The present document can't read!
Please download to view
Please download to view
Download Mozaffar and Schedler 2002
International Political Science Review http://ips.sagepub.com The Comparative Study of Electoral Governance—Introduction Shaheen Mozaffar and Andreas Schedler International Political Science Review 2002; 23; 5 DOI: 10.1177/0192512102023001001 The online version of this article can be found at: http://ips.sagepub.com/cgi/content/abstract/23/1/5 Published by: http://www.sagepublications.com On behalf of: International Political Science Association (IPSA) Additional services and information for International Political Science Review can be found at: Email Alerts: http://ips.sagepub.com/cgi/alerts Subscriptions: http://ips.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.co.uk/journalsPermissions.nav Citations http://ips.sagepub.com/cgi/content/refs/23/1/5 Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 International Political Science Review (2002), Vol 23, No. 1, 5–27 The Comparative Study of Electoral Governance—Introduction SHAHEEN MOZAFFAR AND ANDREAS SCHEDLER ABSTRACT. Electoral governance is a crucial variable in securing the credibility of elections in emerging democracies, but remains largely ignored in the comparative study of democratization. This article develops some basic analytical tools to advance comparative analysis and understanding of this neglected topic. It conceptualizes electoral governance as a set of related activities that involves rule making, rule application, and rule adjudication. It identifies the provision of procedural certainty to secure the substantive uncertainty of democratic elections as the principal task of electoral governance. It argues that electoral governance, while socially and institutionally embedded, matters most during the indeterminate conditions that typically attend democratization. Finally, it outlines a research agenda that covers the comparative study of the structures as well as the processes of electoral governance. Keywords: Democracy • Democratization • Elections • Electoral governance • Electoral institutions The unprecedented 2000 US presidential election afforded us an unforeseen natural experiment to reinforce the central message of this thematic issue: Electoral governance matters. Events and controversies surrounding the election in Florida underscored the importance but also illustrated the paradox of electoral governance in securing credible democratic elections. Both scholarly recognition and systematic analysis of the role of electoral governance in securing credible elections are hindered by the fact that elections in established democracies tend largely to be routine events that produce results outside the unacknowledged “margin of error” that exists in all democratic elections. Because democratic elections entail the largest peacetime mobilization of the national population in a short time span and require the coordination of hundreds of individuals engaged in hundreds of different activities, they are almost always infected with errors 0192-5121 (2002/01) 23:1, 5–27; 020421 © 2002 International Political Science Association SAGE Publications (London, Thousand Oaks, CA and New Delhi) Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 6 International Political Science Review 23(1) stemming from, among other things, defective ballots, incomplete and inaccurate voter rolls, exclusion of registered voters, inaccuracies in counting, tabulating and recording votes, and human mistakes. These inaccuracies define the margin of error in all elections. There is, in other words, no such thing as a perfect democratic election. But to the extent that these errors are random and do not ex ante determine the results, electoral credibility obtains. Credible electoral routines thus tend to obscure the importance of electoral governance in securing them. It is only when “elections go bad” (Issacharoff, Karlan, and Pildes, 2001) that electoral governance attracts critical scrutiny. In established democracies, elections go bad not because they embody deficiencies, mistakes and inaccuracies. They go bad when these otherwise random problems, usually caused by the sheer magnitude of coordinating hundreds of discrete tasks to create a stable framework for electoral competition, systematically affect the outcome of that competition. And these problems systematically affect the electoral outcome when that outcome itself falls within the margin of error defined by them. In Florida, a statistical dead heat produced by the closest presidential race in over a century sharply accentuated the otherwise negligible effects of routine electoral errors on the final result. It highlighted problems widely known to exist but too innocuous to worry about and too expensive to correct. Helped by “real time” television coverage, the virtual tie precipitated a penetrating public scrutiny of the overall structure and processes of electoral governance in the state and the country. Paradoxically, then, electoral governance attracts serious attention not when it routinely produces good elections but when it occasionally produces bad ones. It is this paradox that has obscured the empirical relevance and analytical significance of electoral governance. As far as “people take for granted the administrative dimension of elections” (Pastor, 1999a: 76) they tend to overlook the critical role of electoral governance in securing the credibility and continued legitimacy of democratic elections. This role is obviously important, although not well examined nor understood, in established democracies. But it has a special resonance in emerging democracies, where deliberate electoral manipulation and systematic fraud by recalcitrant authoritarian rulers unwilling to give up power have often blocked, derailed or truncated transitions to democracy. Growing, but as yet unsystematic evidence, alongside the considerable amount of practical expertise accumulated by international donors, indicates that ineffective electoral governance is an important cause of many flawed elections witnessed in transitional regimes over the past three decades. Effective electoral governance alone does not guarantee good elections, of course, because a complex variety of social, economic and political variables affect the process, integrity, and outcome of democratic elections. But good elections are impossible without effective electoral governance. Since the late 1980s, an impressive network has emerged of national and international organizations active in election monitoring and democracy assistance.1 But students of democratic politics have been slow to recognize the practical relevance and analytical import of the new field. Electoral governance remains a “neglected variable” in the study of political democratization (Elklit and Reynolds, 2000; Pastor, 1999b). This neglect stems in part from the normative orientation of discussions that have focused on developing evaluative criteria to assess deviations from preconceived (and often somewhat idealized) notions of democratic progress. While we acknowledge the importance of normative Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 7 considerations, our goal in this Introduction is to stress the analytic significance and utility of electoral governance in the study of political democratization. Systematic examination of electoral governance requires attention to several analytical (conceptual, theoretical, and methodological) issues. The main conceptual tasks involve specifying the meaning of electoral governance, conceptualizing its institutional location in the political system, and disaggregating it into researchable procedural and structural variables.2 The theoretical core issues concern the origins, political logic, and causal implications of electoral governance. Finally, the central methodological challenge consists in elaborating viable strategies for studying electoral governance. While we do not pretend to resolve all these challenges in the remainder of this Introduction, we do wish to advance a good way in developing a substantive agenda as well as an analytical framework for conducting comparative research on electoral governance. Levels of Electoral Governance Elections involve more than voting. The formal act of casting a ballot is preceded by electoral competition in which rules defining, among other things, the electoral formula, district magnitudes, district boundaries, and assembly size coordinate the strategic choice of voters, parties, and candidates over votes and seats (Cox, 1997; Lijphart, 1994; Taagepera and Shugart, 1989). Electoral governance is the wider set of activities that creates and maintains the broad institutional framework in which voting and electoral competition take place. It operates on three levels: rule making, rule application, and rule adjudication. Rule making involves designing the basic rules of the electoral game; rule application involves implementing these rules to organize the electoral game; rule adjudication involves resolving disputes arising within the game (see Table 1). Above the first level, the “meta-game” of constitutional rule making defines who possesses the authority for defining the rules of electoral governance. In democratic transitions, struggles for changing rules are often intertwined with struggles for changing meta-rules. While acknowledging the foundational role of constitutional rules, we focus our attention here on the three lower levels. First, at the level of rule making, electoral governance involves the design of institutions that define the basic framework of democratic elections. Rules of electoral competition and rules of electoral organization configure this framework. Students of electoral systems have focused almost exclusively on the political consequences of the rules of electoral competition. The study of electoral governance looks at their political origins and covers a wider basket of electoral rules. Traditional electoral rules covering suffrage rights, rules of representation, assembly size, district magnitude, district boundaries, and electoral calendars form part of the agenda. But the basic framework of democratic elections also includes additional sets of rules that have been largely neglected in extant scholarship: the formal rules that govern voter, party, and candidate eligibility and registration; rules regulating election observation; laws and regulations that affect the resource endowments of parties and candidates (their access to money and media); rules prescribing the method of counting, tabulating, and reporting votes; and laws establishing the structure, jurisdiction, and operational framework of election management bodies and dispute settlement authorities. Second, at the level of rule implementation, electoral governance coordinates the tasks of diverse personnel and organizes the execution of a complex array Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 8 International Political Science Review 23(1) TABLE 1. The Three Levels of Electoral Governance. Levels 1. Rule Making Choosing and defining the basic rules of the electoral game. (a) Rules of Electoral Competition: Elements – Formula – District magnitude – District boundaries – Assembly size – Electoral time table – Franchise – Voter registration – Party and candidate registration – Campaign financing and regulation – Election observation – Ballot design – Polling stations – Voting, counting, and tabulating – Election management bodies – Dispute settlement authorities – Registration of voters, candidates, parties – Registration of election observers – Voter education – Electoral organization – Voting, counting, and reporting – Admission of complaints – Processing of cases – Publication and implementation of rulings (b) Rules of Electoral Governance: 2. Rule Application Organizing the electoral game. 3. Rule Adjudication Certifying election results and resolving disputes. of interdependent activities to establish the stable institutional basis for voting and electoral competition. Electoral governance as rule application evokes an idealized image of bureaucratic routine and mechanical execution of coherent sets of known and tested rules, an image erroneously conveyed by the familiar notions of “electoral administration” and “electoral management.” Yet, in transitional regimes and new democracies, the complex task of organizing electoral contests is much closer to the “diabolic” paradoxes of politics (Weber, 1919) than to the tedium of bureaucratic routine. It involves pursuing and reconciling three “conflicting imperatives” (Gould, 1999): administrative efficiency, political neutrality, and public accountability. These three goals are “interdependent but contradictory” (ibid.: 439). Electoral officials cannot neglect any one of them, nor can they maximize them all together. In the context of material scarcity and political distrust that is common to many emerging democracies, electoral credibility can obtain when electoral governance succeeds in meeting the three challenges and balancing the trade-offs they involve.3 Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 9 The challenge of administrative efficiency: It is precisely at the level of rule application that electoral governance is most susceptible to errors, not only because of the sheer magnitude and complexity of the tasks that need to be accomplished but also because of the large number of people involved and the authorized discretion they exercise in accomplishing their tasks. Elections involve the largest peacetime mobilization of the national population. Their organization and conduct therefore require complex logistical exercises usually under severe time constraints. To obtain a credible election process and unproblematic voting, these very complexities and constraints dictate a high level of central coordination and strategic planning. Universal adult suffrage and equal participation dictate easily accessible voting places, which requires locating them across the country and staffing them with adequate and trained personnel possessing sufficient devolved authority to manage local contingencies. In federal systems, administrative exigencies also require balancing central control with local autonomy. However, unrestrained local autonomy can engender such marked variations in patterns of electoral governance as to diminish the credibility of the election process.4 The inordinate complexity and interdependence of the multifaceted activities that comprise the election process also demand some form of project planning spelling out the mission of election management, clarifying the strategic objectives to be achieved, and specifying the mechanisms for achieving them in a timely fashion. Electoral governance as rule application consists of innumerable technical activities whose efficient organization and execution determine the credibility of elections. Inadequate attention to operational details can seriously compromise that credibility. An emphasis on project planning does not necessarily imply bureaucratic rigidity, however. Rather it draws attention to the importance of utilizing lessons learned from previous elections to reduce the marginal cost of future elections. The challenge of political neutrality: In transitional regimes, opposition parties are deeply suspicious of electoral processes, and usually for good reasons. Authoritarian rulers seeking electoral legitimation typically attempt to manipulate the structures and processes of election administration to secure favorable outcomes (Schedler, this issue). Election authorities thus have to perform their duties in a way that dissipates suspicions over their political neutrality. Putting it colloquially, they have to show that they are more than just pliant puppets of the regime. The stringent norms of behavior specified in the code of conduct formulated by the International Institute for Democracy and Electoral Assistance (International IDEA) provide useful guidelines for election officials to maintain strict neutrality.5 How successful electoral authorities are in depoliticizing the management of elections depends in part on how successful they are in achieving administrative efficiency. In the context of limited financial and human resources prevailing in most emerging democracies, organizational excellence is difficult to achieve. Some irregularities are bound to happen. Yet opposition parties tend to read technical failures as indicators of fraud. The “intersection between political suspicion and technical incapacity” (Pastor, 1998: 1) may spark ardent disputes with a potential of derailing the whole process. Preventing technical problems from contaminating an electoral process with corrosive suspicions is not an easy assignment, though. Deep-seated distrust often impedes pragmatic trade-offs between organizational efficiency and the establishment of institutional safeguards against fraud and manipulation. Yet, as the Florida controversy in the Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 10 International Political Science Review 23(1) 2000 US presidential election indicates, even in democracies with extended electoral histories and well-established generalized norms of apolitical bureaucracy, the presumption of administrative impartiality may be called into question where the institutional design of electoral governance fails to separate the exercise of public authority from political partisanship. The challenge of public accountability: Electoral governance involves the exercise of discretionary authority constrained by formal rules. Formal rules set boundaries to permissible behavior but do not eliminate discretion. They are seldom sufficiently clear, specific, and consistent to realize the bureaucratic ideal of mechanical rule application. The ambiguity and indeterminacy that inevitably dwell in electoral codes require election authorities to exercise some measure of administrative discretion. The exercise of this discretion, however, may put into question administrative efficiency and political neutrality (Schedler, 2000)— hence, the demand for public accountability. Accountability involves three key dimensions: information, justification, and enforcement (Schedler, 1999c). Strategies of accountability in electoral governance have hitherto focused on the first aspect. Nonpartisan election monitoring, partisan oversight, public opinion polling, and official sunshine policies are all aimed at maximizing the transparency of the electoral process. Transparency helps to prevent, detect, and correct irregularities as well as to document the integrity of election organizations. The presence of party representatives in consultative bodies, media scrutiny, and academic investigations of election management oblige election officials to publicly justify their decisions in the light of legal provisions, normative principles, and material constraints. Finally, enforcement provisions specifying sanctions for dereliction of duties are designed to discourage abuses of discretionary authority in electoral governance through material incentives. While indispensable for making electoral governance efficient, neutral, and credible, securing accountability may undermine these very goals. Facilitating reasonable trade-offs hinges on the choice of institutions and strategies of accountability. For instance, the central deployment of “police patrols” may sometimes be less effective for controlling local election officials than a reliance on decentralized mechanisms of “fire alarms” (McCubbins and Schwartz, 1984). Bureaucratic rigidity in the “keeping of detailed records and the establishment of audit trails” (Maley, 2000: 6) may divert scarce resources from substantive activities. In sum, the challenge is not to maximize but to optimize accountability. Third, at the level of rule adjudication, electoral governance involves the mediation and settlement of disputes arising out of the process and the results of voting and electoral competition. Settling disputes over election results is a vital function of electoral governance, as shown by the decisions of the various local, state and federal appeals courts, and, most dramatically, by the decision of the US Supreme Court in the 2000 US presidential elections. Rule adjudication prominently involves the authoritative resolution of disputes that arise from ambiguities in complex election rules and operational problems in their implementation. For example, when the Ghana Election Commission’s vague guidelines failed to clarify whether the old thumbprint voter identification cards used in previous elections or the new photo identification cards would be valid for the 7 December 2000 national elections, the Ghana Supreme Court determined four days before the elections that both cards would be considered valid. The controversy arose because only 80 percent of the old cards had been replaced with Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 11 new ones by the time elections were held (Temin, 2000). By providing institutional mechanisms for amelioration of electoral errors and peaceful management of resulting political conflicts, the impartial and expedient adjudication of electoral disputes represents a cornerstone of the procedural legitimacy of democratic elections. Institutionalization of Democratic Uncertainty Competitive elections are the hallmark of modern representative democracy. As the institutionalized means by which large numbers of people participate peacefully in selecting and removing governments, they are the primary, albeit not the only, source of democratic legitimacy.6 In electoral politics, as in other spheres of modern life, formal institutions are seen as legitimate only when their substantive outcomes are indeterminate. Political actors will accept the uncertainty of outcomes in electoral competition to the extent that the rules that organize the competition do not ex ante determine those outcomes. In electoral contests as well as court proceedings, procedural legitimacy presupposes substantive uncertainty (Luhmann, 1983). Substantive uncertainty, then, is a reliable measure of the impartiality of institutions. If winners and losers are known beforehand, institutions appear biased. The “discriminating factor” between “efficient” (impartial) and “redistributive” (biased) institutions lies in “the uncertainty of the outcomes they produce” (Tsebelis, 1990: 117). The close association between procedural legitimacy and substantive uncertainty poses the paradoxical challenge of “institutionalizing uncertainty” (Przeworksi, 1988: 63). The paradox is that substantive uncertainty requires procedural certainty.7 It is this paradox that defines the central task of electoral governance: organizing electoral uncertainty by providing institutional certainty. Distinguishing between substantive and procedural uncertainty enables a more nuanced understanding of variations in political actors’ risk-aversion. Authoritarian and democratic actors, for instance, exhibit different attitudes towards uncertainty. While the former attempt to reduce the uncertainty of outcomes, the latter attempt to reduce the uncertainty of institutional rules. If the central task of electoral governance is to institutionalize democratic uncertainty, failure to do so may originate at any of its three levels. Authoritarian actors may try to reduce electoral uncertainty by mobilizing bias at the stage of institutional design, implementation, or dispute settlement. As long as electoral governance artificially reduces electoral uncertainty at any of its three levels, electoral processes are unlikely to be perceived as fully democratic or, by implication, fully legitimate. First, the primary “challenge of a transition is to negotiate electoral rules that all parties can accept and respect” (Pastor, 1999b: 15). Unless opposition parties perceive the basic rules of competition to be “reasonably fair” (O’Donnell and Schmitter, 1986: 58) they will refuse to participate in the electoral contest or fail to accept its results. If “for whatever reason, right or wrong, [electoral] rules are perceived as illegitimate, then democracy is in trouble” (Taagepera, 1998: 80). Second, reaching political compromises at the initial stage of institutional choice does not preclude future organizational failures. New electoral institutions are chosen under the pressure of political uncertainty surrounding regime transitions. Their choice reflects short-term calculations that do not accurately anticipate future contingencies. Operational problems inherent in bureaucratic Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 12 International Political Science Review 23(1) routines and administrative decision-making reduce organizational efficiency. Any or all of these factors, as well as a lingering authoritarian impulse to contain electoral uncertainty, are likely to undermine the ability of electoral governance to meet minimum standards of administrative integrity at the level of rule application, precipitating the decision of opposition parties to withdraw from flawed elections and reject their results. Third, the resolution of electoral disputes represents the concluding act of electoral contests. Failure at this final stage may ruin any advances made at prior stages. Controlling the judges can be the easiest way of controlling electoral outcomes (Eisenstadt, this issue). But the reverse is true as well. When the election management founders, electoral conflict adjudication may still provide an institutional safety-valve. Organizational failures may be “redeemed” by judicial actions that resolve complaints fairly and expediently. The entry of political parties into electoral competition implies their acceptance of its basic rules, which in many instances they helped design in the first place. But experience, pragmatism, and self-interest often lead them to qualify their endorsement of rule application, making their acceptance of the electoral process contingent on the expected capacity of existing conflict mediation agencies to rectify eventual administrative irregularities. Indeterminate Embeddedness The claim that competitive elections foster democratic legitimacy is obvious. The claim that electoral governance contributes substantially to this effect is obvious intuitively, but less so analytically. To what extent, if any, does electoral governance determine the democratic quality of elections? Doesn’t that quality depend simply on the incumbent parties’ willingness and capacity to manipulate the process or hold free and fair elections? Does electoral governance have an autonomous impact or does it merely express the prevailing correlations of power between incumbents and opposition? How much variance does it explain in patterns of political democratization? We possess few systematic answers to these questions. This is due largely to the fact that serious interest in electoral governance is the product of the practical experience of international election monitoring and democracy assistance over the past two decades. Most practitioners and scholars who have participated in these activities and have been concerned with issues of electoral integrity have been more interested in evaluating than in systematically explaining democratic progress. They have treated electoral manipulation not as an explanatory variable but a measure of democratization, not as a cause but an indicator. Accordingly, they have focused attention on measurement problems, on the intriguing question of how to conceptualize, observe, measure, weight, and aggregate violations of democratic norms that may occur during the electoral process (see Choe and Darnolf, 1999; Elklit and Svensson, 1997). Given the absence of an extensive, analytically rigorous literature, we cannot present a coherent set of theoretical insights grounded in systematic examination of the role of electoral governance in the process of democratization. Yet, recognition of two basic ideas seems to be an indispensable starting point for comparative research: (1) the institutional and societal embeddedness of structures and processes of electoral governance and (2) the intrinsic indeterminacy of context in situations of regime change. Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 13 (1) Like all democratic institutions, electoral governance institutions neither emerge from nor operate in a vacuum. History and context shape their choice and consequences, not inexorably or uniformly, but by defining the structure in which differentially endowed actors make strategic choices. Analytically, then, electoral governance institutions are best understood as “embedded institutions” (Grofman et al., 1999; Przeworksi and Teune, 1970). Their origins and unfolding trajectories flow out of power struggles and normative choices that are grounded in historical configurations of institutional, cultural, and socioeconomic factors. Thus, contextual features, such as macrolevel variables of the political system and microlevel choices of political actors, shape not only the choice of rules that configure the structure and process of electoral governance but also their operational effectiveness and political outcomes. Some select examples provide support for the preceding theoretical statements. In Senegal with a short democratic history under a single-party dominant “semidemocratic” regime, the incumbent president, concerned with retaining executive control and determining the pace and scope of incremental democratization, agreed to the demand of an increasingly unified opposition for an independent election commission with only oversight responsibility, retaining the administrative responsibility for the organization and conduct of elections within the ministry of the interior (Mozaffar and Vengroff, 1999, forthcoming). In democratic transitions in the former Soviet Central Asian Republics of Kazakhstan, Kyrgyzstan and Uzbekistan, electoral governance institutions emerged as negotiated settlements of power struggles between regional and central elites entrenched by protracted communist rule (Jones Luong, 2000). Finally, in the 1996 elections in Ghana, a country with previous, albeit checkered, democratic experience and a tradition of inter-elite accommodation, the open and constructive attitude of the electoral commission, exemplified especially by its willingness to involve the opposition parties closely in key decisions on the organization of the election, played an important role in overcoming lingering doubts about its political impartiality (Gyimah-Boadi, 1999). These examples obviously do not exhaust the questions that could frame systematic analysis of the contextual embeddedness of electoral governance. For instance, what, if any, is the relationship between the form of government (parliamentary or presidential) and the structure, process, and outcome of electoral governance? How does the territorial organization of power (unitary or federal) affect electoral governance? Do party systems (two-party or multi-party) affect the form and function of electoral governance? And how do patterns of ethnopolitical diversity affect the choice, operation and outcomes of electoral governance institutions? (2) In their seminal essay, O’Donnell and Schmitter (1986: 3) described regime transitions as highly indeterminate periods of “extraordinary uncertainty.” Since then it has become commonplace among students of democratization to conceive democratic transitions as times of “intense political uncertainty” (Bratton and van de Walle, 1997: 10) in which, in the extreme, “anything might be possible” (Goodin, 1998: 47, emphasis removed). Transitions thus appear as periods of institutional crisis where the basic rules of the political game look uncertain. For agents of electoral governance, the uncertainties of regime change involve risks as well as opportunities. Democratic transitions magnify the intrinsic contradiction between control and legitimacy that is routinely present in political life. Authoritarian rulers who feel compelled to hold competitive elections in the face Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 14 International Political Science Review 23(1) of domestic and international pressures confront the dilemma of risking their (even nominal) democratic credentials if they do not restrain their desire to manipulate the election process to secure favorable outcomes. Convoking elections thus opens a window of opportunity. The window may be small and opaque, but it cannot be sealed hermetically. In all but the worst electoral farces, agents of electoral governance possess some degree of freedom in fostering credible elections (Schedler, this issue). This may be true even in harshly authoritarian regimes, such as Uruguay in 1980 and Chile in 1988. Military rulers in both countries lost plebiscites they themselves convoked, thanks not least to a resuscitated pre-authoritarian tradition of clean election management. Structure of Electoral Governance Numerous contextual factors shape the process of electoral governance. The prevailing correlation of forces between incumbents and opposition is perhaps the most obvious determinant of the fate of electoral processes. But there are good reasons to expect that formal institutions may matter as well. Rule-configured formal structures may decisively constrain the otherwise unfettered exercise of power by political elites. However, among the innumerable rules that prescribe formal roles and responsibilities for the organization and regulation of elections, which do we expect to be decisive? Which variations in institutional design do we expect to matter most? We suggest that an initial research agenda on the institutional foundations of electoral governance should pay attention to institutional choices along six basic dimensions: centralization, bureaucratization, independence, specialization, delegation, and regulation. Centralization. The decentralized structure of electoral governance found, for instance, in Switzerland, the United Kingdom, and the United States, is widely considered to be unworkable in new democracies. The fear that local powerbrokers might hijack electoral processes has blocked any move in this direction. Central control over the organization of national elections is viewed as indispensable for preventing the possible feudalization of electoral governance. Still, in federal systems, electoral designers have to decide whether to make the national election management body responsible for the organization of subnational elections as well. Democratizing Mexico, for example, developed a dual structure whereby each federal state replicates the national system of electoral governance, complete with its own electoral institute and tribunal in charge of state and municipal elections. Yet, electoral reforms in the mid-1990s extended the jurisdiction of the federal electoral tribunal to local conflicts, erecting it as the appellate court of last instance in all electoral matters. An interesting comparative research question is if and how appellate courts in federal systems introduce elements of central control into otherwise decentralized systems of electoral governance, thus homogenizing both electoral laws and electoral practices at subnational levels.8 Bureaucratization. Institutional designers have to decide whether to establish a permanent election commission and, if so, whether to put it at the top of a permanent bureaucratic apparatus (Mozaffar, this issue). Ad hoc commissions that have to start “from scratch at every new election” have become “extremely rare” among contemporary democracies (López-Pintor, n.d.: 57, 59). Still, a wide Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 15 variation exists in the size of the full-time professional staff of electoral commissions. The Mexican Federal Electoral Institute, for example, employs a permanent staff of over 10 000 officials, while to run a national legislative election the Indian Election Commission borrows some 4.5 million staffers “from various governmental agencies” and places them “temporarily under [its] total command” (Gill, 1998: 166). An independent electoral commission with a permanent administrative apparatus raises questions about intra-organizational processes, especially with respect to patterns and effectiveness of control and accountability (Schedler, 1999a). For example, central control of lower officials may help to increase accountability and reduce errors due to local discretion but at the cost of inefficiency and rigidity. The problem is typically one of balancing related but mutually competing, not mutually exclusive, imperatives that underpin and animate all organizations. Variations in central control over recruitment pose additional dilemmas, since the quality, independence, and commitment of personnel are crucial for institutional credibility. The dilemma is especially salient in transitional elections where, as in Cambodia in 1998 (Neou, 1999: 155), authoritarian incumbents’ success in staffing electoral bureaucracies confront democratic actors with the problem of dealing with loyalist holdovers in public administration. Independence. In most Western European democracies, responsibility for organizing elections rests with the central state, with the ministry of the interior usually executing that responsibility under the supervision of either a judicial or a multiparty body. A reliable tradition of bureaucratic neutrality is the precondition for leaving the responsibility for electoral governance to the state. In most new democracies, however, the absence of trustworthy state bureaucracy “usable” (Linz and Stepan, 1996: 10) for democratic purposes renders such an option unworkable. Not surprisingly, therefore, removing electoral governance from executive control has turned into a rallying cry of democratizing forces all over the globe. Establishing an independent electoral commission, in fact, has become a compelling international norm, a sine qua non of electoral credibility. With the notable exceptions of some Eastern European and francophone African countries (Mozaffar, this issue), new democracies have “almost invariably [moved] in the direction of establishing independent electoral commissions independently of [their] historical political traditions” (López-Pintor, n.d.: 41). The idea of establishing such “islands of integrity” (Eigen, 1996) by means of institutional independence carries strong intuitive appeal. We possess little systematic knowledge, however, of how different formal rules translate into actual practice. In principle, we should be prepared to find a similar tendency students of central bank independence have identified in developing countries. In contexts of high “informal institutionalization” (O’Donnell, 1996), the legal independence of central banks turns out to be “a questionable proxy for behavioral independence” (Maxfield, 1999: 286). If the realm of electoral governance displays the same loose coupling between formal institutions and actual practice, we should see deviations in both directions. We should see institutions that are independent on paper, but intimidated, colonized, or neutralized in practice by governmental authorities. And we should see electoral authorities acting independently despite their formal subordination to government agencies. The Peruvian elections board for the 2000 elections exemplifies the former situation, Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 16 International Political Science Review 23(1) while the Nicaraguan electoral commission of 1990 illustrates the latter. Clarifying the conditions under which formal rules of electoral governance are likely to be effective is thus a promising topic of research. Specialization. The administrative and judicial functions of electoral governance can be fused, as in Costa Rica’s Supreme Tribunal of Elections endowed in 1949 with full responsibility for both electoral management and electoral certification. Or they can be separated, with the responsibility for settling electoral disputes being assigned to either another branch of government, a specialized tribunal, or the ordinary court system. Legislative “self-certification” was highly popular in the Americas throughout the nineteenth century (and still is in the United States). But its conflictual fusion of roles analyzed by Fabrice Lehoucq (this issue) has discredited it beyond repair.9 In Mexico, successive electoral reforms elevated the autonomous Federal Electoral Tribunal, founded in 1986, as the final arbiter over all electoral disputes. Finally, reliance on the ordinary court system for the settlement of electoral disputes is the most widespread solution in contemporary democracies. In principle, it works as well as ordinary courts work. If the judicial system is biased and ineffective, it may actually subvert progress achieved in the impartial and professional administration of elections. In Ghana, for example, legal challenges of the 1996 legislative elections tended “to drag on seemingly interminably” (Gyimah-Boadi, 1999: 118). The ineffectiveness, inconsistency, and timidity of appellate courts ended up creating the strong impression that even in the face of evident irregularities, “once declared, reversal of election results is most unlikely” (ibid.). The 2000 post-electoral disputes in the United States showed that the judiciary may lose its aura of professional reliability and legal impartiality when called upon to adjudicate divisive political conflicts. Delegation. Where the ruling party cannot be trusted to run fair elections, a logical option for incumbents and opposition is to establish structures of mutual restraint. Above all, they may agree to concur in the appointment of the members of the national election commission. Such structures of cooperation may take two forms. Parties may develop a quasi-consociational scheme of “power sharing” with a pluralistic “partisan” body of electoral administration that includes opposition representatives as powerful “veto players” alongside representatives from the ruling party. In this scheme, political parties appoint their own representatives (as in Honduras) or a legislative majority appoints representatives on the basis of formal party quota (as in El Salvador until 1993).10 Alternatively, instead of nominating party representatives, the major parties may select individuals (either experts or notables) with a reputation for independence and impartiality. Instead of sharing power in a multiparty commission, they may delegate power to a presumptively neutral, non-partisan body. Canada, Australia since 1984, Mexico since 1994, and Venezuela in 1998 and 1999 fit this “delegative” scheme. If trusted “third parties” exist either in the judiciary or in civil society, political parties may decide to abdicate their authority to determine the composition of top electoral offices. They may leave the appointment of election managers either in the hands of citizens (as in Paraguay since 1992), elected local and regional assemblies (as envisaged in the 1987 Haitian constitution), the judiciary (as in Brazil and Costa Rica), or civil associations (as in Guatemala and Peru). These non-partisan bodies (Haiti excepted) usually appoint non-partisan election officers. In Colombia, by contrast, judicial authorities elect party representatives in Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 17 numbers roughly proportional to the composition of the national parliament (Jaramillo, 1998: 217).11 The three ideal-typical modes of electoral governance—power sharing, delegation, and abdication—open up an intriguing field of research. How do different institutional solutions affect the efficiency and credibility of electoral commissions? What strategic and contextual factors account for the choice of one solution over another? For instance, it is plausible to assume that high levels of distrust between parties in democratizing countries motivates them to set up independent election-management bodies (López-Pintor, n.d.: 52, 55). However, while distrust of the government may explain why opposition parties struggle to dismantle executive dominance in the organization of elections, it cannot explain parties’ micro-institutional choices. It cannot explain, for example, why parties sometimes cede control over electoral governance to independent experts (the delegative design) and sometimes set up multiparty commissions and retain control (the consociational design). One possible explanation lies in the structure of the party system. Political parties are likely to decide on power sharing in a twoparty system and choose a delegative design in a more fragmented multiparty system (Molina and Hernández, 1999). In the former system, a partisan commission gives veto powers to both players. In the latter, some parties may form exclusionary alliances against others. Regulation. As principal-agent theory tells us, distrustful “principals” may try to limit the damage their “agents” may cause by constraining them through detailed “contracts.” That is, principals may attempt to reduce agents’ discretion by sealing them in bureaucratic cages. Electoral reformers often intend to do the same. They try to minimize the discretion of electoral bureaucracies by subjecting them to extensive legal regulation. In the course of the country’s democratic transition, Mexico’s electoral law, for instance, has experienced a major distrust-driven shift from loose subregulation to dense overregulation (Schedler, 2001). Here, again, the current state of comparative research presents more questions than answers. For instance, what is the role of electoral regulation under authoritarian regimes? Under which conditions do opposition parties trust the effectiveness of legal changes? What is the role of international diffusion in the drive towards tight electoral regulation? How much do formal changes of electoral rules depend on changes in the correlation of forces between government and opposition to become effective? And for how long do formal constraints survive the distrust that gives birth to them? Process of Electoral Governance Systematic analysis linking structure and process requires specifying and identifying the empirical components of the process. This is difficult for electoral processes. One obstacle for empirical research lies in the inordinate complexity of electoral processes. Another derives from the fact that assessing both the effectiveness of electoral governance and the democratic quality of elections often requires exploring systematic irregularities that by their very nature are difficult to document. The reliability and validity of data measuring the process components of elections are likely to be suspect. Attempts to surmount these methodological problems have produced four approaches to identifying and collecting relevant data: (1) comprehensive, (2) selective, (3) subjective, and (4) indirect. Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 18 International Political Science Review 23(1) (1) The comprehensive approach: Inspired by the practical experience of international election monitoring and democracy assistance of the past twenty years, the so called checklist approach suggests studying electoral processes in their entirety to detect irregularities at any point along the way. Starting out with a narrow search for irregularities on election day, international monitors have come to compile comprehensive lists of items requiring evaluation in order to reach a meaningful judgment on the quality of a given election. Some scholars suggest adopting a similar approach. For instance, addressing the question of “what makes elections free and fair,” Elklit and Svensson (1997) propose a “checklist of election assessment” that contains over forty independent aspects to be evaluated before, during, and after election day, from basic liberties to impartial dispute settlement. In a later, more explicit effort at translating the checklist approach into a research agenda, Elklit and Reynolds (2000: Table 1), suggest that the systematic study of electoral governance should cover close to fifty “important elements,” from appointment procedures of election officials to the rules governing publication of electoral results. No doubt all the aspects included in such comprehensive checklists look important. But are some more important than others? And how do all the procedural details add up to a final judgment? As Pastor (1998: 159) notes, and Elklit and Svensson (1997: 36) readily admit, checklists do not tell us “how to weight each item on the list” nor do they provide “a formula to aggregate the answers.” In addition to such methodological problems, checklists pose enormous practical challenges. Many electoral processes are of breathtaking administrative and logistical complexity, roughly proportional to the size and heterogeneity of the country in question. The investment of time and money required for collecting adequate data for even minimally systematic analysis would be prohibitive. While indispensable for reaching nuanced judgment on the state of electoral democracy in a given country, the approach is largely unworkable for comparative analysis. (2) The selective approach: This approach restricts analysis to specific issues of electoral governance and is thus more useful for comparative research. Choe (1997) offers a comparative evaluation of election management in South Korea, Sweden, and the United Kingdom. But three countries might be about the largest N a comprehensive study of electoral governance is able to cover. Combining the virtues of analytical depth and comparative breadth will perforce dictate a focus on specific issues. The volume on political-party financing edited by Burnell and Ware (1998) as well as the collaborative work on voter registration edited by Courtney (1991) are good examples of the comparative study of single procedural aspects of electoral governance. (3) The subjective approach: If the complexity of electoral processes and the opacity of systematic irregularities impede in-depth analysis of electoral governance, scholars may rely on the subjective perceptions of the major victims of electoral manipulation, the opposition parties. Accordingly, some authors have advanced operational definitions of democratic elections that rest entirely on actor perceptions. Robert Pastor has been most explicit in renouncing the pretension of assessing electoral processes independently of opposition parties’ own evaluative judgments. A “free and fair” election, he stipulates, “is one in which the major parties all accept the process and respect the results” (Pastor, 1998: 160). Conversely, a “flawed” election is “one in which some or all of the major political parties refuse to participate in the election or reject the results” (Pastor, 1999b: 15).12 Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 19 Substituting the “subjective” perceptions of political actors for “objective” indicators of electoral integrity makes practical sense. A reliance on participant perspectives has been a common practice in the study of other spheres of illicit activity that are also closed to public scrutiny. In corruption research, for instance, the Corruption Perception Index published annually by the NGO Transparency International has turned into the most widely used comparative measure of political and administrative corruption. The index aggregates the perceptions of “international business leaders, risk analysts, and business journalists on the relative degree of corruption in more than fifty countries” (Galtung and Pope, 1999: 275). Developing an analogous Electoral Integrity Perception Index, based on opinion polls of opposition leaders, grass roots activists, journalists, or citizens, could shed light into the black box of electoral manipulation. In the absence of viable alternatives, the relative reliability of actor perceptions provides an additional justification for using them in assessments of electoral quality. When opposition parties hold an electoral process to be free and fair, we have few reasons not to believe them. In their uphill battle against the advantages of incumbency, opposition parties usually have a strong interest in free, fair, and clean elections. Accordingly, they have a strong interest in denouncing acts of electoral manipulation and are not likely to under-report such violations of democratic norms. Still, we should not exclude the possibility that they may overreport them as well. Opposition parties by and large do not boycott elections “out of fear of losing but because they believe no one is listening” (Pastor, 1998: 161). Nevertheless, we should take into account that they may be tempted to exaggerate the phenomenon. Lehoucq and Molina (forthcoming) as well as Eisenstadt (1998) have studied the politics of electoral fraud in historical Costa Rica and contemporary Mexico, respectively, on the basis of legal denunciations of fraud. Yet these authors have been sensitive to the incentives that may drive opposition parties to misrepresent the quality of an election. While Lehoucq and Molina argue that the electoral laws in force deterred parties from formulating frivolous claims, Eisenstadt (1998: 30–36) scrutinizes a broad sample of case files in order to distinguish between generic photocopied “knock-off” complaints and serious claims of irregularities. Thus, even if it seems fruitful to take opposition perceptions as an indirect measure of electoral integrity, we should not forget the difference between conceptual understanding and operational indicators of democratic elections. Collapsing perceptions and realities may lead not only to misjudgments about the democratic quality of an election, but also to abandoning the possibility of studying public perceptions and public discourse in their own right. The comparative study of elite and mass perceptions of electoral manipulation is a field too rich and unexplored to be closed off by operational definitions that blur the distinction between the operational measure of electoral legitimacy and the underlying concept of electoral quality (Schedler, 1999b).13 (4) The indirect approach: Allegations of electoral manipulation become irrelevant when opposition parties win an election. But if the incumbent manages to win, it is often hard to assess the quality of the electoral processes it organizes, and it is also hard to know whether the incumbent would have actually quit power in the case of losing the election. Some authors think we neither know “what might have happened” in the counterfactual case of an opposition victory, nor can we “assess the degree of repression, intimidation, or fraud for each election . . . in a reliable way” (Przeworski et al., 2000: 24).14 They conclude that it is only by Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 20 International Political Science Review 23(1) looking at electoral results that we can learn whether multiparty elections are democratic or not. If the proof of the pudding is in the eating, the proof of democracy is in the alternation of government. Only if competitive elections allow for alternation in power do they provide tangible proof of their democratic character. As long as they confirm incumbents in office, we have to remain skeptical as to their democratic nature, even if they were “considered to be free and fair by observers” (Geddes, 1999: 116). Relying on the substantive results of elections resolves by circumvention two methodological problems. It circumvents the complex challenge of assessing the democratic quality of electoral processes. And it circumvents the complex challenge of assessing the chances that incumbents may eventually reverse electoral results that favor the opposition. The four approaches to studying the process of electoral governance invite pragmatic recombinations, dictated by research objectives. The checklist approach seems appropriate for reaching normative judgment on the democratic quality of elections in a given country. By contrast, reliance on alternation in government allows classification of political regimes for large-N research without having to address complex issues of electoral integrity. For the comparative study of processes of electoral governance, combining the “selective” with the “subjective” approach might be useful. On the one hand, when traveling from case studies based on comprehensive checklists to more comparative studies we need to be sensitive to the perceptions of political parties. Their criticisms may reveal ways of identifying the critical issues of interest. On the other hand, we may double-check allegations of electoral manipulation by studying the points of vulnerability opposition parties are concerned about. If we are interested in the politics (and not just the technicalities) of electoral manipulation, we have to listen to opposition parties. We cannot understand, let alone predict, the political consequences of electoral irregularities without comprehending the discourses and strategies of opposition actors. But if we want to do more than public opinion polling, we should be prepared to check their claims of electoral manipulation on the basis of available evidence in order to reach independent judgment on the quality of a disputed election. Comparative and Historical Perspectives The contributions to this thematic issue of the International Political Science Review undertake the first systematic expeditions into the vast territories of electoral governance charted in this Introduction. Their perspective is comparative and historical, and their purpose empirical and conceptual. Even as they strive to advance our empirical knowledge of electoral governance and democratization, they develop valuable analytical tools to study the phenomenon. Overall, the articles focus on the institutional side of electoral governance, exploring the complex origins as well as the problematic consequences of institutional choices. Revising the electoral history of the Americas, Fabrice E. Lehoucq addresses an issue of paramount importance to the institutional design of electoral governance: Can parties police themselves? His answer is in the negative. The traditional design of electoral governance in presidential systems assigns the responsibility for organizing elections to the executive, and the responsibility for certifying election results to the legislature. Mutual restraint between the two branches is supposed to guarantee the impartial conduct of elections. Lehoucq shows, however, that in the modern world of political parties the nineteenth-century formula of electoral Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 21 checks and balances leads either to interbranch collusion, where one party controls both branches, or to interbranch confrontation in the case of divided government. The author develops a simple game in extensive form to demonstrate the structural propensity to corruption and conflict inherent in the separation-of-power model. In addition, he cites ample historical evidence to support his strong institutionalist argument that leaving electoral governance in the hands of contending parties has historically been a recipe for interbranch conflict and political instability. The article invites a fundamental rereading of the institutional development of the Americas. In neither North nor South did democratic instability grow out of history or culture; rather, it was born out of the failure to institutionalize impartial institutions of electoral governance. Those countries that took alternative institutional paths and established nonpartisan election commissions broke out of the cycle of rigged elections and partisan conflict. Where electoral governance was able to provide the requisite procedural certainties of democracy, politics could turn from armed conflicts over the rules of the game to peaceful competition within the rules of the game. In his article on electoral court failure, Todd A. Eisenstadt delivers a systematic assessment of the autonomy and effectiveness of emerging electoral institutions in Mexico’s federal states. He starts out with methodological reflections on the study of electoral fraud, followed by an analytical synthesis of the complex interactions opposition parties and the ruling Institutional Revolutionary Party (PRI) sustained during Mexico’s protracted transition. He then examines the emerging electoral courts at the state level. Attentive to the possible gap between formal rules and informal practice, he studies both sides of the institutional coin—the legal structure as well as the actual working of the fledgling tribunals. On the one hand, he provides an insightful classification of formal institutional development from the late 1980s to the late 1990s. In three successive waves of legal reform that roughly paralleled the pioneering electoral reforms at the federal level, state electoral tribunals changed from non-autonomous to autonomous institutions with a brief interregnum at the intermediate level of semi-autonomy. On the other hand, Eisenstadt studies the extent to which formal reform was followed by changes in court behavior as well as in party behavior towards judicial proceedings. He finds that legal changes did give way to behavioral changes. Electoral tribunals actually managed to overcome various forms of institutional failure (caused mostly by external interference by the ruling PRI), while opposition parties started to use them as they got more and more autonomous in form as well as in substance. Still, the original data presented by the author reveal an intriguing time-lag between institutional reform and actual compliance. Medium levels of court autonomy multiplied post-electoral conflicts rather than channeling them into the judicial arena. This temporal pattern of institutional development provides an important general insight. Formal institutions matter. But even if they look nice on paper they have to prove in practice that they matter before actors start taking them seriously. New electoral institutions may be weak and ineffective. But they may also be counterproductive, and generate hidden side-effects that run counter to democratic principles. In his contribution, Frederic C. Schaffer analyzes the possible exclusionary consequences of democratic reform. May clean election reforms, he asks, cause citizens to drop out of electoral participation? Drawing upon a wide range of historical and contemporary evidence, he argues that under certain circumstances progressive reforms may indeed have depressive effects on Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 22 International Political Science Review 23(1) voter participation. Furthermore, despite the universalistic justification of legal reforms, their exclusionary effects often show a clear bias against certain groups of voters. Designed to control electoral corruption, legal reforms may remove incentives or set up obstacles to the participation of specific groups of citizens. Schaffer discusses three kinds of regressive effects progressive reforms may engender: the disenfranchisement of opposition voters; the attrition of local party activism; and negative vote-buying (with parties purchasing the abstention of their opponents’ supporters). Such effects are partly intentional. But they are also the unforeseen consequences of decentralized actors adapting strategically to, and creatively subverting, the new legal constraints. Whether incumbents are able to devise biased election laws, the author observes, depends on the ability of party agents to identify and act upon opposition voters. It also depends on prevailing correlations of power. If opposition parties are weak, unchecked incumbents may be tempted to implement legal changes that are neutral in form, but discriminatory in effect. As Schaffer stresses in his cautious conclusions, his findings should not be read as an argument against democratizing reform. The question is not whether to make elections clean, but how to reconcile electoral integrity with electoral inclusiveness. In his article on the structures and origins of election management bodies (EMBs) in sub-Saharan Africa, Shaheen Mozaffar shifts attention from the consequences of electoral institutions to their origins. He first constructs an original database that classifies EMBs in 41 African countries according to their degree of institutional autonomy. He finds that sub-Saharan African countries have followed the global trend towards independent election commissions. Slightly more than half the countries in Africa have established legally autonomous EMBs in the course of their democratic transitions, while over onefourth have established at least semi-autonomous EMBs. Mozaffar then develops an ordered probit model to estimate the effects of four variables on the choice of EMBs: (1) institutional legacies of colonial rule, measured in terms of the anglophone tradition of institutional devolution and decentralized governance versus the francophone and lusophone tradition of statism and centralized governance; (2) political legacies of postcolonial African authoritarian regimes, measured in terms of the frequency of (often restricted) elections and the level of political competition they permitted; (3) structures of ethnopolitical cleavages that shape the power relations and inform the institutional preferences of prodemocracy groups and authoritarian incumbents; and (4) the resulting political negotiations between them. The model generally confirms the expected effects of these variables on the choice of EMBs. Predicted probabilities calculated from the probit coefficients correctly predict the likelihood that an African country will choose a non-autonomous, semi-autonomous or an autonomous EMB due to the separate effect of each independent variable. Anglophone institutional legacies are more likely than francophone ones to influence the choice of autonomous EMBs. High frequency of elections under post-colonial authoritarian regimes is likely to foster the choice of non-autonomous EMBs. But high levels of political competition under postcolonial authoritarian regimes, high ethnopolitical fragmentation, and the incidence of political negotiations over the choice of new democratic institutions all increase the likelihood of African countries choosing autonomous EMBs. To complete this thematic issue, Andreas Schedler opens his article on the dynamics of democratization by elections with a comparative look at the “menu of Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 23 electoral manipulation” ruling parties have at their disposal to control electoral outcomes. His list includes some old themes: fraud, repression, and unfairness. But it also features some issues that have received much less attention: the manipulation of actors and cleavages, and the manipulation of rules of competition. Schedler argues that manipulated elections are deeply ambiguous and thus profoundly contested processes. As a result, rather than establishing an institutional equilibrium, they tend to trigger a self-reinforcing spiral of democratization by elections. Schedler introduces the heuristic model of a “nested” two-level game to capture the interaction between electoral reform and electoral competition that drives such “self-subversive” processes. He outlines the causal interaction and strategic interdependence of the two levels. He describes the basic strategic choices and strategic dilemmas actors face in iterative cycles of conflict. He analyzes the uncertainties of outcomes, relations of power, and strategic responses that characterize the game. Finally, he explains how actors may cope with the ambivalent and uncertain nature of the game, namely, by devising seemingly contradictory strategies and by privileging one level of the game over the other. Notes 1. The International Institute for Democracy and Electoral Assistance (International IDEA) in Stockholm, the International Foundation for Election Systems (IFES) and the National Endowment for Democracy (NED) in Washington, DC, the UN in New York, and the Centro de Asesoría y Promoción Electoral (CAPEL) of the Inter-American Institute of Human Rights in Costa Rica, are among the most prominent international players in the field. Much of the evidence on the political relevance of electoral governance is contained in the reports written by these institutions in connection with technical assistance projects. In addition, they have made laudable efforts to make the enormous amounts of existing materials readily available to scholars and practitioners worldwide. The Administration and Cost of Elections (ACE) Project, jointly sponsored by International IDEA, the UN, and IFES, offers the most extensive electronic collection of electionrelated documents [www.aceproject.org]. 2. See Pierre and Peters (2000) for a more general discussion of governance as structure and process. 3. Our triangle of “conflicting imperatives” is partially inspired by Michael Maley who stresses five fundamental challenges of electoral administration: political neutrality, logistics, accountability, decentralization, and project planning (Maley: 2000: 6–7). 4. This was arguably the only incontrovertible point in the otherwise highly controversial decision of the US Supreme Court that ended the Florida recount and handed George W. Bush the presidency. Specifically, a 7 to 2 majority determined that different criteria used by Florida’s 67 counties to distinguish a valid from an invalid ballot violated the equal protection clause of the Fourteenth Amendment of the US Constitution. 5. See “Ethical Principle 2: Election administration must be non-partisan and neutral” (International IDEA, 1997: 10–12). 6. Public deliberation, political accountability, bureaucratic rationality, and material welfare are some other sources of democratic legitimacy. 7. The paradox can be explained by the bounded rationality of self-interested actors and the veil of ignorance (Rawls, 1971: 136–142) that obscure future outcomes of current choices. Both reduce capacity for accurate calculations of costs and benefits of alternative strategies under new institutional arrangements. Hence, strategically rational actors prefer to limit the discretion of other actors over alternative possibilities. But without resorting to the coercive instruments of authoritarian governance, they cannot do so without limiting their own discretion. Subjecting one’s own interests as well as Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 24 International Political Science Review 23(1) those of one’s adversaries to the uncertainty of competitive elections based on electoral rules agreed ex ante thus restricts the opportunity of all to arbitrarily change unfavorable outcomes ex post (Przeworski, 1991: 12–13, 40–45). 8. On appellate institutions as mechanisms of hierarchical “political control,” see Shapiro (1981). 9. The United States apparently remains the only country in the Americas in which the legislature has the final authority for electoral certification, with the US Congress responsible for congressional and presidential elections and state legislatures for legislative and gubernatorial elections. 10. Sometimes, the consociational principle of proportionality (Parteienproporz) works as an informal understanding that at least each of the major players has a “quota” to fill, as in Venezuela until 1993 (see Molina and Hernández, 1999: Table 2). 11. As in most classification schemes, some countries cannot be assigned to either of the categories. Mixed cases, such as Uruguay, assign the authority to appoint high election officials to various institutions. 12. Elklit and Reynolds take a similar route when conceptualizing the “quality of an election … as the extent to which the entire electoral process is seen as legitimate and binding by political actors” (Elklit and Reynolds, 2000: 2). 13. In addition, delegating the assessment of electoral quality to opposition parties offers little guidance in cases where the “major political parties” differ in their public judgment. In the course of Mexico’s transition process, for example, the right-wing National Action Party and the left-wing Party of the Democratic Revolution have adopted contrasting strategies to cope with electoral manipulation (Gómez López, 2000). 14. Granted, we never know what would have happened in counterfactual scenarios. Still, we usually do know reasonably well what might have happened, considering given sets of actors, correlations of power, and institutional constraints. References Bratton, M. and N. van de Walle (1997). Democratic Experiments in Africa: Regime Transitions in Comparative Perspective. Cambridge: Cambridge University Press. Burnell, P. and A. Ware (1998). Funding democratization. Manchester: Manchester University Press. Choe, Y. (1997). “How to Manage Free and Fair Elections? A Comparison of Korea, Sweden, and the United Kingdom.” Göteborg University. Choe, Y. and S. Darnolf (1999). “Evaluating the Structure and Functional Role of Electoral Administration in Contemporary Democracies.” 95th Annual Meeting of the American Political Science Association (ASPA), Atlanta, 2–5 September. Courtney, J.C. (ed.) (1991). Registering Voters: Comparative Perspectives. Cambridge, MA: Center for International Affairs, Harvard University. Cox, G. (1997). Making Votes Count: Strategic Coordination in the World’s Electoral Systems. New York: Cambridge University Press. Eigen, P. (1996). “Combating Corruption Around the World.” Journal of Democracy, 7(1): 158–168. Eisenstadt, T.A. (1998). “Electoral Justice in Mexico: From Oxymoron to Legal Norm in Less Than a Decade.” Working Paper, Carter Center, Atlanta, GA. Elklit, J. and A. Reynolds (2000). “The Impact of Election Administration on the Legitimacy of Emerging Democracies: A New Research Agenda.” 96th Annual Meeting, American Political Science Association (APSA), Washington, DC, 31 August–3 September. Elklit, J. and P. Svensson (1997). “What Makes Elections Free and Fair?” Journal of Democracy, 8(3): 32–46. Galtung, F. and J. Pope (1999). “The Global Coalition Against Corruption: Evaluating Transparency International.” In The Self-Restraining State: Power and Accountability in New Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 25 Democracies (A. Schedler, L. Diamond, and M.F. Plattner, eds), pp. 257–282. Boulder, CO: Lynne Rienner Publishers. Geddes, B. (1999). “What Do We Know about Democratization after Twenty Years?” Annual Review of Political Science, 2: 115–144. Gill, M.S. (1998). “India: Running the World’s Biggest Elections.” Journal of Democracy, 9(1): 164–168. Gómez López, A. (2000). “Juegos políticos: Las estrategias del PAN y del PRD en la transición.” Facultad Latinoamericana de Ciencias Sociales, PhD dissertation, Mexico City. Goodin, R.E. (1998). “Keeping Political Time: The Rhythms of Democracy.” International Political Science Review, 19(1): 39–54. Gould, A.C. (1999). “Conflicting Imperatives and Concept Formation.” The Review of Politics, 16(3): 439–463. Grofman, B., S.C. Lee, E.A. Winckler, and B. Woodall (eds) (1999). Elections in Japan, Korea and Taiwan Under the Single Non-Transferable Vote: The Comparative Study of an Embedded Institution. Ann Arbor: University of Michigan Press. Gyimah-Boadi, E. (1999). “Institutionalizing Credible Elections in Ghana.” In The SelfRestraining State: Power and Accountability in New Democracies (A. Schedler, L. Diamond, and M.F. Plattner, eds), pp. 105–121. Boulder, CO: Lynne Rienner Publishers. International IDEA (1997). “Code of Conduct: Ethical and Professional Administration of Elections.” International Institute for Democracy and Electoral Assistance, Stockholm. Issacharoff, S., P. Karlan, and R. Pildes (2001). When Elections Go Bad. New York, NY: Foundation Press. Jaramillo, J. (1998). “Los órganos electorales supremos.” In Tratado de derecho electoral comparativo de América Latina (D. Nohlen, S. Picardo, and D. Zovatto, eds), pp. 205–249. Mexico City: Fondo de Cultura Económica. Jones Luong, P. (2000). “After the Break-Up: Institutional Design in Transitional Societies.” Comparative Political Studies, 33(5): 563–592. Lehoucq, F.E. and I. Molina (forthcoming). Stuffing the Ballot Box: Fraud, Electoral Reform, and Democratization in Costa Rica. Cambridge: Cambridge University Press. Lijphart, A. (1994). Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies. New York: Oxford University Press. Linz, J.J. and A. Stepan (1996). Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. Baltimore, MD: Johns Hopkins University Press. López-Pintor, R. (n.d.). “Electoral Management Bodies as Institutions of Governance.” United Nations Development Programme, New York. Luhmann, N. (1983). Legitimation durch Verfahren. Frankfurt/Main: Suhrkamp. Maley, M. (2000). “Administration of Elections.” In International Encyclopedia of Elections (R. Rose, ed.), pp. 6–13. Washington, DC: Congressional Quarterly Press. Maxfield, S. (1999). “A Brief History of Central Bank Independence in Developing Countries.” In The Self-Restraining State: Power and Accountability in New Democracies (A. Schedler, L. Diamond, and M.F. Plattner, eds), pp. 285–292. Boulder, CO: Lynne Rienner Publishers. McCubbins, M. and T. Schwartz (1984). “Congressional Oversight Overlooked: Police Patrols versus Fire Alarms.” American Journal of Political Science, 28(1): 165–179. Molina, J. and J. Hernández (1999). “La credibilidad de las elecciones latinoamericanas y sus factores: El efecto de los organismos electorales, el sistema de partidos y las actitudes políticas.” Cuadernos del CENDES, 41: 1–26. Mozaffar, S. and R. Vengroff (1999). “Institutional Choice as Nested Games in Emerging Democracies: The Case of Senegal.” 95th Annual Meeting of the American Political Science Association (APSA), Atlanta, 2–5 September. Mozaffar, S. and R. Vengroff (forthcoming). “A ‘Whole System’ Approach to the Choice of Electoral Rules in Democratizing Countries: Senegal in Comparative Perspective.” Electoral Studies. Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 26 International Political Science Review 23(1) Neou, K. with J.C. Gallup (1999). “Conducting Cambodia’s Elections.” Journal of Democracy, 10(2): 152–164. O’Donnell, G. (1996). “Illusions About Consolidation.” Journal of Democracy, 7(2): 34–51. O’Donnell, G. and P.C. Schmitter (1986). Transitions from Authoritarian Rule: Tentative Conclusions about Uncertain Democracies. Baltimore, MD: Johns Hopkins University Press. Pastor, R.A. (1998). “Mediating Elections.” Journal of Democracy, 9(1): 154–163. Pastor, R.A. (1999a). “A Brief History of Electoral Commissions.” In The Self-Restraining State: Power and Accountability in New Democracies (A. Schedler, L. Diamond, and M.F. Plattner, eds), pp. 75–81. Boulder, CO: Lynne Rienner Publishers. Pastor, R.A. (1999b). “The Role of Electoral Administration in Democratic Transitions: Implications for Policy and Research.” Democratization, 6(4): 1–27. Pierre, J. and B.G. Peters (2000). Governance, Politics and the State. New York: St. Martin’s Press. Przeworski, A. (1988). “Democracy as a contingent outcome of conflicts.” In Constitutionalism and Democracy (J. Elster and R. Slagstad, eds), pp. 59–80. Cambridge: Cambridge University Press. Przeworski, A. (1991). Democracy and the Market: Political and Economic Reform in Eastern Europe and Latin America. New York: Cambridge University Press. Przeworski, A., M.E. Alvarez, J.A. Cheibub, and F. Limongi (2000). Democracy and Development: Political Institutions and Well-Being in the World, 1950–1990. Cambridge: Cambridge University Press. Przeworksi, A. and H. Teune (1970). The Logic of Comparative Social Inquiriy. New York: Wiley Interscience. Rawls, J. (1971). A Theory of Justice. Cambridge, MA: Belknap Press. Schedler, A. (1999a). “Las comisiones y la pirámide: La recentralización conflictiva del poder en el IFE.” Política y Gobierno, 6(1): 187–222. Schedler, A. (1999b). “Civil Society and Political Elections: A Culture of Distrust?” The Annals of the American Academy of Political and Social Science, 565: 126–141. Schedler, A. (1999c). “Conceptualizing Accountability.” In The Self-Restraining State: Power and Accountability in New Democracies (A. Schedler, L. Diamond, and M.F. Plattner, eds), pp. 13–28. Boulder, CO: Lynne Rienner Publishers. Schedler, A. (2000). “Incertidumbre institucional e inferencias de imparcialidad: El caso del Instituto Federal Electoral.” Política y Gobierno, 7(2): 383–421. Schedler, A. (2001). “Distrust Breeds Bureaucracy: Democratization and the Formal Regulation of Electoral Governance in Mexico.” Public Integrity, 3(2): 181–199. Shapiro, M. (1981). Courts: A Comparative and Political Analysis. Chicago: University of Chicago Press. Taagepera, R. (1998). “How Electoral Systems Matter for Democratization.” Democratization, 5(3): 68–91. Taagepera, R. and M.S. Shugart (1989). Seats and Votes: The Effects and Determinants of Electoral Systems. New Haven, CT: Yale University Press. Temin, J. (2000). “A compelling national election drama in Ghana.” The Sunday Boston Globe, 24 December. Tsebelis, G. (1990). Nested Games: Rational Choice in Comparative Politics. Berkeley: University of California Press. Weber, M. (1988) . “Politik als Beruf.” In Gesammelte Politische Schriften (J. Winckelmann, ed.), pp. 505–560. Tübingen: J.C.B. Mohr. Biographical Notes SHAHEEN MOZAFFAR is associate professor of political science at Bridgewater State College and research fellow of the African Studies Center at Boston Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009 MOZAFFAR/SCHEDLER: Introduction 27 University. He has published widely on the colonial state, democratization, ethnic conflicts, and electoral systems in Africa. ADDRESS: Department of Political Science, Bridgewater State College, Bridgewater, MA 02325, USA [e-mail: [email protected]] ANDREAS SCHEDLER is professor of political science at the Facultad Latinoamericana de Ciencias Sociales (FLACSO) in Mexico City. He coordinates the EGO Electoral Governance Network [http://flacso.edu.mx/ego] and chairs the IPSA Research Committee on Concepts and Methods [www.concepts-methods.org]. He is just completing a book entitled The Politics of Impartiality: Democratization and Electoral Governance in Mexico. ADDRESS: FLACSO, Facultad Latinoamericana de Ciencias Sociales, Camino al Ajusco 377, Col. Héroes de Padierna, Delegación Tlalpan, CP 14200 Mexico City, Mexico [e-mail: [email protected]] Acknowledgements. Mozaffar thanks the National Science Foundation for financial support and the Boston University African Studies Center for research support. Schedler acknowledges support from the Austrian Academy of Sciences through the Austrian Program for Advanced Research and Technology (APART). Downloaded from http://ips.sagepub.com at CAPES on September 15, 2009