The Rise of Election Monitoring
WHAT MAKES ELECTIONS
FREE AND FAIR?
Jergen Elklit & Palle Svensson
Jergen Elklit is associate professor of political science at the University of Aarhus, Denmark. He has served as an advisor on electoral systems and democratization and as, an election monitor in Africa, Asia, and Europe. In 1994 he was a member of the South 4fricän Independent Electoral Commission. PalIe Svensson is associate professor of political science at the University of Aarhus and has advised on and monitored elections in African countries. An earlier version of this essay was presented at the Nordic Political Science Association’s Triennial in Helsinki in August 1996.
It was late in the afternoon in Kampala on 31 March 1994. Journalists were waiting impatiently for an announcement from international election observers. United nations officials stated for the third time their argument that the observers. Should declare the March 28 elections for Uganda’s Constituent Assembly “free and fair~” But the election observers avoided that phrase. They had monitored only part of the electoral process; moreover, they knew that calling the elections “free and fair” would hinder or preclude discussion of the problems they had discovered. In the end, the elections, in Uganda – which were no worse than many other elections that have taken place in emerging democracies—were not declared “free and fair.”
As this incident shows, election observers encounter great pressure - and not just from overeager journalists – to judge whether the elections in question were “free and fair,” Indeed, sometimes it seem that this is all people want to know. “Free and fair” has become the catchphrase of UN officials, journalists, politicians, and political scientists alike. It exemplifies what Giovanni Sartori once called “conceptual stretching.” “The wider the world under investigation, the more we need conceptual tools that are able to travel.” But what actually constitutes a “free and fair” election? Does the phrase mean only that the election was “acceptable,” or does it imply something more?
International organizations have long been involved in monitoring and assessing elections and referendums. Especially notable has been the UN’s role in referendums on independence, which began to take place in the late 1950s. Before the UN could recognize former colonies and trust territories as independent states, it had to know whether these votes had been “free and fair.”2 This concept supposedly made its first appearance in a report on Togoland’s 1956 independence referendum.3
The UN’s involvement in the November 1989 referendum in Namibia was fundamentally different: In that case the vote was not just an element of the colony’s long liberation process but also an integral part of the UN peacekeeping efforts in the area. In February 1990, the UN supervised presidential and legislative elections in Nicaragua. Interestingly, this was done at the request of the country itself, and as part of an assessment of the entire electoral process, not just of election-day events. Thus the UN acquired a major role in the electoral process of an independent member country—something that not all UN members saw as a positive development.
Subsequent elections and, referendums in which the UN has been directly involved, either as part of peacekeeping efforts or because the countries in question .sought its approval include those in Haiti (December 1990), Angola (September 1992), Cambodia (May 1993), and Mozambique (October 1994). One might add to that list Eritrea (April 1993), South Africa (April ‘1994), and Malawi (June 1993 and May 1994), though the UN’s involvement in these cases was less extensive and due in part to other factors.4
Besides the civil war – torn countries noted above, many other nations have taken dramatic steps toward democracy during the past decade. In many cases, individual Western countries have provided support these developments, in other cases, the primary actors have been international organizations other than the, UN (especially the Organization of American States, the European Union, the Organization for Security and Cooperation in Europe, the Inter-Parliamentary Union, and Commonwealth Secretariat). Both national and international governmental organizations (NGOs) have also become involved. Many, of these NGOs have received substantial funding from governments other public sources.5
Over the past decade, countless election ., observers have been dispatched to every region of the globe. This increased activity has been accompanied by an intensified demand for standardized assessment criteria, but the development of “checklists” has been hindered by disagreement over what should be included. In addition, cooperation among different countries organizations and election authorities has been uneven. Thus a discussion of the basis on which an election or referendum can be labeled “free and fair,” or at least “acceptable,” long overdue.
Although criteria for declaring an election “free and fair” have been developed in various contexts, translating such theoretical concepts into a comprehensive list of factors to consider has proved difficult. Equally daunting are the methodological problems of determining whether a particular electoral process meets the established criteria and combining the different “measurements” on various dimensions into a single score. One approach is to study various aspects of the process (e.g., the electoral system, the voter-registration system, media access, campaign rules, ballot counting) and then assess whether conditions within each area promote or hamper the freedom and fairness of the election.’ Here we take a different approach. Drawing on the work of Robert Dahl, we start by examining the relationship between elections and. Democratic development. This provides a basis for defining the concepts “free” and fair.” We then present a list of assessment criteria and examine the value of such a list in actual practice.
It is not surprising that politicians and voters in formerly colonized or nondemocratic countries - as well as individuals, countries, and international organizations that subscribe to democratic principles – take a great interest in elections and referendums. Yet this has contributed to development of a distorted picture of the process of democratic transition: The poll itself has become the focus of attention, acquiring an importance that has no basis in either democratic theory or practical politics.
A common misperception is that any country in which elections have been held without too many obvious irregularities can be called a “democracy.” This attitude has been most easily discerned in U.S. policy toward some South and Central American countries, but it can also be identified in the foreign policy of other nations.’ Yet if the ultimate objective is to encourage continuous development toward a well functioning democracy, the prerequisites of democratic elections must not be ignored.
Robert Dahl has identified a number of “institutional prerequisites” of democracy. One of these is free and fair elections. Yet Dahl does not indicate what he means by “free and fair,” other than that “elected officials are chosen in frequent and fairly conducted elections in which coercion is comparatively uncommon.”11 This leaves several questions unanswered: Can elections be free and fair if part of the adult population has no right to vote? What if not all adult citizens have the opportunity to run for office? And what if there is no freedom of speech, assembly, or movement? In other words, do “free and fair” elections not require the fulfillment of all of Dahl’s other prerequisites of democracy? Dahl himself has argued that elections should be held later in the democratic-transition process than has often been the case. This implies that a number of other preconditions of democracy must be met, at least in part, before free and’ fair elections can be held.’2
The concepts “free” and “fair,” then, must be clearly defined and distinguished from other preconditions of democracy. They must also be translated into specific criteria that can be used to evaluate elections. The electoral process itself must be broken down into its component phases, each of which corresponds with certain evaluation criteria.
Freedom, as Dahl notes, contrasts with coercion. ‘Freedom entails the right and the opportunity to choose one thing over another. Coercion implies the absence of choice, either formally or in reality: either- all options but one are disallowed, or certain choices would have negative consequences for one’s own or one’s family’s safety, welfare, or dignity.
Fairness means impartiality. The opposite of fairness is unequal ‘treatment of equals, whereby some people (or groups) are given unreasonable advantages. Thus’ fairness involves both regularity (the unbiased application of rules) and “reasonableness (the not-too-unequal distribution of relevant resources among competitors).
In practice, it is not always easy to distinguish between freedom and fairness, and any categorization of various elements of the electoral process should be approached with caution. In general, however, the “freedom” dimension should include elements relating to voters’ opportunity to participate in the election without coercion or restrictions of any kind (with the ‘possible exception of economic limitations). Thus “freedom” primarily deals with the “rules of the game.”
Which of the two dimensions is more important? ‘We hold that freedom must be given priority, because it is a precondition for democracy and for elections as a means to that end: Without rules granting formal political freedoms, the question of the fair application of rules is meaningless, and the question of equality of resources irrelevant.
It may be just as difficult to distinguish between the two aspects of fairness as between freedom and fairness, because both regularity and reasonableness involve the notion of impartiality. Yet they should be separated to the extent possible. Regularity—which involves impartial application of the election law, constitutional provisions, and other regulations - is the more specific of the two and must be present in a high degree before an election can be accepted. Reasonableness - which concerns securing roughly equal opportunities for the exercise of political freedom - is more general, and is impossible to achieve in full or even to a high degree. In fact, we know of no democracy that has distributed relevant political resources equally among political competitors.
Thus in assessing the fairness of an election, or a referendum, it is more important to discern how the rules are applied than to determine whether individuals and groups have ideal opportunities. This does not mean that reasonableness is unimportant. Indeed, a broadening of access to relevant political resources and opportunities is a clear indication of a regime’s progress toward democracy. In competitive elections, the opportunities available to various groups are especially important. There should be no question of any particular group or political party having a greater chance of winning the election than any other group. The standard formulation used in the preparatory phase of the April 1994 elections in South Africa—the notion of “leveling the playing field”—epitomizes this aspect of “fairness.”’3
In addition to clarifying core concepts, it is important to distinguish among events before, during, and after the actual polling. The election day itself is only part of the electoral process; thus observation missions consisting of short stays around election day are fundamentally flawed. The pre-election period is especially important: it is at this stage that observers must assess whether the election law and the constitution guarantee the freedom of the voters, and verify that relevant resources are not too unequally distributed among competing parties and candidates. The importance of evaluating pre-election day events has been increasingly acknowledged by the UN and other organizations, particularly since the Nicaraguan operation of 1990.
The period after the actual polling must also be considered. At this stage, the crucial issue is the fair and regular application of rules. The counting of ballots should be carefully controlled to prevent fraud, the
results should be reported immediately, and complaints about the electoral process should be handled impartially.
Combining the two principal dimensions of election assessment with the three observational phases yields the checklist presented in Table I. The checklist is a useful device, but it is not without problems. Although it is based both on relevant literature’ and on practical experience, it is “certainly not exhaustive. Moreover it represents a
schematic outline of the assessment process, not a detailed and unambiguous set of instructions. This gives rise to a twofold problem of reliability. First, election observers may disagree on the extent to which the individual criteria have been fulfilled. Second, the list does not indicate the relative importance of the various criteria, If some criteria have been fulfilled, while others have not, or if a certain criterion has been fulfilled only partially,- the observers must rely heavily on their own judgment. Nevertheless, some general guidelines can be provided.
Although the election law (and related regulations) of the country in question may not be ideal in the eyes of international election observers, the observers’ main duty is to determine whether or not the electoral.
Table I — Checklist for Election Assessment
Freedom of movement A transparent electoral process
Freedom of speech (for candidates, the media, An election act and an electoral system
voters, and others) that grant no special privileges to any Freedom of assembly political party or social group
Freedom of association Absence of impediments to inclusion
Freedom from fear in connection with the election in the electoral register and the electoral campaign Establishment of an independent
Absence of impediments to standing for election and impartial election commission
(for both political parties and independent candidates) Impartial treatment of candidates by
Equal and universal suffrage the police, the army, and the courts of law Equal opportunities for political parties
and independent candidates to stand for
Impartial voter.education programs
An orderly election campaign (observance
of a code of conduct)
Equal access to publicly controlled media
Impartial allotment of public funds to
political parties (if relevant)
No misuse of government facilities
for campaign purposes
On Polling Day Opportunities to participate in the election Access to all polling stations for
representatives of the political parties,
accredited local and international
election observers. and the media
Secrecy of the ballot
Absence of intimidation of voters
Effective design of ballot papers
Proper ballot boxes
Impartial assistance to voters (if necessary)
Proper counting procedures
Proper treatment of void ballot papers
Proper precautionary measures when
transporting election material
Impartial protection of polling stations
After Polling Day Legal Possibilities of complaint Official and expeditious announcement of
Impartial treatment of any election
Impartial reports on the election
results by the media
Acceptance of the election results by
process conforms to that law. The checklist in Table 1 should prove useful in this regard: It can help observers keep track of the various components of the electoral process and compare them against the election law (with which the observers should be thoroughly familiar).
0f course, electoral processes differ, as do political and social conditions and democratic-transition processes themselves. Thus it is not possible to attach absolute weights to the various criteria listed in Table1, for their importance varies with the electoral context. In general, however items in the “free” column are more important than those in the “fair” column, and within the “fair” column, correct and impartial application of the election law and other relevant regulations is more important than ideal opportunities for political competition. Whereas freedom is a necessary - though not sufficient – condition for an election’s acceptability, the combination of freedom and the fair application of electoral rules is both necessary and sufficient for such acceptability. For an election to be free and fair, however, the main competitors should have had at least some access to campaign resources and the media, even if that’ access was not fully equal.
What happens before and after polling day is at least as important as what happens on polling day itself. In particular, the observance of political freedoms in the pre-election period is a prerequisite for the acceptability of an election, and even for the mounting of an electoral ‘observation. If these rules are broken, the election cannot be declared acceptable, much less free and fair. After the election, all that is required is voluntary acceptance of the outcome by all serious political contenders. If the results are disputed, it is essential to analyze the
reasons for the disagreement and to observe how complaints are treated.
Of course what happens on pooling day is also important. Yet election day activities are often overemphasized; moreover, they are not always reported adequately. It is not enough merely to observe and report irregularities; rather, they must be evaluated in relation to reasonable expectations. What matters is how widespread they are, how serious they are, whether they represent a clear tendency (especially in favor of current officeholders), and how significant they are in affecting the final results.
To be sure, irregularities should be noted, and suggestions for remedying them should be given. But irregularities that are a result of deficits in technical capacity or experience are less serious than deliberate attempts to manipulate the results. In fact, irregularities on polling day seriously threaten freedom and fairness only to the extent that they are extensive, systematic, or decisive in a close race.
In addition to determining which criteria have been fulfilled and deciding on the relative importance of the various items observers should also judge whether the election or referendum under the given circumstances reflects the expressed will of the people: This is, after all, the main reason for conducting elections, and irregularities and technical problems should be assessed from this perspective.
Observers should also evaluate the ‘election in the context of the specific democratic-transition process. Will the election despite possible technical shortcomings—stimulate further democratization by increasing respect for political freedoms, strengthening adherence to the election law, enhancing political contestation through broader access to relevant resources, involving more people in the political process, or improving the quality of the political debate? Although some would categorize this as a “political” judgment, it can be argued that it legitimately falls within the domain of election observation. If observers are to view an election not as an isolated event but as part •of the democratization process, they cannot avoid considering whether and how it contributes to that process.
It should be emphasized, however, that while observers may go beyond a narrow technical assessment of elections to evaluate the degree to which the preferences of the electorate have been expressed and the role that the election played in the democratization process’: they still do’ not have license to pass judgments of a broader nature? An election should not be deemed acceptable because it contributes to political stability or law and order in the country or the region. Such judgments may be both relevant and expedient, but it is not the role of election observers to make them. All they can do is deliver relevant information about the electoral component of the overall situation: it is up to national governments and international ‘ bodies to draw the political conclusions.
It is not easy to establish the precise line between legitimate an illegitimate assessments; moreover, not all election observers acknowledge the distinction is readily, apparent when a delegation of election observers include a number of parliamentarians, who are used to making political judgments and willing to take responsibility for them. Such observers often refuse to accept the inherent limitations of their role. Asimilar problem of demarcation arises when international organizations or national governments have difficulty separating the electoral assessment from an assessment from an analysis of the political consequences of that assessment. Again, these are two different kinds of activities, which should be carried out by different organizations.
A few examples will illustrate some of the problems of election observation and assessment mentioned above. The June 1992 parliamentary elections in Mongolia were praiseworthy at east one respect: the only slightly reformed, former communist Mongolian People’s Revolutionary Party (MPRP) had introduced a by-and-large exemplary election law, which was observed in all essentials. Yet the MPRP had also installed an electoral system that was perhaps the most undemocratic in the world (majority elections in multimember constituencies, with each voter required to cast exactly the same number of votes as the number of parliamentary seats to be filled by the constituency). Predictably, the MPRP won 93 percent of the available seats with only 57 percent of the vote. Of course, 57 percent is a clear majority. Yet this left more than 40 percent of the electorate virtually without parliamentary representation at a time when their country’s social and political systems were being totally reformed. What does this mean for the election’s status as “free and fair”? Can an election conducted under such a system even be termed “acceptable”? Fortunately, the electoral system was replaced with an ordinary first-past-the-post system for the June 1996 parliamentary elections.
In the case of Kenya’s December 1992 presidential, parliamentary, and local elections, many elements of the electoral process were questionable.’6 There was considerable evidence of manipulation on the part of President Daniel arap Moi and his party, the Kenya African National Unity (KANU). Yet the situation appeared to improve as polling day approached, resulting in a relatively orderly vote, all things considered. The poor showing of the opposition parties was due largely to their own failure to work together and only in part to the various tricks of KANU, the chairman of the electoral commission, and others. Can such elections be termed “free and fair”?’7 To what extent can a degree of progress toward democracy compensate for irregularities and disregard of the rules?
In elections for Uganda’s Constituent Assembly in March 1994, political parties—which had been associated with the country’s bloody ethnic clashes—were forbidden to field candidates, while individual candidates were given carte blanche, a decision that provoked considerable dissatisfaction. Moreover, a serious voter-registration problem arose. The plan was to complete the voting in one day and count the ballots before dark: The voters were therefore distributed among polling stations of no more than six hundred electors. Technical difficulties, however, prevented the voter lists from being published; consequently, people did not know which polling station to go to. This resulted in a good deal of confusion on election day. Does the exclusion of political ‘parties from the electoral process preclude an election from being “free and fair?” And do technical problems with voter registration render an election unacceptable?
In South Africa’s April 1994 elections for parliament and regional assemblies, considerable efforts were made to involve all citizens and parties in the process of democratization and reconciliation. Yet it was difficult to ensure equality of opportunity for the country’s many different political formations and social groups. There were plenty of
administrative and procedural problems as well. What are we to make of the dual character of this particular electoral process? Was it appropriate that the South African Independent Electoral Commission (IEC) itself issued the verdict of “substantially free and fair”?
In Tanzania’s October 1994 local elections, only about half the electorate registered to vote—a disappointing figure compared with those of previous local elections. On the other hand, candidates from parties other than the autocratic Revolutionary Party of Tanzania (CCM) were nominated, and some were actually elected. Yet because the full election result~ were not reported, it was impossible to know the extent to which the opposition had succeeded in wresting local-council seats from the CCM, and how strong the opposition parties were nationally. It is difficult to regard such an election as “free and fair,” but might it not pass as “acceptable?”
Making Analytical Distinctions
These and other cases make it clear that it is difficult—perhaps impossible—to establish precise guidelines ‘ for assessing elections wherever they occur. Nevertheless, it is possible to establish some analytical distinctions.
If we consider the two main dimensions of freedom and fairness, it seems evident that some elections can be characterized as free and fair, even if they are not perfect. All or almost all elections in well-established democracies presumably fit this description. In these cases, however, there is no perceived need to invite international observers, making this first category mainly of academic interest.
It is also evident that some ejections are not “free and fair’~ owing to the violation of a large number of key criteria. In many cases the countries conducting such elections do not even bother to invite international observers, who would in any case be unlikely to come, for fear” of being seen as endorsing the elections. In other cases, observers may declare the election not free and fair. Albania’s parliamentary balloting of May 1996 is a recent case in point. Most village-committee elect ion in the People’s Republic of China also fall into this category. Between the extremes, however, lie many cases in which elections cannot be labeled ‘free and fair” because of any number of short-comings, but in which it would be unreasonable explicitly to declare them not “free and fair.” Perhaps they are free in a formal sense, but fairness is limited in practice—or perhaps they are free only to some extent, but rather fairly conducted within those particular limits.
When observers take circumstances into account and adopt a broad view – especially in terms of the possibility of progress towards greater political competition and participation - they may deem such elections “acceptable” even if they fall short of being “free and fair.” These are not only the most difficult cases to assess, but also the ones that international observers are most likely to witness: The governments in these countries are eager to obtain the international community’s stamp of approval as a means of boosting their internal legitimacy and gaining / external recognition.
In practice, then, elections and referendums are most likely to fall within the shaded area (between curves a and c) of Figure 1, representing balloting that is neither clearly free and fair nor clearly not free and fair, but acceptable when technical limitations and prospects for progress toward democracy are taken into account. In these cases, analysis of the application of a country’s election law should take into account not only the criteria listed in Table 1, but other factors that may help observers determine how strictly those criteria should be applied. For example, if the election is the first in a country undergoing a transition from authoritarianism to democracy, a relatively loose application of the criteria may be indicated (curve a). On the other hand, if there is reason to believe that the authorities have rigged the election, a stricter application is justified (curve b). Even under more favorable conditions, it is neither reasonable nor methodologically feasible to insist on complete fulfillment of all the criteria before declaring an election or a referendum “free and fair.” Something less (curve c) may suffice, as long as basic freedoms exist, the election law is for the most part applied impartially, and the main competitors have reasonable access to campaign resources and the media.
A B C
- Free +
Given the prevalence of elections and referendums that fall within the shaded area (between curves a and c), these special considerations deserve close attention. Because of the difficulty of distinguishing clearly between the dimensions of “free” and “fair,” assessments of such ballots can be represented by a straight line connecting the two most extreme points (the diagonal line in Figure 1). Thus the crucial point is where the diagonal leaves the shaded area (in a southwestern direction): it is here that we find the threshold between “acceptable” and “not acceptable.”
In determining the acceptability of a given election, observers should focus on the degree of conformance to the country’s election law and related regulations, considering not just election-day event but also the periods before and after the balloting. Also important are whether the preferences of the voters have been adequately expressed and what contribution the electoral process has made to the overall process of –democratization.
Yet another issue to consider is the closeness of the electoral race. Should observers be more permissive in cases in which the winner triumphs by a wide margin, possibly indicating that irregularities would not have affected the outcome? This argument, while plausible, is nevertheless problematic. Even huge wins can be manufactured moreover, the distinction between wide and narrow margins is subjective. Finally, given the importance of elections as learning opportunities for voters as well as for candidates, political parties, and election authorities, a strong case can be made for not setting electoral standards too low in newly emerging democracies.
A Complex Process
The phrase “free and fair” cannot denote compliance with a fixed, universal standard of electoral competition: No such standard exists, ‘and the complexity of the electoral process makes the notion of any simple formula unrealistic. Election observation requires the simultaneous use of multiple scales to achieve valid and reliable measurements of complex phenomena. These problems especially affect the large segment of elections that are neither clearly free and fair nor clearly not free and fair.
Election observers, therefore, face a dilemma. They might simply avoid using the phrase “free and fair,” but at the cost of opening the door to its use by others who have less knowledge or understanding of the situation. Alternatively, observers can elect to use the phrase as convenient shorthand, but at the cost of exposing themselves to all manner of criticism grounded in intellectual or moral considerations.
This does not mean, however, that election observation and assessment are hopeless tasks; it is indeed possible to draw general conclusions about how best to conduct such activities. In the borderline cases described above—the crucial “in-between” category—observers should identify their evaluation criteria as clearly as possible while at the same time acknowledging that, their conclusions rest to some degree on estimates and subjective judgments. In arriving at a conclusion it is vital that they consider whether electoral competition shows qualitative improvements over previous elections (especially in terms of the freedom dimension and the regulatory aspect of the fairness dimension) of course, it goes without saying that the course of events should reflect the preferences of the electorate.
The fulfillment of this last criterion is one reason for the widespread acceptance of the IEC’s designation of the April 1994 elections in South Africa as “substantially free and fair.”20 This factor is also emphasized by Guy Goodwin-Gill, who tries to strike a balance between explicit standards (which he claims do indeed exist) and what he calls “the special conditions of the general situation”—that is, whether the elections reflect the will of the people and are conducted in a positive atmosphere.21
The decision to declare an election “acceptable” also involves the election observer’s willingness to enter into a dialogue, not with the political authorities of the country in question, but with himself as to the possibilities of the situation. If the election, despite irregularities, seems to reflect progress toward a more democratic government, observers may choose to give it the benefit of the doubt. This may displease more moralistic observers who take a black-and-white view of the world, but so be it.
This essay has sought to develop a checklist with greater practical application than the ones that have emerged so far. Yet we do not consider it possible to develop a set of guidelines that is equally applicable to all elections and referendums in emerging democracies. The September 1996 elections in Bosnia-Herzegovina provide a good illustration of the complexity of the electoral process and the difficulty of rendering a final judgment. Before polling day, many freedoms were grossly violated, both in the Bosnian Federation and in Republika Srpska. Among them was the freedom of movement, which is usually taken for granted and therefore not included on checklists of assessment criteria. The election authorities were, of course, aware of this state of affairs, but decided to proceed cautiously in order to avoid jeopardizing the entire electoral process, which was seen as crucial to achieving peace and stability in the long term. Other major pre-election problems concerned voter registration, freedom of speech, media access, and intimidation of voters. On polling day itself, local reports of orderly elections abounded; at the counting stage, however, rumors of fraud in some areas began to circulate.
Questions can also be raised about the situation after polling day, especially with regard to the electoral authorities’ willingness to address complaints and the acceptance of the election results. Of course, the problems that arose in the latter area may have been due to some people’s refusal to acknowledge the legitimacy of the new political institutions.
— On the whole, then, the evidence would seem to indicate that the elections in Bosnia-Herzegovina were not free and fair, or even acceptable. Yet they were accepted by the international community owing to their presumed importance for stability and peace in the region. The lesson is clear: Elections vary so much from one case to another, with new and complicated political situations constantly arising, that previous observation experiences can provide only limited help.
— Despite the rapid growth of election observation over the past decade, the task of establishing criteria for evaluating elections still has a long way to go.
I.Giovanni Sartort, Concept Mtxformation in Comparative Politic\, Political Science Review 64 (December 1970): 1034.
2.A good overview of such efforts is provided in Yves Beigbeder, International Moni. taring of Plebiscites, Referenda and National Elections: Self-Determination and Transition to Democracy (Dordrecht, the Netherlands: Martinus Nijhoff. 1994), 94—143.
3.Jon M. Ebersole, “The United Nations Response to Requests for Assistance in Electoral Matters,” Virginia Journal of International Law 13 (Fall 1992): 94. See also Beigbeder, International Monitoring of Plebiscites. Referenda and National Elections.
4.Larry Garber and Clark Gibson, Review of United Nations Electoral 1992—93 (Project INT/9l/033. United Nations Development Programme, 18 August 1993).
5 Ebercole. “United Nations Response”: Jennifer McCoy, Larry Garber, and Robert Pastor Pollwatching and Peacemaking.” Journal of Democracy 2 (Fall 1991): 02—14.
6. See also Lessons Learnt—International Election Observation. Seventeen Organizations Share Experiences on Electoral Observation (Stockholm: International Institute for Democracy and Electoral Assistance, 1995), 13.
7. Jergen Eiklit, “Is the Degree of Electoral Democratization Measurable? Experiences from Bulgaria, Kenya. Latvia, Mongolia and Nepal,” in David Bcetham, ed,. Defining and Measuring Democracy (London: Sage, 1994). 89—Ill.
8. Guy S. Goodwin-Gill, Free and Fair Elections: International Law and Practice
(Geneva: Inter-Parliamentary Union. 1994), 27—80: William C. Kimhcrling, A Rational Approach to Evaluating Voter Registration. in John C. Courtney. cd.. Registering Voters. Comparative Perspectives (Cambridge, Mass.: Center for International Affairs. Harvard
University, 1991). 3—Il.
9. A good discussion of this point is round in Georg Srnensen. Democracy and Democratization: Processes and Prospects in a Changing World (Boulder, Cola.: Westview. 1993).
10. Philippe C. Schmitter and Terry Lynn Karl, ‘What Democracy Is ... and Is Not, Journal of Democracy 2 (Summer 1991): 75—88.
II. Robert A. Dahl. Democracy and Its Critics (New Haven: Yale University Press, 1989), 221. DahI actually refers not to “democracies” but to polyarchics. We use the term “democracy” here both for the ideology and for existing political regimes in order to conform to ordinary usage.