Look at these foreign observers. What they see is only the surface; they don’t know anything about our country.
—Nepalese voter outside a polling station, 20081
DESPITE CONTENTIOUS DEBATE over the years about whether it is putting the cart before the horse,2 the international community continues to push countries to hold elections as a way to promote freedom and democracy. Indeed, international election monitoring has become the primary tool of democracy promotion.3 Today diverse organizations flock to observe elections all over the world and broadcast their findings to the domestic and international communities. These efforts have become a true growth industry, involving global and regional intergovernmental organizations as well as nongovernmental agencies and organizations (Figure 1.1). Given that countries have traditionally guarded elections as a strictly domestic affair and a sacred hallmark of sovereignty, the rapid expansion of monitoring is stunning.
International monitors often play central roles in election dramas. Consider Georgia, where in 2003 denouncement of election fraud by international and domestic monitors helped trigger the Rose Revolution.4 Four years later, President Mikheil Saakashvili responded to sudden political riots by calling a presidential election for early 2008. To boost votes in the first round and prevent opposition voters from uniting against him in a runoff, he combined the implementation of social welfare programs with campaigning, stacked the central election commission (CEC) with partisan members, and occasionally used intimidation and pressure.5 The international community feared further instability. The West was pulling for Saakashvili, Russia for the opposition, leaving the election observers in a difficult and prominent position. The Financial Times noted on the eve of the vote: “Pressure is mounting on more than 1,000 international observers who will play the key role in deciding the legitimacy of votes cast at some 3,400 ballot stations.”6
Yet despite the sweeping prevalence of international monitors, global political developments are unsettling: After 2005, the democratic gains of the past two decades have stagnated, perhaps even begun to recede. In 2009, the year marking the twentieth anniversary of the fall of the Berlin Wall, freedom declined in no less than forty countries. This was the fourth consecutive year that declines trumped gains and the longest continuous period of deterioration in the forty years of reporting by Freedom House, the independent watchdog organization. The downward trend continued in 2010. With backsliding in Honduras, Madagascar, Mexico, Mozambique, Niger, Ukraine, and several others countries, by 2010 the number of what Freedom House calls “electoral democracies” dropped to 115—its lowest level since 1995.7 It remains to be seen whether the Arab Spring will bring any relief at all to this downward slide. The elections in Kyrgyzstan in 2005, Pakistan in 2008, and Afghanistan in 2009 and 2010, among many others, were all monitored by international observers, yet these elections made it painfully obvious that elections cannot be equated with democracy and, furthermore, that simply holding an election does not ensure progress toward democracy, even if international actors invest heavily in monitoring it.
Figure 1.1: Number of national-level election missions per year, 1975–2004
So is international election observation a good idea? Is it worth all the effort put into it? Does it actually promote democracy by strengthening elections? It would be naïve to expect all monitoring efforts to succeed or to infer from these broader developments that election monitoring itself is failing. Furthermore, regardless of the trends, elections remain a necessary component of a democratic society.8 Yet, the signs of slippage in democracy and freedom around the world are clearly alarming. Given that measures of democracy rightly lean so heavily on the quality of elections, the declining scores suggest that in some countries the quality of elections is not improving or may even be worsening. This makes it more pressing to ask whether election monitoring is worthwhile. Furthermore, monitoring has become such a central tenet of democracy promotion that it is imperative to examine its role. Although monitors do not have as much prominence in every election as in the Georgia case, when they do, it is usually in the more critical and interesting cases. The domestic and international media listen to their statements, as do governments around the world. Thus, what international election monitors say and do is of great consequence.
Unfortunately, the answer to the question of whether international election monitoring is a good idea is: We do not really know. Despite the significance of international election monitors, their activities receive little real scrutiny. Critics were vocal in the early years of election monitoring, but they usually based their criticism on their unique experiences with particular elections.9 Today, commentators occasionally question individual missions, as when the press accused the International Republican Institute (IRI) of withholding exit poll results after the 2007 election in Kenya,10 but—by and large—few commentators question their credentials and most simply treat them as a force for good. This is true of scholars, who repeatedly point to international election monitors as an effective way to improve elections without providing any evidence.11 It is also true of the media. For example, reporting on the downfall of a corrupt regime in Ukraine in 2004, The New York Times argued that the election monitors’ report “lent credibility to Mr. Yushchenko’s opposition movement and his supporters’ mass demonstrations, provided a basis for an international outcry, and helped lead to a complaint to the Supreme Court, which nullified the voting.”12 Naturally, international monitoring organizations likewise promote their own brand, arguing that they strengthen democratic institutions, boost public confidence, and deter fraud, intimidation, and violence.
Yet, as early critics noted, international election monitoring organizations are highly complicated actors and monitoring is a complex undertaking. Despite the experience they have gained over the years, they face several serious challenges. Elections are much more than a polling exercise: They begin months before polling day, and they involve a legislative framework, extended campaigns, and complicated administrational and logistical issues. Assessing elections is difficult, organizations have limited capacity, and, on top of that, organizations have to juggle multiple political and practical concerns. Although they do not like to speak too openly about them, monitoring organizations are aware of the problems and many try to address them. However, the will to improve varies considerably among the motley profusion of organizations and solutions are rarely apparent and often difficult. On some issues, organizations are stuck between a rock and a hard place. For example, they gain their leverage from their ability to legitimate or invalidate elections, yet this very task of assessment can also lead to thorny political entanglement. Even when monitoring organizations can prescribe solutions, they often lack the capacity to follow up and are at the mercy of domestic politicians to implement them.
Thus, it is not as straight forward as proponents suggest to assert that international election monitoring is worthwhile. Given their intrusiveness into domestic affairs and the weight their opinions receive, a critical third-party perspective on their activity is necessary. As the Roman poet Juvenal asked in his Satire IV in which a man places male guards outside his wife’s house to prevent her adultery: “Quis custodiet ipsos custodes?” [Who will guard the guardians?]. In a world that places so much emphasis on elections and on international election monitoring, this book assesses the guardians.
By injecting themselves into the domestic political process, monitoring organizations raise many interesting questions about their conduct and effects and, by extension, about the motivations of the international actors who sponsor them. For example: Do monitoring organizations actually reduce election violence by their presence or mediation?13 Do monitors influence domestic politics in other ways, for example, by influencing the decision of opposition parties to boycott elections?14 And what role do international monitors play in the training and effectiveness of domestic monitors?
This book touches on many of these questions, but it focuses exhaustively on two central and related questions: Do monitors assess elections accurately and objectively? Do monitors help improve the quality of elections? By focusing on the credibility of international institutions and the methods the international community uses to promote good domestic governance, these two questions focus the book on fundamental issues of global governance and democracy promotion.
The purported raison d’être of international monitors—their core mission—is to provide reliable and accurate information to the international community and to domestic actors.15 This role is particularly important in countries without credible domestic watchdogs such as a free media, an independent judicial system, or domestic observer groups.16 By taking on the role of producers of such information, however, monitoring organizations inevitably also become “legitimizers,” because they assess whether the election conformed to recognized principles or accepted rules and standards and thereby determine the legitimacy of the elected officials.
Although some organizations claim that they do not make categorical or simplistic “free and fair” or “thumbs up/thumbs down” statements, most organization do just that—or at a minimum are perceived by domestic and international audiences to be doing just that. Indeed, the official commission created by Kofi Annan to review the contested 2007 election in Kenya notes that “one of the most common purposes of electoral observation is to assess the legitimacy of an electoral process.”17 Partly due to the international community’s obsession with elections as the litmus test of democracy,18 election monitoring is, by extension, often the primary tool the international community uses to assess the legitimacy of governments.19 If international election monitors signal that elections were satisfactory, adequate, fair, legitimate—or whatever language they may employ—this has consequences for both international and domestic acceptance of the outcome. When Viktor Yanukovych claimed victory in the 2010 Ukraine presidential election, this did not prompt a second Orange Revolution, as it had in 2004, when international monitors disputed his claim. Instead, Prime Minister Yulia Tymoshenko, who had opposed him for the presidency, dropped her election challenge partly because this time international monitors had approved of the election, thus reducing her political ammunition.20
Yet, are monitoring organizations as impartial as they profess? Assessing elections is difficult. Monitors can only cover a fraction of polling stations and can only stay for a limited time at each station. Thus, choices are necessary. They may make pre-election assessment trips or have delegations in countries far in advance, but their resources are still limited, they lack local knowledge, and they may be up against politicians who work to deceive them. Thus, the efforts of international observers sometimes meet with cynicism, as expressed by the Nepalese voter in the chapter’s opening quote.
In addition to these logistical challenges, sometimes the political pressures on monitoring organizations are considerable. In the 2008 election in Georgia, the problematic pre-election period was followed by a fairly organized and peaceful polling day, although some precincts were chaotic and had problems with the ink used to safeguard against multiple voting. The counting was also slow and had “procedural shortcomings.”21 When exit polls showed Saakashvili with 52 to 53 percent, barely enough to avoid a second round, the opposition cried foul. The observers did endorse the election, albeit hesitantly.22 However, reactions were highly polarized as to the validity of that assessment; U.S. and Russian officials made contradictory statements and an Organization for Security and Co-operation in Europe (OSCE) observer openly criticized the mission.23 Thus, assessing the quality of an election is frequently contentious, and when more than one organization monitors an election, the monitoring organizations sometimes generate controversy by disagreeing on their assessments.
It was perhaps with such complications in mind that, when monitoring began to spread in the early 1990s, the renowned legal scholar, Thomas Franck, noted the importance of considering “the legitimacy of the emerging international rules and processes by which the governance of nations is increasingly being monitored and validated.”24 In other words, what rights does the international community at large have to assess and judge elections around the world? And when organizations do so, do they really base their opinions on “the highest standards for accuracy of information and impartiality of analysis”?25
Because of the practical, ideological, normative, and political difficulties inherent in monitoring, the quality of the monitors’ assessments cannot be taken for granted. This is an issue in global governance in general. Numerous monitoring bodies exist in global governance, but many of them are ineffective. This is particularly true in areas related to quality of government such as human rights, labor rights, gender equality, and similar issues on which governments have incentives to distort information about their less acceptable behaviors. Much of this monitoring occurs through self-reporting to various treaty organizations. Is the quality of election monitoring different from these processes? Do monitors provide more reliable information because they are present on the scene? Does the quality of the information vary between the different monitoring organizations or across different electoral contexts? If the quality of information varies, what does this mean for the legitimacy of international election monitoring itself and for the legitimacy that organizations bestow on governments? Thus, the question of quality of monitoring information has important normative implications as well as implications for the design of monitoring regimes more generally.
In addition, the quality of election monitoring assessments is important for the broader study of the nature of transnational actors. In the past this research has tended to assume that transnational actors are neutral and benign. Only recently have scholars begun to explore how the politics and preferences of transnational actors influence their behavior,26 and subsequently their ability to advance democracy both domestically and in international governance. Studying what factors influence the quality of monitors’ information encourages a deeper inquiry into the politics and norms of transnational actors in global governance.
To study the quality of information, this book asks a series of questions to help understand the motivations and methods of the actors involved: Why did election monitoring evolve in the first place? What sorts of organizations first became active, and what were their motivations? What countries invited monitors in the early days and why, and has the motivation to invite monitors since changed? How has the monitoring industry as a whole changed over time? When evaluating elections, what sort of considerations might monitoring organizations make? Is it possible to detect patterns in their assessments? Chapters 2–4 address these and other questions about the quality of election monitoring information.
Most international election organizations seek not only to inform domestic and international actors about the legitimacy of elections, but also to improve the quality of elections. Indeed, the main thrust of election observation is to promote good elections as an essential building block to better democracy. Election monitoring has indeed become the central component of the democracy promotion efforts of many organizations and governments. A study of whether election monitoring improves elections therefore gets at the core of many prominent democracy promotion programs around the world.
Unsurprisingly, international monitoring organizations voice great confidence in their own effectiveness. The European Union (EU), for example, notes: “Election observation can contribute to strengthening democratic institutions, build public confidence in electoral processes and help deter fraud, intimidation and violence.”27 The claims of other organizations are similar, or even stronger.28 The Asian Network for Free Elections (ANFREL) even argues that “the presence of international election observers has been proven effective in deterring and detecting violence and fraud as well as in providing greater confidence to candidates, political parties and the voting public.”29
Unfortunately, however, in reality very little is known about the effects of election monitors. Some scholars claim that monitoring is effective, arguing, for example, that it “limits the capacity of incumbents to engage in large-scale fraud,”30 that it has been “proven effective time and again in detecting and documenting deficiencies, manipulation, and fraud, thereby challenging the legitimacy of rulers who seek to stay in power through rigged elections,”31 and that “election monitoring not only facilitates reasonably fair elections but the development of basic democratic institutions and habits as well . . . [and] has thus become the central element of a rapidly developing international regime to preserve and extend democracy.”32 However, these claims are mostly unproven assertions; like those of the international election monitoring organizations themselves, they have not been subjected to thorough examination.
Instead, the existing research consists primarily of a vast set of case studies that examine a variety of issues through the lens of a given election or a smaller set of regional elections.33 Moreover, these case studies disagree about the effectiveness of international monitoring. Some case studies showcase benefits. For example, a study of the 2003 election in Armenia shows that although it was fraudulent, monitors lessened fraud in the polling stations they visited.34 Several case studies also credit monitors with increasing the electorate’s confidence in other elections. However, other studies strongly criticize international election monitors for being biased, unprepared, and under-resourced and question their ability to have any influence whatsoever on electoral struggles for political power.35
Together, these case studies present a valuable collection of research on international election monitoring, but given their disagreements they generate more questions than answers. Although practitioners and area specialists have paid attention to international election observers since the late 1980s, a comprehensive global study of whether international monitors improve elections across countries and over time is needed.36
Therefore, the second focus of this book is on whether international election monitors improve the quality of elections. Given the logistical and political challenges to their efforts to assess elections, as discussed above, skeptics would have plenty of reasons to question claims that monitoring organizations could actually influence the behavior of politicians in any way. Nevertheless, theoretically, monitors may be able to improve elections through several mechanisms.
First, monitors may be able to change the incentives facing the politicians. International monitors raise the cost of cheating by signaling increased international concern, calling greater attention to problems, and strengthening domestic critics. As the cost of cheating increases, politicians cheat less, or as one Carter Center (CC) observer expressed it, “they are more likely perhaps to play according to the rules.”37 International monitors may also raise the benefits of honesty by playing a verification role that makes it harder for opponents to dismiss honest victories as stolen. These changes in incentives may not always be sufficient to decrease cheating meaningfully. As this book explores, cheating is not an either/or choice, but a matter of degree. However, the rationalist expectation is that if an individual election is monitored, then the likelihood of cheating in that election decreases. The election is therefore more likely to be of higher quality.
Second, monitors may be able to change the conditions on the ground in various ways that facilitate improvements in the election process. Monitors make detailed recommendations that reinforce the message about what the international community expects. They can also help build capacity in several ways that can facilitate better implementation of electoral standards. Furthermore, over several elections the repeated interaction between external and national actors may socialize countries into norms and behaviors through persuasion and teaching.38 These domestic activities may help national actors improve their conduct of elections.
These channels of influence are complementary and may work through both constructivist and rationalist logics, that is, they may work through a combination of norms and incentives. Indeed, often the actual mechanisms of influence will be difficult to distinguish. For example, some studies have found that governments respond to shaming,39 a strategy whereby international actors strongly criticize governments publicly. It is hard to say whether governments react to such shaming because it alters their incentive structure or because they respond to the normative arguments. The same is true with international monitoring. Slow responses to the long-term engagement of monitors need not mean that the responses result only from socialization; domestic politicians may just as well be reacting to change in incentives over time. Regardless, both of these schools of thought provide theoretical reasons to expect that international monitors can improve the quality of elections through various mechanisms in both the immediate and the longer terms.
Yet many scholars remain skeptical that third-party actors ever really influence domestic politics or the behavior of governments. Realists have long dismissed international law and institutions as window dressing,40 arguing that any apparent influence merely reflects the fact that countries self-select into various international activities and commitments that they are predisposed to keep. And even if the mechanisms discussed above theoretically can occur, realists would contend that they are but a drop in the bucket—far too weak to exert any meaningful influence. Thus, several studies on a variety of topics ranging from the environment to human rights have found international efforts to curb government behavior ineffective.41 Following this line of thinking, international election monitors—who do not present a terribly formidable force—should have little influence. The harsh reality is that in politics power is everything: If incumbents risk losing power, they have no incentive to improve elections. They may even try to foil the efforts of international observers. This is perhaps why the skepticism about the effect of external actors is particularly pronounced among democratization scholars. Indeed, most accounts of democratic transitions remain decidedly domestic. Some scholars do acknowledge international influences,42 but many consistently downplay their effect on democratization43—even when actors resort to direct intervention.44
Who is right? Is international election monitoring useless, or can it actually improve elections under some conditions? In Chapters 5–8, this book examines this question in depth by asking a series of questions: Can monitoring deter cheating? Do politicians outwit monitors by simply shifting the way they cheat? What happens when monitors repeatedly return to a country? Do their efforts pay off? Do governments implement their advice? Do improvements, if any, last?
Given the lack of systematic study to date and the disagreements about the ability of international actors to influence domestic politics in general, these questions are important for academic, practical, and normative reasons. They go to the heart of global efforts to promote democracy. They represent the core liberal belief that it is possible to spread norms to reluctant governments, to bring about change in recalcitrant states. Thus, they are the questions this book sets out to answer.
The lack of research on the quality of election monitoring and the accompanying lack of consensus about their influence is understandable; just as assessing the quality of elections is difficult, so is assessing the quality and effects of monitors.45 Outcomes of democracy promotion efforts are particularly hard to measure,46 and electoral fraud has, perhaps wisely, received only scant attention from scholars.47
One challenge is that neither an objective measure of fraud nor an authoritative definition of a competitive election exists. Many irregularities go unnoticed, and the electoral process and context differs in every country. This makes it difficult both to know the truth and to apply a consistent measure across elections in different countries and at different times. However, an effort to reach consensus on these questions is beginning to emerge. For example, the European Commission has compiled a compendium of international legal commitments that in various ways outlines the details of the responsibility of states to hold elections and of the necessary elements of a competitive election.48 Experts from the CC are spearheading a similar and even more comprehensive effort.49 The Organization of American States (OAS) has sought to develop a standardized method of assessment.50 Likewise, scholarly work on the range and nature of election fraud is converging.51 Most important for the execution of this study, monitoring reports of elections have long shared a basic structure. Most regard, as will this book, an election as a long and comprehensive process, starting months before polling day and including the completion of tabulation and handling of disputes. They all include a common list of areas of assessment such as the legal framework, the media, the conduct of the campaign, the administration of the election, and so forth. And while there may not yet be a consistent way to measure fraud, in practical terms, a shared definition of fraud exists. Although some organizations may choose to ignore these behaviors, no international monitoring organization, for example, ever argues that it is acceptable for the incumbent to control the media, that vote buying is an acceptable cultural idiosyncrasy, or that dead voters belong in the registry. Thus, while legal scholars, politicians, and practitioners from around the globe may continue to debate the finer points of what constitutes a proper competitive election, on an operational level the consensus is quite workable.
A bigger problem, however, is that, even for experts who have devised long checklists of essential elements of proper competitive elections, these cannot be weighted and aggregated into a unified measure that applies uniformly across contexts. Experts at international monitoring organizations continually express this frustration. Thus, even if a fairly wide theoretical consensus about what constitutes competitive and free elections is emerging among scholars and practitioners, assessing the quality and context of all the factors in a given election is inherently subjective and subject to measurement errors. Furthermore, monitored elections may suffer from a “reporting effect” similar to that on the issue of human rights: It may appear as if the problems are more pronounced in monitored elections simply because more information is available.
Another fundamental challenge is that monitoring efforts are never administered randomly. This makes it difficult to isolate the effects of monitors because whether an election is monitored may well be related to the expected conduct of the election. Exacerbating this problem, as Chapter 2 shows, the rise of international election monitoring has coincided with a global rise in democracy.
In response to these challenges, this book seeks to be as comprehensive and thorough as possible in understanding and evaluating the broader phenomenon of election monitoring. Extensive immersion in the topic for six years has made it possible to follow developments and engage in debates on a level that reduces the risk of making shallow and misguided inferences. The study takes a pragmatic approach to building knowledge by combining inductive and deductive approaches though the simultaneous consideration of both facts and theories.52 As a result, the initial theories and ideas have been revised and refined through continued deepening of the inquiry. Furthermore, the focus is not exclusively on causal theorizing, but on deepening the understanding of international election monitoring and developing generalizations that can provide insights for policy.
To analyze the issues from as many angles as possible, the research relies on multiple methods of investigation. It combines historical inquiry, systematic comparative case studies, formal coding of the content of monitoring reports, and quantitative analysis, including descriptive statistics as well as more sophisticated modeling. It draws on varied sources, such as primary documents, interviews, and secondary sources.
Each method serves a unique purpose. The historical analysis in the following chapter places the study within a broader context and lays the foundation for the analysis of factors that explain the rise of monitoring. The historical analysis also connects international election monitoring to broader world developments such as the liberal efforts to promote freedom and democracy.
The comparative case studies investigate the role of monitors in countries over many elections. These case studies are updated through 2009. This long-term perspective provides understanding of conditions that influence the effectiveness of monitoring efforts over time and it also improves insights about the quality of the influence that monitors may have on a country. The case studies also help reduce spurious inferences by tracing whether the activities of monitors are plausibly connected with the development of policies and conduct of elections in each country. Situating the monitoring activities within richer narratives highlights other factors that were also influencing the developments in the individual countries. This helps to identify possible alternative explanations and conditions that modify the effect of the monitoring activities.
The quantitative data have been generated specifically for the project. The data cover 1,324 national elections from 1975 to 2004 of which about one-third were monitored. The data, which are described further in Appendix A, are from the Project on International Election Monitoring, which coded more than forty thousand pages of election monitoring documents, such as reports, press releases, and interim statements from more than six hundred monitoring missions. These documents provide the formal record of the experiences and conclusions of election monitors, and their coding allows for statistical analysis of what would otherwise be insurmountable information. The resulting dataset has been titled DIEM, the Data on International Election Monitoring. This dataset was supplemented by coding of more than four thousand pages of reports from the U.S. State Department on Human Rights Practices, providing comparable data covering not only the elections that were monitored, but also many that were not monitored.
The data are used in several ways. First, descriptive analysis helps reduce the complexity that arises from a global study. The project uses simple aggregation of data to cast light on the patterns of monitoring and reveal trends and differences between categories of data. The data are also used inferentially to analyze the questions of information quality and effect. Here a variety of different statistical models are used to examine the robustness of initial findings. In particular, the analysis of effects of monitoring takes advantage of some of the cutting-edge methods for addressing the selection problem discussed above.
Thus, although the empirical challenges make it tempting to set aside efforts to study election monitoring, this book argues that the topic is too important to justify ignorance. Monitoring has become the flagship of democracy promotion. It is costly, and, more important, the monitors’ assessments can have serious consequences in both domestic and international politics. The study of election monitoring can also illuminate important debates in political science and law about the ability of international actors to influence domestic politics and the legitimacy of transnational actors in global politics. Examining the role of international election monitoring is therefore highly compelling, even if it must necessarily be done cautiously. That is what this book sets out to do.
The book’s findings and arguments are complex, but the basic message is clear: International election monitoring is imperfect, but worth improving. The book refutes arguments that international efforts cannot curb government behavior and that democratization is entirely a domestic process. Yet it also boosts the critics who argue that democracy promotion efforts are deficient and that most outside actors are regrettably powerless. Although most monitoring information is credible, the book reveals flaws in the supposed objectivity of monitoring organizations and shows that these problems cannot be isolated to a few organizations. The book also shows that international monitoring works best as a reinforcement tool rather than a transformational tool. Monitors can promote progress, but only under certain conditions. Among these are constructive international incentives, domestic pressures for reform, less conflict-prone, zero-sum settings, and international monitoring organizations that are persistent, capable, and free of political baggage. The conclusion crystallizes several dilemmas that the international community faces regarding its use of international election monitoring to promote democracy. It also raises a host of questions to stimulate discussion about the quandaries the international community faces both regarding election monitoring specifically, but also about the legitimacy and value of the global promotion of democracy and liberal values more generally.