Electoral Institutional Change and
Democratization: You Can Lead a Horse to
Water, But You Can't Make it Drink
JØRGEN ELKLIT
Drawing on evidence from first and second elections in emerging democracies, this
article addresses issues pertinent to understanding the interplay between institutional
design and change on the one hand and on the other hand the course of the
democratization process, its background and concomitants. Because the electoral
system (understood broadly, that is, not only as seat allocation rules) is often seen as
an institution conducive to the legitimation of emerging - and therefore fragile democracies,
it is important to examine how the development of the electoral system
has influenced the democratic transition outcome, in a number of cases. Cases
considered include Kenya, Mongolia, Nepal and Tanzania, while comparisons with
countries outside the third world such as Bosnia-and-Herzegovina and South Africa
throw additional light on the argument. By seeking to avoid both the fallacy of
electomlism and the fallacy of anti-electoralism, the article argues that the electoral
system is a crucial factor behind the unsatisfactory course of the democratization
process in many countries. More emphasis should be put on sustaining the different
prerequisites of democracy, while simultaneously more effort should be put into the
preparation of elections, in order to progress beyond simple and inadequate electoral
democracy. The technical refinement of the electoral process, while far from sufficient
to guarantee democratic development, can be a step in that direction. Credible and
transparent elections are conducive to internal legitimacy and the dynamic of the
political-electoral process will gradually bring along more contestation, more
participation, and the enjoyment of more rights and liberties.
Electoral and Other Democracies
The fallacy of electoralism1
has increasingly been recognised - at least in
the scholarly community - as a problem which aptly denotes the mistake of
confusing the holding of elections with the advent and development of
democratic regimes. Contemplating what has happened to 'the third wave'
Jørgen Elklit is an Associate Professor in the Department of Political Science at the University of
Aarhus, Denmark. An earlier version of this article was presented at the European Consortium of
Political Research workshop on 'Democracy in the Third World: What Should Be Done?', in
Mannheim in late March 1999. The text in its present form has benefited from comments by
Deborah Schein, Palle Svensson, and the participants in the Mannheim workshop. The constructive
assistance of these good people is highly appreciated.
Democratization, Vol.6, No.4, Winter 1999, pp.28-51
PUBLISHED BY FRANK CASS, LONDON
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 29
of democracy, Larry Diamond adds a welcome extension to the debate
about the relationship between the holding of elections and democratization
- and thereby also to the conceptualisation of democracy - by focusing on
three distinct categories of non-authoritarian regimes: pseudo-democracies,
electoral democracies, and liberal democracies.2
What the three categories
of 'democracies' have in common are elections for public office taking
place intermittently, but the categories differ decisively in the degree to
which they actually allow meaningful competition and participation. In
Diamond's words:
Contemporary minimalist conceptions of democracy - what I term
here electoral democracy, as opposed to liberal democracy commonly
acknowledge the need for minimal levels of civil freedoms
in order for competition and participation to be meaningful. Typically,
however, they do not devote much attention to the basic freedoms
involved, nor do they attempt to incorporate them into actual
measures of democracy.3
The crucial distinction between electoral and liberal democracies is related
to whether or not political rights and civil liberties are primarily seen as
crucial to ensuring meaningful electoral contestation and participation, or if
they are also cherished between elections and are allocated wider importance
to ensure other democratic functions as well.4
The third category consists of pseudo-democracies, which according to
Diamond are 'less than minimally democratic but still distinct from purely
authoritarian regimes. [They] ... have legal opposition parties and perhaps
many other constitutional features of electoral democracy, but fail to meet
one of its crucial requirements: a sufficiently fair arena of contestation to
allow the ruling party to be turned out of power'.5
Pseudo-democracies
obviously take different forms, but the common denominator is that they
tolerate the existence of opposition parties. This is potentially important and
not only theoretically - since the presence of legal opposition parties
could be the foundation upon which future democratic development will
build.
Diamond also suggests that the 'free' rating in the Freedom House
surveys of freedom in the world6
offers the best available empirical indicator
of liberal democracy. Much can be said in support of that claim, as
demonstrated by a comparison between the Freedom House survey
methodology and the various elements Diamond lists for the purpose of
clarifying how liberal democracies differ from the minimalist/formal/
electoral democracies.7
However, Freedom House still presents the electoral
democracies as the more comprehensive category which then subsumes the
'free' countries (Diamond's indicator of the number of liberal democracies).
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
30 DEMOCRATIZATION
We have here an obvious example of the phenomenon discussed by Collier
and Levitsky, where they juxtapose 'moving up the ladder of generality' and
the establishment of 'diminished sub-types'.8
The methodological problems involved in assigning scores to the
various items on the political rights and civil liberties scales, the multidimensionality
of the measurement of these two interwoven clusters of
variables, the inherent problems of validity and reliability, the absence of a
more elaborate weighing procedure, and the arbitrariness in setting
thresholds between categories cannot be put aside as mere methodological
truisms which one should not be concerned about. The point has been
demonstrated elsewhere, based on a scrutiny of the electoral process in five
transitional countries.9
Obviously, the empirical indicator of Diamond's electoral democracy
category should be a country's appearance on the Freedom House list of
countries which are electoral democracies without at the same time being
also in the Freedom House category of 'free countries'. In 1997-98, that is
in early 1998, there were probably 36 such countries (that is, 117 electoral
democracies compared with 81 free countries). The number of pseudodemocracies
can then roughly be estimated as the remainder of the Partly
Free category (21 countries in the 1997-98 count), permitting some
borderline cases. One problem encountered during the classification of
countries for this article (see Table 2 below) is that countries scoring 3 on
civil liberties were not easily classified as liberal democracies. A greater
number of electoral democracies than the reader might suspect from the
account given so far should therefore be expected.
The gross Freedom House figures indicate that the sizes of the four
categories outlined by Diamond have not changed much since 1995. This
apparently supports his point about stasis, namely 'that the third wave of
democracy has come to a halt, and probably to an end'.10
However, such a
drastic conclusion requires - as a minimum - annual transition tables
between categories (three-by-three or four-by-four) which neither Diamond
nor Freedom House have provided. Nor have such tables been presented in
recent discussions of whether illiberal democracy is increasing or declining
in the world. This discussion cannot be settled by looking only at the
increasing percentage of countries among the electoral democracies ranked
by Freedom House as free."
Diamond points to three fundamental differences between full-fledged
liberal democracies and electoral democracies: (1) qualitatively better
elections as well as the absence of reserved power domains for the military
or other forces not accountable to the voters in the liberal democracies, (2)
'vertical' accountability of elections is complemented with 'horizontal'
accountability in liberal democracies, and (3) extensive provisions also
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 31
between elections for political and civic pluralism, as well as for both
individual and collective freedoms.12
The basic features of liberal democracy
are spelled out in nine specific elements which overlap considerably
with Robert Dahl's seven institutional requirements for polyarchy.
Diamond's own recommendation, after his conclusion about the possible
end to the third wave of democracy, is to engage in undertakings which in
all three kinds of democracies will be conducive to the consolidation - or
further development/completion in the two least developed categories - of
the democratic regime, namely institutional, policy, and behavioural
changes.
The Case for Conducting Qualitatively Acceptable Elections in
Transitional Democracies
In immediate response to Diamond's recommendation,13
the argument
presented here is that the development of a fully legitimate electoral regime
- which might later develop into a more complete (deepened, consolidated)
liberal democracy - requires an electoral process acceptable to all major
stakeholders, including the ordinary voters.
Obviously, the holding of acceptable elections is only a necessary, not a
sufficient, condition for a development towards a full-fledged liberal
democracy. All attempts to improve the electoral process, however, will
necessarily increase the social costs of subverting the democratization
process. This is a crucial element in negative consolidation, that is efforts to
avoid democratic regression, which is at least as important as positive
consolidation. It is certainly one important means of counterattacking the
reduction of an electoral democracy to a pseudo-democracy or something
worse, a point made by many writers.14
This is not to imply that one should concentrate on the electoral field
only, far from it. Diamond indicates a number of causes also worthy of
attention if the intent is to consolidate democracies within all three
categories and the various intermediary forms. But it will be instrumentally
valuable to improve on the electoral component, not only in the liberal
democracies as part of an ongoing effort to increase the democratic score,
but also - and even more so — in electoral democracies which must be lifted
up one category, if democracy is to be completed, as has been pointed out
by Schedler.15
The claim of this account is thus that all variants of democratic
transition, development, and consolidation - positive as well as negative in
all not fully consolidated democracies, in the third world and elsewhere,
will benefit from a renewed and reinforced interest in, analysis of, and
improvement to the electoral system and the entire electoral process.16
Such
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
32 DEMOCRATIZATION
activities are important tasks in a democratization context because they are,
directly and indisputably, conducive to the legitimation of new and fragile
democracies, as discussed by Lijphart and Waisman.17
Only when voters
experience meaningful contestation and participation in the political
process, within the framework of relevant institutions and a certain socioeconomic
development, and only when they understand, however
imperfectly, the importance of political rights and civil liberties, will they
develop some kind of normative commitment to democracy. That is an
ingredient in the consolidation process which should not be forgotten, in
spite of the focus that many analyses train on the importance of elites and
their behaviour.18
In her discussion of the prospects of democracy in Central America,
Karl outlines an 'optimistic' as well as a 'pessimistic' scenario. According
to her, the development in the undemocratic, pessimistic direction is
facilitated if - among other things - elections are generally seen as flawed,
losers are inadequately represented and so on.19
There is no reason why this
should only be valid in relation to Central American countries. The same
reasoning will probably apply to most, if not all, third world countries
experiencing democratic transition. Moreover, there is actually no reason to
restrict the point to third world countries only, as it will probably apply to
most countries in political transition.
This is not the same as saying that qualitatively better elections will in the
course of time solve all problems, for that would be a regrettable relapse to
electoralism. As Bratton also believes, it is possible to steer a middle course
between the two fallacies of electoralism and anti-electoralism.20
Many
observers have recently pointed out that the prerequisites for democracy are
in scarce supply, which explains the tendency to democratic stagnation and
retrenchment encountered in some parts of the world.21
Crucial social,
economic, and political cultural conditions are obviously lacking, and the
same might apply at the institutional level. Qualitatively acceptable elections
and the framework for such elections are only some of the many institutional
preconditions which must be in place before we can be truly optimistic about
future democracy in the third world. But that is just one more reason for
engaging in the process of developing and refining the electoral process and
the understanding of what is needed in order to have 'free and fair' elections
as an element in progressing towards a consolidated democracy.22
It should be possible to identify the phases where the electoral process
has failed, in order to provide the tools for doing better if and when
conditions are conducive to qualitatively better elections, which in any case
requires that the incumbent government/party will actually allow a true
democratic process to unfold. The recent focus on political elite behaviour
- sometimes carried out under the label of agency factors - makes it obvious
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 33
that the democratic intentions and inclinations of such political and
administrative elites might be tested by looking at how they legislate and
regulate the electoral process. Do such elites actually do what is needed or
do they do less? That appears to be the main question. And if they do less,
what are the reasons for not doing better: is it lack of knowledge of electoral
processes, lack of money, or lack of democratic intent, or some combination
of these?
The first electoral process requirement which one should be concerned
about is one stated not only by Diamond in the quote already cited but also
by many others, namely the provisions for political and civic pluralism, as
well as for both individual and collective freedoms. This concern coincides
with the main tendency in recent discussions of how to develop and refine
post-election evaluation procedures and approaches. The key question
('Was it free and fair - or at least acceptable?') to a considerable degree
hinges on what actually happened before polling day - and even before the
electoral process started.23
This perspective, namely that it is necessary to focus mainly on the prepolling
period if one's intention is to improve the quality of elections in a
democratic perspective, implies stressing the presence of, respect for, and
unhindered use of relevant political and civil rights and freedoms. These are
the very qualities that Dahl convincingly argued are prerequisites which
must be in place, in other words must be institutionalised, before we can
even think of elections being potentially free and fair. However, these rights
and liberties are of value not simply during the time of election campaigns
alone. It is democratically unacceptable to restrict the use of political rights
and civil liberties to a period every fourth or fifth year called election time.
The democratic process cannot be and should not be restricted in time. Dahl
puts it this way:
In this perspective, free and fair elections are the culmination of the
process, not the beginning. Indeed, unless and until the other rights
and liberties are firmly protected, free and fair elections cannot take
place. Except in countries already close to the thresholds of
democracy, therefore, it is a grave mistake to assume that if only
leaders of a non-democratic country can be persuaded to hold
elections, then full democracy will follow.24
It has been depressing to see how little concern for the complexities of the
transition processes western politicians and development aid administrations
and also many first time election observers have been able to
muster. And it has been surprising to observe the swiftness with which
many academics have forgotten that the course of democratization is also
dependent on a complicated and interwoven set of structural preconditions,
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
34 DEMOCRATIZATION
including political history and culture, literacy levels, the existence of a
middle class, or the level and distribution of wealth. The literature on the
relative importance of socio-economic conditions and political factors in
Latin America has been convincingly reviewed by Mainwaring, but it is
worth asking whether an even better understanding of the relationship
between the two would follow if the situations during first and second
transitions were analysed separately.25
It is worth contemplating if the second transition (the consolidation
phase) is generally more dependent on structural factors, while the first
transition (the regime change) might also or possibly more so depend on
agency factors. The possibility of checking the democratic intentions and
inclinations of political and administrative elites by looking at the
legislation and regulation of electoral processes then becomes particularly
pertinent, as the first transition will normally include the founding election,
while the second transition will include second and subsequent elections.
This perspective might help explain the remarkable decline in quality that
Bratton's survey found between founding and second elections in subSaharan
Africa. Be that as it may, both founding and second (or later)
elections can be used to evaluate the intentions of ruling elites.26
One surprising consequence of the ordinary international approach to
support of democratization and elections has been a general lack of
willingness to engage in. serious discussions of election quality, both before
and after the election event, and in connection with first and second
elections. The impression is easily gained that the mere holding of an
election was enough for the international community - with or without
foreign observers - and that the interest in seriously evaluating the entire
election exercise was minuscule outside academic circles, no matter how
many foreign observers were flown into the country.
Inside academic circles the ability and willingness to engage in election
evaluation has also differed as is evident from comparisons of different
evaluations or analyses of the same elections. It is not difficult to understand
this as election evaluation is a risky and difficult business because of the
interwoven methodological, theoretical, and empirical complexities of
grading performances along different process dimensions without having a
clear grading scheme and some agreed-upon weighing schemes.27
It is in this context interesting to note how difficult the development of
adequate election evaluation tools has been and how the current front line
in election observation methodology, development and refinement has
moved in the direction of upgrading the in-depth analysis of the pre-polling
period, instead of heavily investing in massive polling day observation.
Electoral systems28
are often seen as encompassing only allocation rules
for converting votes to seats. It can, however, be argued that a broader
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 35
electoral system concept exists covering the entire electoral process from
start to end, that is all formal and informal rules and practices which might
have a bearing on how political attitudes in the electorate are converted into
a certain distribution of seats in parliament. As has probably become
evident already, this discussion employs the broader electoral system
concept, integrating into it the system for allocating seats to parties on the
basis of the votes cast.
What Should Be Done?
The willingness of political and administrative elites to engage in activities
which will be instrumental in achieving a level of democracy over and
above that of an electoral democracy is not particularly difficult to measure
and evaluate. The main test is whether key decision-makers are willing to
engage in legislation and regulation which focus on the general democratic
situation and not only on the more narrow or trivial election related
activities. Basic political rights and civil liberties are one such issue obviously
a sine qua non - and so is the overall institutional set:
up for the
election, such as the electoral law, the seat allocation system, and the
provisions for having a truly independent and impartial electoral commission.
Election quality is not achieved by concentrating on election day
activities only, far from it, even though they are still important - but only if
the course of events during the pre-polling phase is acceptable.29
The prevalence of this understanding is increasing, and election
administration and electoral system advisors as well as organisers of
electoral observation missions are therefore well advised to pay considerably
more attention than previously to the pre-polling phase. What
goes on before polling is the main key to both success and failure. An
indication of the elements which are particularly important during the three
main phases is given in Table1, which also shows how the various elements
can be tentatively attributed to the two traditional dimensions of electoral
observation.
It is possible to approach the issue of what should be done by looking at
genuine third world electoral success stories. They are, however, few and
far between. It might also be difficult to decide how much of these
experiences can be generalised and how much is to be attributed to
particular conditions which evade generalisations. Recent examples of good
quality elections include Namibia since 1989, South Africa 1994 and 1999,
Ghana 1996, and Mongolia 1996.
It is much easier to find examples of problematic third world elections
which have contributed to setbacks to the democratization processes. It
might, however, be equally difficult in relation to such cases to tell what can
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
36 DEMOCRATIZATION
TABLE 1
KEY ELEMENTS IN AN ACCEPTABLE ELECTORAL PROCESS
'Free' 'Fair'
Before • Freedom of speech
polling • Freedom of assembly
day • Freedom of association
• Freedom from fear in
connection with the election
• Freedom of movement
• Absence of impediments to
standing for election
• Equal and universal suffrage
• A transparent electoral process
• An election act and a system for seat
allocation which grants no special
privileges to any party, group, or person
• An independent and impartial electoral
commission (and administration)
• Impartial voter education programmes
• Absence of impediments to inclusion in
the electoral register
• Adequate possibilities for checking the
provisional electoral register
• An orderly election campaign
• Equal access to public mass media
• No misuse of government facilities for
campaign purposes
On • Opportunity to participate in the
polling election
day • Absence of intimidation of voters
• Secrecy of the ballot
• Adequate provisions to ensure that voters
only vote once
• Well designed ballot papers without serial
numbers
• Access to the polling stations for
accredited party
• Representatives and election observers
• Impartial assistance to incapacitated
voters
• Proper treatment of void ballot papers
After • Legal possibilities of complaint
polling • Adequate possibilities for resolution
day of election related conflicts
' Proper counting and reporting procedures
• Proper precautionary measures when
transporting election material and securing
polling stations
• Impartial reports by the media on election
results
• Impartial treatment of election complaints
' Acceptance of the election results by all
involved
Source: J. Elklit and P. Svensson, 'What Makes Elections Free and Fair', Journal of Democracy,
Vol. 8, No. 3 (1997), p.37.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 37
be generalized and how much is to be attributed to particular conditions
evading generalization. Thus, Burnell's classification system for democratic
failures and their explanation does only provide more general points directly
related to the problem at hand, even though the most appropriate category
might often be 'false starts/premature democratization' and the best explanation
comprises a mixed bag of several of the explanations suggested.30
Another classification which might prove helpful in classifying elections for
analytical purposes has recently been provided by Merloe who in a
discussion paper distinguishes between post-conflict elections, breakthrough
elections, consolidation elections, elections in democratically backsliding
countries, and elections taking place as part of a managed transition.31
It might, furthermore, be difficult to say precisely what should have been
done and what might have worked better if implementation had actually
been feasible. In some cases problems were only identified after the event;
in other cases they were clearly spotted before the event, brought to the
attention of the relevant political or administrative authorities, and then
nothing or too little was done. So there is ample scope for improvement, if
the political and administrative readiness is there - and if it is not (as has
sometimes been the case) what kind of election - and what kind of
democracy - are we then dealing with? An electoral democracy voluntarily
sliding backwards? Merloe's five types of election contexts add an
interesting new dimension to this discussion.
Table 2 below, 'Elections Classified by Type of Democracy and
Quality', represents a first attempt to classify a sample of first and second
elections in new democracies according to the type of democracy and the
general quality of the election. The idea is to present qualitatively
acceptable and less acceptable elections together in order to see if this will
help generate useful conclusions. The basis for the classification is relevant
Freedom House scores in combination with an assessment of the election's
congruity with the criteria already displayed in Table 1- before, during, and
after the event. Any evaluation of the quality of an election must include the
degree to which it was 'free and fair', but it should also include a concern
for the level of inclusiveness (over and above the suffrage component of the
'freeness' dimension) as well as the technical quality (see below).
Suggestions for other elections that might be included in Table 2 are invited.
The next step is to identify common features within two groups of
elections: elections scoring high and elections scoring low on the combined
quality dimension. If the first group is tentatively seen as consisting of
elections in the two highest categories on the quality dimension, it appears
that what connects these rather different election situations is the intention
and the willingness of political and administrative authorities to have
elections of good quality in combination with a certain degree of respect for
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
38 DEMOCRATIZATION
the rule of law. The basic claim is thus that the key factor to explain the
overall character of these elections is the determination of the political (and
administrative) elite to have and to conduct elections which could be
considered acceptable (or even better than just acceptable, if possible) by
most local political stakeholders, as well as by the international community.
Precisely because of this determination (which might obviously be caused by
different motives for different actors, including that of the incumbent
president or party being convinced that they will win easily and therefore can
allow themselves to have elections which are acceptable by international
standards), the electoral process is as far as possible conducted according to
generally accepted norms and standards.32
This argument is, of course, related
to some basic points in Di Palma's discussion of democracy crafting.33
Exactly the opposite is the case when we look at elections in the two
categories at the bottom end of the quality scale. Obviously, agency factors
are not only conducive to a positive democratization; they can also - within
a given set of structural factors - cause the democratic development to slow
down or even fall back - and to do so both during the initial transition phase
and during the consolidation phase.
TABLE 2
ELECTIONS CLASSIFIED BY TYPE OF DEMOCRACY AND QUALITY
Liberal
Democracies
Good quality
Reasonable quality South Africa 1994, 1997
South Korea 1997
Below average
quality
Low quality
Electoral
Democracies
Botswana 1965+
Ghana 1996
Mongolia 1996
Namibia 1994
Malawi 1994
Mexico 1997
Mozambique 1996
Nepal 1991
South Korea 1996
Zambia 1991
Mongolia 1992
Tanzania 1995
Bosnia-and-
Herzegovina
1996, 1997
Pseudo-
Democracies
Singapore 1997
Kenya 1992, 1997
Nigeria 1999
Zambia 1996
Zimbabwe 1995
Note: 'Free' countries in the Freedom House lists are classified as electoral democracies, not as
liberal democracies, if their Civil Liberties score is 3 or worse. The order of elections
within cells is alphabetical.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 39
It is important to avoid a circular argument here. For this reason the
elections are classified according to how they score on the indicators in
Table 1 and only then are conclusions drawn about possible explanations of
the pattern that emerges, rather than proceeding by showing how such
explanatory factors contribute to election quality. Such classification is,
indeed has to be, qualitative in nature. However, a finer, possibly multidimensional
grading system might ruin the possibilities of drawing clear-cut
conclusions, for many fewer elections would be banded together.
It might, however, subsequently be instructive to see what specific
administrative and technical solutions have been chosen in these particular
situations, as this might be an adequate operationalisation of a well thought
out and orderly electoral process - a kind of benchmark - against which the
intentions and capabilities of political and administrative elites can then be
evaluated. In this way, Table 3 below might be instrumental in identifying
ways and means of improving and developing the electoral institution as
such, while issues related to political rights and civil liberties, allocation of
sufficient funding and so on must be taken into account separately.
The Electoral Process
The electoral process should be seen as unfolding in a number of
chronological systematic steps, some of which are also partly simultaneous
and partly overlapping. The structuring of the entire electoral process in ten
steps was first suggested by Bill Kimberling in a presentation of problems
related to voter registration.34
In Table 3 two more steps (steps 11 and 12)
have been added to give a slightly more complete coverage of the entire
electoral process. The outline of the twelve steps of the standard electoral
process demonstrates that there is a logical-systematic sequence of electoral
process functions to perform, which should come more or less in the order
indicated in Table 3 if the process is to unfold in a natural, orderly way and
without problems deriving from the order in which matters are addressed.
Good quality elections are then elections where the full respect for and
procurement of basic preconditions are accompanied by an election
preparation process in which the electoral calendar looks more or less like
Table 3, with ample time allowed for all the different phases. It is also a
precondition that legislation has provided an institutional framework
conducive to a meaningful contestation - 'a level playing field' - and a
high level of inclusiveness and participation - and that the law and all
accompanying rules and regulations are duly implemented at all levels.
This latter, partly more technical aspect of the electoral quality concept
includes for example that logistical matters are addressed adequately, that
counting and tabulation of the results are precise, and that ballot paper
accounts are correct.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
40 DEMOCRATIZATION
TABLE 3
THE TWELVE STEPS OF THE ELECTORAL PROCESS
Step Electoral process functions Clarifications
• Establishment of the legal
framework for the electoral
process
• Establishment of adequate
organisational management
structures, i.e. systems for
managing the electoral process,
including securing the adequate
financial and other means
• Demarcation of constituencies
and polling districts
4 • Voter education and voter
information
5 • Voter registration
6 • Nomination and registration of
political parties and candidates,
i.e. providing ballot access
7 • Regulation of the electoral
campaign
8 • Polling
9 • Counting and tabulating the
vote
10 • Resolving electoral disputes
and complaints; verification of
final results; certification
11 • Election result implementation
• Includes the entire electoral legislation, i.e. from
the constitutional framework to rules and
regulations
• The seat allocation system is also covered under
this heading
• The character and composition etc. of the
electoral commission, if any
• The relationship between the electoral
commission and its implementing arm, i.e. the
election administration
• Only the physical implementation, since the
relevant legislation belongs under Step 1
• This also goes for the decision on what decisionmaking
and administrative structures to apply, i.e.
if - for instance - a separate demarcation
commission is foreseen
• Might overlap chronologically with Steps 5 and 6
Must come before voter registration, but might
need to be repeated later also
• Might overlap chronologically with Steps 4 and 6
• Rules for public scrutiny of the voters' roll
• Might overlap chronologically with Steps 4 and 5
• Spending rules, if any
• Rules for access to publicly owned media, if any
• Code of conduct?
• Rules regulating the presence of party agents,
domestic and international election monitors etc.
• Counting at polling stations or centralised?
Are opposition parties allowed to check the
count?
• The electoral court system, if any
• Time limits for handling of electoral complaints
and disputes
• Are those elected actually allowed to perform
their roles, i.e. are there any administrative or
other hindrances?
• Are those elected actually willing to take office,
that is to perform their role?
12 • Post-election handling of election
material; production of the official
election statistics; archiving;
closing the books
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 41
Polling itself comes quite late in the process - as Step 8 - so it both
presupposes and depends on how well the seven preceding steps have been
performed. Because of this, it is regrettable that up until now the political
focus both locally and internationally has primarily been on the polling
process itself and the immediate outcome of that process in a narrow sense,
that is the distribution of votes among parties and candidates. It is also
remarkable, as more and more observers and analysts of elections in third
world countries seem to agree that too little attention has been paid in the
past to the early steps of the electoral process and too much attention to
what goes on at polling day, which for many reasons is unfortunate.
If it is a correct observation that third world elections of an acceptable
quality are only possible if and when political and administrative elites are
genuinely interested in having such elections - which then tends to be yet
another precondition - it becomes obvious why so many election and
democratization advisers are fighting an uphill battle and why so many
opposition party leaders are continually being frustrated. Multi-party
elections might be called and take place according to current legislation, but
the general electoral and administrative framework and the state of mind of
central political and administrative actors might still be such that good
elections —meaning elections conducive to some kind of democratic
improvement - are not really possible.
Some of the problem areas in the early phases of the electoral process
are listed and commented on below. Implementation of these different
elements where they are not yet in place will most certainly improve the
electoral process, so one claim is that this is what should be done. But
another claim is that the attitudes of the political-administrative elites
towards the implementation of these provisions is an adequate way of
testing the willingness of these elites to engage in democratic consolidation,
even though it may not be the only way of doing so and may not always be
the best. But it is easy to analyse the willingness to engage in electoral
institutional reform, because nothing will happen if key decision-makers
have not realised that this is a conditio sine qua non for improvement in the
score for overall election quality. This is where the problems lie, as
demonstrated by many third world countries, some of which featured in
the lower part of Table 2. Obviously, the political heavy-weights have
to be convinced one way or the other that this is what should be done, if
they want to be seen as having elections of an acceptable quality (but they
only rarely do). Obviously, their main concern is international legitimisation
for their regime.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
42 DEMOCRATIZATION
Areas of Concern
Step 1: Establishment of the Legal Framework for the Electoral Process
• The entire set of electoral laws, by-laws, rules and regulations and so on
should be in place and fully articulated before the calling of the election:
The rules of the game should be known before kick-off - and they should
not be changed at half-time! It is also an important principle that electoral
legislation should be understandable for the largest possible number of
voters. It is easy to mention countries where these principles have not
been adhered to. Sometimes it was not possible, as in Bosnia-andHerzegovina
1996 and 1997, or in South Africa 1994, where even the
country's constitution had to be changed during the process. In other cases
the incumbents wanted to make life more difficult for the opposition
parties, as in Kenya in 1992 when the Attorney-General used his power to
shorten the period for parties to nominate their parliamentary candidates
without even informing them.35
In yet other cases, the electoral legislation
is difficult to comprehend. The law has been amended over and over
again, which makes it difficult to read and understand as the provisions
have gradually become more inconsistent. In such cases a clear,
consolidated version is not much to ask for.
• The legislative framework includes provisions concerning the electoral
system (in the narrow sense), that is the system for allocating seats on the
basis of the votes cast. The importance of the electoral system for the
development of (1) inclusiveness, (2) accommodation and tolerance in
(previously) divided societies, and (3) party systems of a certain character
is well established - something that is not the subject here. However,
discussions about the criteria for keeping or changing a specific electoral
system36
only rarely attract the attention of incumbent parties who have
benefited from the system already in place (or expect to do so in coming
elections). This is easily explained as yet another case of rational
behaviour, for elites will often only legislate those institutional changes
which they consider instrumental to the advance of their own immediate
interests. An argument can often be made for using a proportional
representation system to achieve inclusiveness and to avoid turning
elections into an all-or-nothing event where the stakes are too high, as is
especially likely to be the case in cultures where losing is something that
makes one feel very ashamed indeed. The positive importance of using
proportional representation systems in several countries in southern Africa
- most notably in South Africa in 1994 and 1999 - is beyond dispute, even
though the debate about electoral systems for the region has been
flourishing since the early 1990s.37
The consequences for the electoral
outcome of a change in the seat allocation system has been demonstrated
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 43
on various occasions, including Mongolia 1992-96. Interesting
discussions have been conducted in many countries including quite
recently Tanzania and Jordan about the merits of different electoral
systems, but only rarely do the beneficiaries of the current systems see
such debates as particularly rewarding to engage in. A special - and
controversial - issue is the use of so-called 'special seats for women' (as
in Tanzania) or similar arrangements to promote female representation in
parliament.
Step 2: Establishment of an Adequate Election Management Structure
• A truly independent, impartial, and dedicated electoral commission is
instrumental for achieving an electoral outcome that at the end of the day
is acceptable to most contestants, particularly the losers. Appointment
procedures contribute to the way in which an electoral commission is
perceived by the political parties, that is to its credibility and legitimacy,
but in practice the appointment procedures are often not conducive to
such feelings among opposition parties. Public vetting of would-be
commissioners is one way of increasing public confidence, but various
models for electoral commissions and their appointment do exist, and it is
primarily the overall performance of the electoral commission which
counts.38
Examples of electoral commissions who have not been able to
gain the trust of some of the electoral contestants are many, and in some
cases there are good reasons for this.39
Obviously, losers might feel
inclined to blame the electoral commission for their poor performance.
Therefore, and because it contributes to the development of the very
essential understanding of the importance of the rule of law, it is
absolutely necessary that electoral commissions perform their tasks in the
most irreproachable way and are perceived as impartial.
• It has often been difficult for law-makers to construct and legislate a good
division of labour between the electoral commission itself and its
implementing arm (under the Chief Electoral Officer, or whatever the
local title is). The electoral commission is sometimes seen as a board of
directors, while the day-to-day problems might require a much more
'hands-on' administration, which can then complicate enormously the
relationship between the commission and its staff.
• Adequate funding of the entire electoral process is a sine qua non, even
though the resources available in the country might be too few to run a
fully acceptable election. Foreign funding has become a possibility in
some cases, but sometimes such money comes too easily, and is
encouraged by the funders as one way of co-opting the players, even
though the desired results might not be achieved.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
44 DEMOCRATIZATION
Step 3: Demarcation of Constituencies and Polling Districts
• Gerrymandering is a particular problem, especially in electoral systems
relying on single-member constituencies. Proportional representation
systems, or majority/plurality elections in multi-member constituencies,
often rely on previously existing administrative divisions and seat
allocation in relation to some objective standard (such as population or
electorate size) which is another attractive feature of such systems.
Gerrymandering in third world countries often takes the form of
allocating more seats to regions where the incumbent party/group is
strong than to other parts of the country (Kenya 1992 and 1997; Nigeria
1999). Different constituency sizes are, however, fully acceptable in
regions where geographical conditions or residential patterns make it
unreasonable to go for full electoral equality (as in the Himalayas, Nepal
1991+).
Step 4: Voter Education and Voter Information
• Impartial and precise voter education must start early in order to ensure
that all potential voters are aware of why and how to register. It should be
the responsibility of the electoral authorities to conduct such campaigns,
particularly in societies and regions with a high rate of illiteracy.
Regionally (or otherwise) skewed information campaigns by public
authorities are not acceptable, while it is unrealistic to demand the same
standards from the political parties. Their contribution should, however,
be appreciated, because of their interest in going an extra mile.
Step 5: Voter Registration
• Politically equality unconditionally requires that all members of the
collectivity are registered as voters so that they can participate in the
elections (as voters and as candidates) if they so choose. That is why the
registration phase is so important, and that is why it is so dismal that it is
often not taken seriously enough by politicians, electoral commissions,
and election administrators. Registration is an excellent opportunity to
start one's voter education campaign and to test the preparedness of the
election administrative apparatus, but the opportunity is not always fully
utilised. Examples are plentiful and clearly connected to the general
standard of population book-keeping in third world countries (Nepal 1991
is an example of a country with an interesting mixture of double- and
under-registration, while Zambia 1991 and particularly in 1996 offers a
standard example of a very modest level of registration quality). Kenya
1992 is another case in point,39
but the problem is certainly a much more
general one. Not to require proper voter registration, as happened in South
Africa in 1994, is not a solution to the problem of inadequate civil
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 45
registration. The South African debate during the first months of 1999
about how to achieve an acceptable level of registration before the 2 June
1999 elections illustrated well how a combination of political,
administrative, and financial problems can complicate an otherwise well
thought out plan.
• Systems for local public display of voters' registers and ample
opportunities for corrections are an integral part of an acceptable
registration system. Copies of the consolidated register should not be
made available to the incumbent party only - it is either all of them or
none of them!
Step 6: Nomination and Registration of Parties
• Parties should be free to organise, to assemble without being hindered by
bureaucratic orders, to register (if that is the part of the politicaladministrative
tradition of the country), and to nominate candidates of
their own choice, selected according to procedures decided by the party
itself according to its own decision-making procedures.41
Restrictions on
these rights are not conducive to a full-fledged democracy. An illustrative
example in this regard is the various restrictions on candidate nomination
in Chinese village committee elections.41
• Other restrictions are sometimes established, as the requirement in the
Nepalese electoral law that parties must field at least five per cent female
candidates in order to allowed participation in the election. The effect is
negligible as most of these candidates are allocated to seats unwinnable
for women, as in some constituencies in the Himalayas.
Step 7: Regulation of the Electoral Campaign
• Equitable access to public media for all registered parties is a must, even
though it is often violated.
• Public funding of political parties and/or candidates - generally or for
electoral campaign purposes only - is a controversial issue, in particular
because it might call forward parties and candidates who are only
participating because of the funding. As such funding is a means to 'level
the playing field' there should ideally be provisions for allocating
sufficient funds also to new and previously unrepresented parties, but that
is exactly what may attract less serious parties and candidates. It is
equally problematic to let previously represented parties allocate funds
only to themselves, since that might perpetuate the party system already
in place and deny new parties a fair chance of access.
• Spending limits on election campaign and related costs both for parties
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
46 DEMOCRATIZATION
and individual candidates is another controversial issue, particularly since
it is difficult to exercise control and auditing can only take place after the
election. Still, such a practice will encourage accountability and as
democracy develops, it will become part of a system of checks and
balances that will be increasingly integrated in the democratic culture. So
the long term impact will be more important that the short term impact.
Vote buying and related practices are a related issue which is even more
difficult to handle.
• The incumbent party must not be allowed to use government resources for
campaign purposes as has happened many times43
Step 8: Polling
• Rumours about election day fraud and misbehaviour are many and often
difficult to control. The presence of well-trained and dedicated polling
station staff is one prerequisite for achieving a reliable polling result. It is
probably correct - as suggested by Bratton44
- that election rigging does
not primarily taking place on polling day, which is another reason for
focusing on the previous steps in the process. However, it is not true that
violations never occur on election day - only that the occasion is less
important than what one would think on the basis of the many election
observer missions. However, newspaper reports about the Nigerian
elections in 1999 indicated clearly that observers were less than happy
with what they noticed.
Step 9: Counting
• The use of decentralised counting (that is at polling station level) is
preferable, because it removes problems related to transportation of
uncounted ballot papers, and it makes it easier for local electorates to
follow and understand the process. A positive consequence of this will be
more confidence-building at the local level when results are reported
accurately. The flip side of decentralised counting is obviously that it
makes it possible to sanction negatively polling districts (wards, villages)
who did not vote 'correctly' - or to use the threat of sanctions as in
Singapore 1997.45
Another argument against decentralised counting is that
it makes it easier in situations with extensive vote buying to check the
delivery of 'the goods'. Party agents and election observers should always
be allowed to follow the count. A further possibility is to combine
decentralised counting with a recount at the national level, to check
accuracy and other features, as in Venezuela.
Step 10: The Electoral Court System
• Provisions for an electoral court system which can handle electoral
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 47
complaints and disputes in an expeditious and impartial manner are
necessary. It can be argued that the number of electoral court cases is a
good indicator of the quality of the work of the electoral commission (and
the conditions they worked under).
Several other areas of concern could also be indicated, but that should not
be necessary. The list demonstrates convincingly that it is possible to point
to specific and important areas within the electoral institution which can be
used simultaneously to (1) evaluate the democratic intent and seriousness of
political and administrative elites in new democracies and (2) to indicate
areas where election administration and system improvement will be most
helpful, if asked for.
Conclusion
The electoral institutions matter for the democratization process and so does
electoral institutional change, be it improvement or the opposite.
A crucial factor when it comes to explaining the level of electoral quality
and the electoral institution's independent contribution to democratic
consolidation appears to be the democratic intent and seriousness of the
relevant political and administrative elites - in combination with respect for
the rule of law - while the amount of foreign financial support poured into
the election or the number of election advisors (not to speak of election
observers) is less important. The international community might be able to
counteract democratic setbacks and de-democratization, but further
democratization will normally have to come from within the country in
question.
Having a full-fledged liberal democracy - as well as elections of good
quality - presupposes respect for basic political rights and civil liberties, not
only formally and at election time, but in actual practice, year in and year
out. It also presupposes that real political competition is possible, and that
there is a vibrant civil society, that non-governmental organisations and
political parties are allowed to develop and flourish, that relevant
information is available through the media and so on. The possibility of full
and equal participation - and active and systematic efforts to ensure
inclusiveness - is another crucial element, the importance of which has
become evident in South Africa in the run-up to the June 1999 parliamentary
and provincial elections.
Actors, whether domestic or international, who are genuinely interested
in promoting a democratic development and a democratic culture in a
country which is not yet a full-fledged liberal democracy should therefore
concentrate their election-related efforts on those first steps in the electoral
process where the returns both in the short and the long run are highest and
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
48 DEMOCRATIZATION
most visible. And it can come as no surprise that actors with the opposite
interest actually disregard suggestions for improvement in the early phases
of the electoral process.
It is within the context of the preconditions stated above that the various
elements of the electoral institution matter, and that includes a truly
independent and impartial electoral commission, the character of the seat
allocation system, and the registration rules and their implementation.
All these elements should be of prime concern to anyone who would like
to see democracy develop and thrive in the third world. It might, however,
be difficult/impossible to achieve in countries where the political and
administrative elites are uninterested in taking the democratic route and
perhaps only do so superficially because that is the price for getting much
needed development assistance from the international community.
It has been argued that the early steps in the process - from legislating
the electoral process to regulating the electoral campaign - should in
particular attract our attention. If the provisions and procedures listed under
the first seven steps of the electoral process are implemented and adhered to
openly and responsibly, then it becomes less important to subject the
remaining elements to detailed observation and monitoring. There is
obviously a balance to strike, but the conclusion is that it is more conducive
to a worthwhile electoral process to focus on and support earlier rather than
later steps in the process.
If sufficient effort is put into ensuring the quality of the electoral process
before polling day, and if the election-related legislation is carried through
in a politically unbiased way, then qualitatively better elections will follow.
One must therefore focus on the legislative and administrative processes in
the early phases, and the budgetary and technical independence of the
electoral administration should be a main aim. The registration phase is also
more important than is often realised, and the same applies to the
strengthening of political parties, as they must be able to present to the
electorate viable alternatives to the incumbent party.
The conclusion in relation to election observation and monitoring is
similarly that such activities should also concentrate on the early phases of
the electoral process. Instead of sending election monitors and observers
(who in any case are usually too few and too ill equipped to pass a reliable
judgement on the quality of the election), it would be better to have expert
teams analyse the country's performance during the seven first steps of the
electoral process, while polling day observation could be left to locals, in
other words party agents and non-governmental organisations with a vested
interest in a credible outcome of the elections, dedicated to monitoring their
country's process towards a more democratic future.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 49
NOTES
1. Terry Karl, 'Imposing Consent? Electoralism versus Democratization in El Salvador', in
Paul Drake and Eduardo Silva (eds.), Elections and Democratization in Latin America,
1980-1985 (San Diego, CA: Center for Iberian and Latin American Studies/Center for
US/Mexican Studies, UCSD, 1986), pp.9-36; Philippe C. Schmitter and Terry Lynn Karl,
"What Democracy is ... and is Not', Journal of Democracy, Vol.2, No.3 (1991), pp.75-88.
2. Larry Diamond, 'Is the Third Wave Over?', Journal of Democracy, Vol.7, No.3 (1996),
pp.20-37.
3. Diamond, op. cit., p.21.
4. Diamond, op. cit., p.25; see also Jørgen Elklit, 'Is the Degree of Electoral Democracy
Measurable? Experiences from Bulgaria, Kenya, Latvia, Mongolia, and Nepal', in David
Beetham (ed.), Defining and Measuring Democracy (London: SAGE, 1994),pp.89-111.
5. Diamond, op. cit., p.25.
6. See for example Adam Karatnycky (ed.), Freedom in the World,The Annual Survey of the
Political Rights & Civil Liberties 1997-1998 (New Brunswick, NJ and London: Transaction
Publishers, 1998).
7. Karatnycky, op. cit., p.592ff.; Diamond, op. cit., pp.23-4.
8. Karatnycky, op. cit., p.3-4.
9. See Elklit, 'Is the Degree ... '; David Collier and Steven Levitsky, 'Democracy with
Adjectives: Conceptual Innovation in Comparative Research', World Politics, Vol.49, No.3
(1997), pp.430-51; for a more general discussion, and also Kenneth Bollen, 'Liberal
Democracy: Validity and Method Factors in Cross-National Measures', American Journal of
Political Science, Vol.37, No.4 (1993), pp.1207-30.
10. Diamond, op. cit., pp.30-31.
11. Fareed Zakaria, 'The Rise of Illiberal Democracy', Foreign Affairs, Vol.76, No.6 (1997),
pp.23-43; Adrian Karatnycky, 'The Decline of Illiberal Democracy', Journal of Democracy,
Vol.10, No.l (1999), pp.112-25, and especially p.116.
12. Diamond, op. cit., pp.23-5.
13. Diamond, op. cit., pp.33-4.
14. Andreas Schedler, 'What is Democratic Consolidation?', Journal of Democracy, Vol.9, No.2
(1998), p.100; Michael Bratton, 'Second Elections in Africa', Journal of Democracy, Vol.9,
No.3 (1998), pp.51-66; Michael Bratton and Nicolas van de Walle, Democratic Experiments
in Africa. Regime Transitions in Comparative Perspective (Cambridge: Cambridge
University Press, 1997), p.236.
15. Schedler, op. cit., pp.94-9; it should be remembered, however, that Schedler uses electoral
democracy as a broader concept than does Diamond.
16. Timothy D. Sisk, 'Elections and Conflict Management in Africa. Conclusions and
Recommendations', in Timothy D. Sisk and Andrew Reynolds (eds.), Elections and Conflict
Management in Africa (Washington, DC: United States Institute of Peace Press, 1998),
pp.148-75; Bratton, op. cit.
17. Arend Lijphart and Carlos H. Waisman (eds.), Institutional Design in New Democracies.
Eastern Europe and Latin America (Boulder, CO: Westview, 1996), p.2.
18. Scott Mainwaring, 'Transitions to Democracy and Democratic Consolidation: Theoretical
and Comparative Issues', in Scott Mainwaring, Guillermo O'Donnell and Samuel Valenzuela
(eds.), Issues in Democratic Consolidation; The New South American Democracies in
Comparative Perspective (Notre Dame, IN: University of Notre Dame Press, 1992),
pp.294-341.
19. Terry Lynn Karl,. 'The Hybrid Regimes of Central America', Journal of Democracy, Vol.6,
No.3 (1995), p.84.
20. Bratton, op. cit., pp.51-2.
21. Thomas Carothers, 'Democracy Without Illusions', Foreign Affairs, Vol.76, No.l (1997),
pp.85-99; Jeff Haynes, Democratic consolidation in the Third World;Many questions, any
answers?, paper for the ECPR workshop on 'Democracy in the Third World: What Should
Be Done?', University of Mannheim, 26-31 March 1999 (London: Department of Politics
and Modern History, London Guildhall University, 1999).
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
50 DEMOCRATIZATION
22. Jørgen Elklit and Palle Svensson, 'What Makes Elections Free and Fair?', Journal of
Democracy, Vol.8, No.3 (1997), pp.32-46.
23. Ibid.
24. Robert A. Dahl, 'Democracy and Human Rights under Different Conditions of
Development', in Asbjørn Eide and Bernt Hagtvet (eds.), Human Rights in Perspective. A
Global Assessment, Nobel Symposium 74 (Oxford: Blackwell, 1992), p.246.
25. Carothers, op. cit., p.92; Mainwaring, op. cit., pp.326-9; see also Doh Chull Shin, 'On the
Third Wave of Democratization. A Synthesis and Evaluation of Recent theory and
Research', World Politics, Vol.47 (1994), pp.153-4 and Peter Burnell, 'Arrivals and
Departures: A Preliminary Classification of Democratic Failures and their Explanation',
Commonwealth and Comparative Politics, Vol.36, No.3 (1998), p.17.
26. Bratton, op. cit.
27. Elklit, 'Is the Degree ... '; Elklit and Svensson, op. cit.
28. Or 'rules', see Rein Taagepera, 'How Electoral Systems Matter for Democratization',
Democratization, Vol.5, No. 3 (1998), pp.68-91.
29. Elklit & Svensson, op. cit., Bratton, op. cit., p.62.
30. Burnell, 'Arrivals and Departures ... '.
31. Patrick Merloe, Lessons Learned and Challenges Facing International Election Monitoring,
paper (Washington, DC: National Democratic Institute for International Affairs, 1999).
32. Among the many available analyses and descriptions of such elections, apart from literature
already mentioned in these end notes, use has also been made of the following: Peter Burnell,
'Taking Stock of Democracy in Zambia', paper for ECPR workshop on 'Democracy in the
Third World: What Should be Done?', Mannheim, 26-31 March 1999; Yonhyok Choe, 'How
to Manage Free and Fair Elections. A Comparison of Korea, Sweden and the United
Kingdom', Gothenburg Studies in Politics 44 (Gothenburg: Department of Political Science,
University of Gothenburg, 1997); Staffan Darnolf, 'Democratic Electioneering in Southern
Africa. The Contrasting Cases of Botswana and Zimbabwe', Gothenburg Studies in Politics
45 (Gothenburg: Department of Political Science, University of Gothenburg, 1997); Election
Observation Centre, Kenya General Elections 1997. Final Report, For Donors' Democratic
Development Group (Nairobi, Jan. 1998); John D. Holm, 'Elections in Botswana:
Institutionalization of a New System of Legitimacy', in Fred M. Hayward (ed.), Elections in
Independent Africa (Boulder, CO and London: Westview Press, 1987); Krishna Kumar (ed.),
Postconflict Elections, Democratization, and International Assistance (Boulder, CO and
London: Lynne Rienner Publishers, 1998), pp.121-47; Samuel S. Mushi and Rwekaza S.
Mukandala (eds.), Multiparty Democracy in Transition. Tanzania's 1995 General Elections
(Dar es Salaam: Department of Political Science and Public Administration, University of
Dar es Salaam, 1997); NEC, The Report of the National Electoral Commission on the 1995
Presidential and Parliamentary Elections (Dar-es-Salaam: NEC, 1997); Palle Svensson
'Prospects for Free and Fair Elections in Kenya 1997-98' (unpublished paper, Aarhus, Aug.
1997). Aili Mati Tripp, 'Political Reform in Tanzania: The Struggle for Associational
Autonomy', paper presented at ECPR Workshop 'Democracy in the Third World: What
Should be done?', Mannheim, 26-31 March 1999 (Madison, WI: Department of Political
Science, University of Wisconsin-Madison, 1999 - a version of the paper is forthcoming in
Comparative Politics); John A. Wisemann, 'Early Post-redemocratization Elections in
Africa', Electoral Studies, Vol.11, No.4 (1992), pp.279-91; John A. Wiseman, 'The Slow
Evolution of the Party System in Botswana', Journal of Asian and African Studies,
Vol.XXXIII, No.3 (1998), pp.241-64.
33. Guiseppe Di Palma, To Craft Democracies. An Essay on Democratic Transition (Berkeley,
CA: University of California Press, 1990).
34. William C. Kimberling, 'A Rational Approach to Evaluating Voter Registration', in John C.
Courtney (ed.), Registering Voters.Comparative Perspectives (Harvard, MA: The Center for
International Affairs, 1991), pp.3-11.
35. Patrick F.J. Macrory, Jørgen Elklit, and Rafael Sabat Mendez, Facing the Pluralist
Challenge. Human Rights end Democratization in Kenya's December 1992 Multi-Party
Elections (Washington, DC: International Human Rights Law Group, 1992), pp. 16-17.
36. Andrew Reynolds and Ben Reilly, with Kader Asmal, Sarah Birch, John Carey, Mona
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014
ELECTORAL INSTITUTIONAL CHANGE AND DEMOCRATIZATION 51
Makram Ebeid, Jørgen Elklit, et al, The International IDEA Handbook of Electoral System
Design (Stockholm: International Institute for Democracy and Electoral Assistance, 1997).
37. Timothy D. Sisk and Andrew Reynolds (eds.), Elections and Conflict Management in Africa
(Washington, DC: United States Institute of Peace Press, 1998); Joel D. Barkan, 'Elections
in Agrarian Societies', Journal of Democracy, Vol.6, No.4 (1995), pp.106-16; Jørgen Elklit
(ed.), Electoral Systems for Emerging Democracies: Experiences and Suggestions
(Copenhagen: Ministry of Foreign Affairs/DANIDA, 1997); Donald L. Horowitz, A
Democratic South Africa: Constitutional Engineering in a Divided Society (Berkeley, CA:
University of California Press, 1991); Andrew Reynolds, "The Case for Proportionality',
Journal of Democracy, Vol.6, No. 4 (1995), pp.117-24; Andrew Reynolds, Electoral
Systems and Democratization in Southern Africa (Oxford: Oxford University Press, 1999).
38. Peter Harris, 'Building an Electoral Administration', in Peter Harris and Ben Reilly (eds.),
Democracy and Deep-Rooted Conflict: Options for Negotiators (Stockholm: International
IDEA, 1998), pp. 308-19.
39. As in Kenya 1992: Macrory, Elklit, and Mendez, op. cit.; Joel D. Barkan, 'Kenya: Lessons
from a Flawed Election', Journal of Democracy, Vol.4 (1993), pp.85-99.
40. Macrory, Elklit, and Mendez, op. cit., pp. 22-4; Barkan, op. cit., p.93.
41. Barkan, op. cit., p.94.
42. Jørgen Elklit, 'The Chinese Village Committee Electoral System', Chinese Information,
Vol.XI, No.4 (1997), pp.1-13.
43. Richard Sandbrook, Transitions without Consolidation: Democratization in Six African
Cases', Third World Quarterly, Vol.17, No.l (1996), pp.69-87.
44. Bratton, op. cit., p.62.
45. Jinshan Li and Jørgen Elklit, "The Singapore General Election 1997: Campaigning Strategy,
Results, and Analysis', Electoral Studies, Vol.18, No.2 (1999), pp.199-216.
Downloadedby[UniversityofCalifornia,SanFrancisco]at12:2301September2014