. Whatever data are used to calculate proportional quotas for
minorities, the ethnic identification of voters will not always remain a
secret, especially in the smaller local constituencies even without formal
identification in the voters’ registers.
29. The explicit reservation of seat
contingents for minorities should always be the strategy of last resort if
anything else fails. It may temporarily be the adequate policy in a country
still marked by civil war and refugees. To rely on the free democratic
competition is in the medium- to long-term a much better strategy, and it
avoids the politically sensitive issues of searching objective criteria and obliging
citizens to define their ethnic identity in an either-or way that may no longer
reflect the social realities in the country. Finally, we have a number of local
contexts where the minorities are not really minoritarian. The Electoral Law
even includes a special clause that somehow ‘protects’ ethnic Croatian
interests by assuring proportional representation in those units in which members of ethnic and national
communities or minorities comprise the majority of population (Art. 9 para
2). As far as minorities are regionally concentrated (like the Italians) they
might indeed not need any assured proportional representation in local councils
because they would automatically be relative strong contenders in those
localities where they live.
4. Preliminary Conclusions
30. Our analysis
of the Croatian Local Electoral Law leads to a somewhat puzzling conclusion.
The regular and the interim provisions of the Electoral Law
need to be analytically distinguished. The regular
provisions represent with some minor exceptions, a valuable approach, but need
specification in the perspective of the forthcoming 2005 elections. The late
adoption of the Law led to a situation where a number of minority-relevant
provisions could not be implemented, but it is beyond the competence of this
author to evaluate to what extent this lacuna results from the difficult
organizational contexts or a lack of political will to guarantee the fair
representation of minorities within elected local and regional institutions.
31. Although the electoral system provided
for in the Law if implemented (and complemented by additional provisions like
the two options presented above) could offer good prospects for ethnic conciliation,
the obligation to present mixed lists in the local and regional contexts is
without any doubt a very strong form of electoral engineering that relies most
probably on some elementary social pre-conditions. Without any political will
to calculate quotas and establish mixed lists such an ambiguous electoral system
might also produce counter-productive effects. The only remaining alternative
is then the establishment of separate rolls, i.e. separate elections for
minorities by minorities.
32. Finally, there is certainly a need to
improve the coherence of the different constitutional and legal provisions that
regulate citizenship and electoral contests. The consistent definition of
ethnic groups and minorities that are entitled to be represented is a
particular urgent case. One should, however, not think that a complete harmonization
of electoral regimes between the different territorial levels will serve best
the interests of the minorities, nor the principles of democratic representation
in general. Whereas minorities need a formal and strong protection at the
national level, minorities operate in different circumstances at lower
territorial levels, and might benefit from less rigid and formalized
proportional rules. A consequent policy of fair representation needs to also
tackle the question of proportional access to public office at the local level
which necessarily goes beyond legislative institutions. And while there are
good reasons to allow external votes for national elections, we can’t see any
in the case of local and regional elections.
