62011 WORLD DEVELOPMENT REPORT 2011 BACKGROUND PAPER REPRESENTATIONAL MODELS AND DEMOCRATIC TRANSITIONS IN FRAGILE AND POST-CONFLICT STATES Jack A. Goldstone George Mason University September 18, 2010 The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the World Development Report 2011 team, the World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. Why Foster Democracy in Fragile and Post-Conflict States? The notion that relatively poor pre-industrial countries, including those just recovering from or even struggling with violent conflicts, should be encouraged and aided in quickly setting up democratic institutions is relatively recent. Throughout the Cold War, and up through today’s dealings with such countries as Egypt, Saudi Arabia, and China, most advanced Western industrialized countries were content to deal with, and even support, authoritarian governments provided they were seen to be providing orderly conditions for economic investment and growth, and not posing an immediate military threat to their neighbors or to Western nations. Although acts of blatant suppression of human rights – such as China’s military action against Tiananmen Square protestors in 1989, or Burma’s crushing of the democratic opposition in 1988 – were roundly condemned by the international community, the latter did not generally engage in active promotion of democracy in authoritarian states. Although the U.S. did seek to support dissidents in communist countries, especially in the U.S.S.R. and eastern Europe, there were many voices, including proponents of the “Asian path to growth,” policy professionals, and economists who argued that a period of authoritarian rule was both desirable and necessary for poorer states in order to create state institutions capable of reshaping their economies, promoting economic efficiencies, and ensuring the sustained investments in public goods needed to put poor countries on track for rapid economic development. This position was challenged by two major developments. First was the collapse of the communist regimes of the Soviet Union and Eastern Europe in 1989-91. While many of the post-communist successor regimes, especially those in Central Asia, failed to sustain a transition to democracy, the end of the Cold War was seen by many as demonstrating the failure of the major alternative model of economic and political development. Liberal democracy became identified as the only political system that would lead to sustained political stability and long- term economic success (Fukuyama 1992). The second major development was the steady spread of democratic aspirations and reforms among developing countries, starting in the 1970s and driven by the demands of their own populations for regimes that would replace the notoriously corrupt and ineffective governments of the post-colonial era with more accountable and responsible governments. This trend began with the collapse of dictatorships in Portugal, Spain, and Greece and the expansion of the European Union, and continued with the development of democracy in South Korea and Taiwan, the replacement of military and one-party regimes and the consolidation of democracy in Latin America from Chile to Mexico, and the emergence of democracy in many African states from South Africa to Mali. According to Freedom House, in 1975 the number of countries that were ‘not free’ exceeded those that were ‘free’ by 50%. However, by 1985 the growth in ‘free’ countries was such that they outnumbered those still ranked as ‘not free,’ and by 2007 twice as many countries were ‘free’ as were ‘not free’ (Freedom House 2008). The spread of democracy since the mid-1970s has helped to make democracy a global aspiration, and resulted in the majority of the world’s countries having governments that are at least formally democratic. Yet this progress has not meant that an irresistible trend to greater democracy and thus better governance has taken hold. First, many of the countries that are formally democratic – that is, they have officials chosen by contested elections – nonetheless 1 continue to be plagued by poor government, including high rates of crime, government corruption, grossly inadequate and mismanaged delivery of social services, and unresponsive and unaccountable officials. Formal elections themselves have proven no guarantee of individual rights, judicial independence and impartiality, and checks on executive authority. Instead, a ‘winner takes all’ attitude has often meant that elections exacerbate, rather than reduce, the patronage power of national leaders and the oppression of ethnic or religious groups excluded from power by the winners. Many of the formally democratic nations thus are what Fareed Zakaria (2007) has labeled ‘illiberal’ democracies, which suffer from a variety of deficits in economic and personal security and freedom, which in turn undermine or stifle economic growth. Second, this post-1970s wave of democratization is now showing signs of slowing down or reversing (Diamond 2008). Many of the ‘illiberal’ democracies have remained in that status for a decade or more, showing little sign of improvements in rights, security, stability, and economic growth. A significant number of elections in emerging democracies have resulted in violence, as excluded groups seek to offset or prevent further losses to ‘winner-take-all’ governments, or regimes fearing a loss of power harshly repress their opponents. New democracies from Ethiopia to Zimbabwe, and even countries there were as economically developed and previously democratic as Venezuela and Thailand, have seen democratic institutions curtailed or overturned as winning parties have imposed curbs on opposition politicians and free media, and manipulated election results. The result has been a spread of one- party rule and increasingly authoritarian governance. The 2009 Freedom House survey reports “a continued erosion of freedom worldwide, with setbacks in Latin America, Africa, the former Soviet Union, and the Middle East.” Despite continued efforts by courageous journalists, opposition leaders, and ordinary citizens to extend and defend democracy, the number of electoral democracies has now decreased every year since 2005. While ‘free’ countries still outnumber those that are ‘not free,’ the current 116 ‘free’ governments is the lowest number of such regimes since 1995 (Puddington 2009). The combination of demands for democracy spreading throughout the developing world, plus the impact that the fall of communism had upon Western elites and their belief in the value of democracy, led to a growing consensus among the leading industrialized nations of NATO and the OECD that authoritarian regimes were irredeemably corrupt, and would ultimately be both unstable and economically unsuccessful. Yet it has also become clear in the last decade that progress toward democracy remains limited and uncertain. Thus it seemed evident that the best way to help fragile and war-torn states that needed to strengthen their government institutions was to assist them in building democratic government (USAID 2002). The Hopes for Democracy as a Solution to the Problem of Institutions in Fragile States Building democracy in war-torn and corrupt states seemed to offer hope for addressing all of their key problems. First, democracy would make governments accountable to their citizens for pursuing sound policies, reducing problems of corruption and ineffectiveness. Second, democracy would give legitimacy to governments that had previously based their authority mainly on coercion and patronage to privileged elites, thus building popular support and improving stability. Third, popular participation in inclusive regimes would end discrimination that had previously empowered particular ethnic or religious or regional groups who exploited others. Fourth, the democratic peace theory – and the experience of the West with Japan and 2 South Korea – argued that democratic states, even from diverse historical and cultural backgrounds, do not make war against other democratic states; thus democratization would reduce international conflicts (Doyle 1983, Russett 1993). Fifth, democratic regimes would more likely invest broadly in public goods to benefit the population, especially in human capital formation, rather than dissipate national wealth in corruption or concentrate it among a privileged elite; this would increase economic growth, reduce inequality, and spread opportunity. This view was strongly promoted by Amartya Sen’s arguments regarding the role of democracy in preventing famines and contributing to broader human development (Sen 1982, 1999). Sixth, gains in women’s rights and human rights more generally have usually accompanied the birth and expansion of democratic institutions; thus democracy would also help empower women and other disadvantaged groups. A seventh benefit, raised after U.S. and allied intervention in Iraq and Afghanistan, was that legitimate and popularly-chosen governments would diminish the appeal of, and deny their territory to, terrorist groups. The list above – based in both political theory and a reading of trends in world history since the 1980s – was impressive and enormously persuasive, raising the hopes of both local populations and Western aid donors and interveners. At roughly the same time, economic development theory was shifting away from the ‘neoliberal’ view that emphasized minimal government and a strong private sector to an ‘institutionalist’ view that economic growth, even the private sector, can only flourish under good governance and sound institutions (North 1990; Rodrik, Subramanian and Trebbi 2004; Acemoglu, Johnson and Robinson 2005). Democracy and decentralization were identified as elements contributing to good governance. Democracy promotion thus became a central goal for international agencies, NGOs, and donors dealing with poor, war-torn, or badly-governed states. Indeed, during the 1990s the creation of democratic institutions, often minimally defined in terms of a written constitution and competitive elections, often appeared to be the end goal for the international community: once the first national elections were held, the job of the donor community in regard to political and security intervention was considered complete, and a nation could return to more ‘normal’ relations of economic guidance and assistance with international donors. Unfortunately, these hopes were based on a somewhat rosy-hued interpretation of democracy, without much regard for the longer-term history of democratic development. The United States, in its own history of democratization, moved through over a decade of weak and ineffective government under its Articles of Confederation, seven decades of slavery and a horrendous Civil War, violent labor struggles and massive urban corruption in the late 19th century, and continued racial violence and struggles up through the 1960s. Creating democracy from scratch thus was enormously difficult even in the United States, and many serious social issues and sources of violence remained unresolved. The happier experiences of Japan and Germany after World War II, and the positive experiences in post-communist countries such as Hungary, the Czech Republic, and Poland, areoften pointed to as examples of the virtues of implanting democracy in war-torn states. Yet these latter countries – all of which had substantial prior experience of democratic governance and of industrialization – may have been poor guides to what to expect in efforts to bring democracy to much poorer states with no prior democratic experience. In fact, the history of efforts to develop democracy in poor, war-torn, and poorly- governed states suggests a number of pitfalls for donors seeking to bring stable democracy to such settings. 3 Pitfalls of Democracy Promotion in Fragile and Conflict-Affected States A central problem is deciding what constitutes democracy. The academic literature is divided. Some scholars prefer the simplicity of a narrow definition in terms of electoral processes – a democracy is a country in which top executive leadership and most legislators are chosen by competitive elections open to all (or nearly all) adult citizens (Przeworski et al. 2000), in accord with a constitutional process that protects basic political and human rights Under this definition, gaining acceptance of a constitution and holding the first free and fair democratic elections for national offices completes the task of creating a democratic country. This definition thus creates a fairly clear and unambiguous exit point for those international agencies and donors seeking to build democracy. However, other scholars prefer a richer definition in terms of the functional perquisites of democratic institutions: competitive elections plus freedom of speech and assembly, and transparency, accountability, representativeness, political equality, rule of law, a vital civil society, institutionalized political parties, free media, etc. (Freedom House 2008, Diamond and Morlino 2005). While this definition no doubt captures more of the elements required for the long-term success of democratic governance, it creates several problems for democracy promotion. First, several of the items are fairly resistant to measurement: what constitutes the ‘rule of law’ or a ‘vital civil society?’ Second, even if one could precisely measure all of these characteristics, how much is enough – when have the diverse institutional underpinnings of democratic society matured enough to allow international the community to reduce its efforts? The major indices that purport to measure national levels of democracy with a quantitative scale – the Freedom House ranking (Freedom House 2008), which scales nations from 1 to 7 on civil and political liberties, and the Policy IV scale of regime characteristics (Marshall and Jaggers 2009), which gives nations a score from -10 (most autocratic) to +10 (most democratic) – have both been widely criticized. While both of these scales are in good agreement in placing countries at the extreme ends of their scales – that is, identifying countries as full democracies or clear dictatorships – they are often inconsistent in placing countries in the middle of their scales, and disagree about the size and direction of small movements in democracy (Munck and Verkuilen 2002, Vreeland 2008). This is because while it is fairly easy to classify states where virtually all the components of democracy are in place, or where all are lacking, there is an immense problem of weighting various components of democracy in countries that are partly democratic and making gradual changes. Which should count more toward democracy – an autonomous judiciary or a potent legislature? How much does one count breadth of electoral participation as opposed to limits on executive power? How much does one weight political equality versus equality in the delivery of essential government services? Thus a country with a score of 4 out of 7 on Freedom House, or 6 out of 10 on Polity IV, = has not achieved full democracy, but the deficits could be in any number of different areas. Countries with middling democracy scores may thus vary considerably in their conditions and specific needs, and the meaning of a small move, say from 3 to 4 on Freedom House scales, or from 6 to 7 on Polity IV scores, is dubious, and may simply mean switching one deficit for another. Direction matters as well – conditions may be very different in a country that has a score of 4 out of 7 on Freedom House that is moving toward democracy (was previously a 6 or 7) from conditions in another country with the same score but which is experiencing a deterioration of democracy (was previously a 1 or 2). The consequence of these ambiguities is that existing scales of democracy are at best rough guides to global trends in democratic 4 governance, but are very weak guides to the growth of democracy in particular states, or for framing policies to assist democratic emergence and consolidation in a given country. That is, these scales are the most inconsistent and ambiguous precisely where the practical needs for democracy assessment are the greatest: in states that are making a transition to democracy, and where small year-to-year changes need to be assessed. For practical purposes, democracy assistance thus has generally taken a two-part approach. First, the task of holding free and fair elections is now a fairly well-defined process, which many donors and agencies can help countries implement in a few years – from drafting a constitution to choosing an electoral commission to carrying out voting and tabulating results. Such electoral assistance and election monitoring is now a readily available option and fairly widespread and consistent element of international assistance to support democratic governance. However, the second part – developing the broader functional perquisites of democracy, including a strong civil society, rule of law, political equality, fighting corruption, and provision of essential education, health, and infrastructure – is far less well defined and uniform. These tasks now are undertaken by a wide variety of bilateral and multilateral donors through diverse programs that vary from donor to donor and state to state. These goals involve more extensive social and political changes, which we do not have well-established methods of carrying out and in most countries can take many years of persistent efforts to bear fruit. Donors thus have come to recognize that concentrating on an initial set of elections would most often lead to only short-term success and leave longer-term problems unresolved. An increasing body of empirical literature suggests that transitions to electoral democracy that occur in the context of unresolved factional differences, or in a state with weak rule of law and corrupt state practices, leads to further political instability and negative impacts on economic growth (Goldstone et al. 2010, Polterovitch and Popov 2007). On the other hand, an open-ended commitment to providing assistance until a country reaches a level of democracy that most Western countries themselves have reached only after a century or more of development is not reasonable or practicable. Thus democracy promotion in fragile and war-torn states typically aims for free elections plus some degree of further social and political change – but how much more, and in what areas, depends on the particular threats to democracy and stability in the specific country receiving aid. The success of democracy promotion thus lies less in following any particular academic model of democratic processes than in the degree to which aid policies are responsive to addressing and offsetting the specific obstacles to democracy in a particular country, and particularly to how well they empower and assist the local groups and trends promoting democratic government. A major issue in democracy assistance is thus how much effort is made to fit nascent democratic institutions to the conditions within a given society, as opposed to simply implementing a narrow (just elections) model or an idealized Western-type set of democratic institutions (secular law, centralized government, all-elected officials). In practice, donor agencies often implement plans based mainly on what they have learned to do in other countries, rather than tailoring their assistance to current and past conditions in each society. This is particularly problematic in that the international democracy assistance community consists of highly varied donors – both multilateral and national donor organizations, and variety of specialized democracy-assistance NGO’s (e.g. the National Democratic Institute, the International Republic Institute, IDEA) with differing experience and strategic views. Yet a good fit with local conditions is vital if democratic institutions are to have the desired degree of legitimacy. After all, while electoral office-holding has become the ultimate basis for legitimacy 5 of political power in most Western nations, it is far from being that in many poorer and fragile states. Tribal office, hereditary rank, and religious sanctity all are competing, and often far more potent, sources of legitimate authority. A more sound, stable, and legitimate democracy may often develop in such cases by incorporating traditional elements of legitimacy. For example, the sultans in Malaysia, tribal leaders in South Africa, and hereditary lords in Great Britain were all given roles in their countries’ developing democratic institutions, albeit subordinate and integrated into broader electoral-driven decision-making institutions. As Francis Fukuyama observes, “It is sometimes preferable to work within the context of imperfect existing institutions than to spend political capital [and enormous, possibly wasted, energy and money] on long-term institutional reforms” (Fukuyama 2007, 39). Rather than conferring legitimacy in and of themselves, electoral procedures may need to gain legitimacy themselves from working with, and within, local norms of traditional legitimacy. A further problem in ensuring that democratic institutions acquire legitimacy is attention to the likely outcome of elections – whether they will in fact produce meaningful participation by a broad public, or polarization in favor of particular groups and disaffection in others. For example, in Iraq the first national elections in 2005 seemed certain to produce such large majorities in favor of the Shi’ite parties that it discouraged Sunnis from participating, and Sunni communities engaged in combinations of formal boycotts by leaders and simple disaffection and non-participation by ordinary Sunnis, undermining the legitimacy of electoral outcomes. After the elections sectarian warfare between Sunnis and Shi’as greatly escalated rather than decreased. In Palestine and Algeria, the introduction of elections into societies that had already been somewhat polarized by Islamic parties exhorting religion-based politics led to victories at the polls by Hamas in Gaza and the Islamic Salvation Front. In the former case, the result was renewed hostilities with missile attacks on Israel prompting an Israeli armed incursion in response; in the latter the result was the seizure of power by the military and a civil war with Islamist groups producing over 100,000 deaths. These latter cases illustrate the high risks involved in developing democracy. Elections and transitions to democracy are often highly destabilizing and violence-producing events, rather than straightforward pathways to stability. The rich democracies are accustomed to think of democratic competition as producing incremental shifts in the balance of power among well- established parties, which have ingrained a willingness to share power and work with the opposition in victory, and to expect a future return to power when in defeat. The situation is wholly different in countries where competing political, regional, or ethnic groups have a history of hostility and view each other with extreme distrust and suspicion. Where parties fear that defeat will mean future exclusion from all political power and economic opportunity, where elections are seen as ‘winner-take-all’ licenses for the victor to act with impunity, where groups that have long enjoyed power and treat it as their right feel threatened with summary dismissal by the voters, then it is unlikely that elections will unfold as they do in Western contexts. More often one of the following scenarios will unfold: the party in power, fearing loss of power in electoral defeat, does all it can to hinder opponents and intimidate opposition voters, stuff ballot boxes, and manipulate voting to ensure a large victory (as with Marcos in the Philippines); or the government will declare the election void or cancel it before voting begins claiming a security threat (as in Algeria); or the group that loses the election (whether current rulers or their opponents) will reject the vote as invalid and launch a military coup or a violent campaign to take the power it feels it rightfully deserves (as in Chile in 1973). All of these outcomes undercut the legitimacy of the government and usually lead to more 6 violence. Thus the hope that building democracy will sharply reduce the risks of violence in fragile and war-torn societies must be qualified. One of the best-evidenced findings of comparative politics is that while both full autocracies and fully institutionalized democracies are quite stable, transitional and intermediate regimes – sometimes called ‘anocracies’ (Fearon and Laitin 2003), ‘partial democracies’ (Epstein et al. 2006, Goldstone et al. 2010), or ‘illiberal democracies’ (Zakaria 2007) – are extremely unstable, being highly prone to reversion to authoritarianism or civil war (Mansfield and Snyder 2005). This is precisely because introducing truly competitive elections into a situation of hostility and fear creates uncertainty, not only over who will win, but over what will be the consequences of that victory. Such uncertainty often leads either or both the government and opposition to seek to control the outcome by other means. In countries with long-standing regional divisions – such as Pakistan and Nigeria – society has bounced back and forth between efforts at democracy building and bouts of military dictatorship for many decades. In countries where a group that has enjoyed a dominant role is threatened with a reversal of fortune under a new democratic government, as with the Arabs in regard to the south and southwest of Sudan, or the Hutu in Rwanda under the agreement that President Habyarimana signed with the Tutsi-led Rwandan Patriotic Front, fear of a future adverse democratic outcome can even propel efforts at genocide. Moreover, a significant number of regimes have learned to follow formally democratic procedures, holding elections with ‘approved’ oppositions, but then manipulate the resources, media coverage, and opportunities of opposition actors so effectively that these elections are not truly competitive. Such semi-authoritarian regimes, e.g. Egypt,Venezuela, or Russia, can be very stable, but do not represent true progress toward democratic government. Indeed, these regimes function like, and tend to increasingly become, fully authoritarian states, with increasingly concentrated and unchecked executive power. Finally, even if all other obstacles are overcome and an elected government peacefully takes office, this provides no guarantee that governance will improve. From 19th century American urban ‘bosses’ to 21st century Nigeria, democratic regimes can often be as corrupt as non-democratic ones. Many formally democratic regimes today score very low on the Transparency International scale of perceived corruption. Thus Haiti, Venezuela, Yemen, Sierra Leone, and Kenya are all ranked by Freedom House as “Partly Free,” and Ukraine is ranked as “Free.” Yet all of them are ranked among the very worst countries in regard to corruption: the first three are respectively scored as 168th, 162th, 154th, and the last three tied at 146th, out of 180 nations. Indeed, the opportunities to buy voters’ loyalties and the expenses of running for office mean that democratic politicians are always desirous of money. The large number of people involved in decision making often means that many more people have to be bought off to make deals for decisions to be made at all, and the intoxication of popular support sometimes makes politicians feel entitled to enrich themselves. In many regions, the prevailing norm that good fortune should be used to provide betterment for one’s family and friends translates into even properly-elected politicians seeking ways to profit from office or direct government largesse on the basis of personal interests (Polterovich and Popov 2007, 77). Moreover, democratic governments can also be notoriously ineffective. Divisions in elected legislatures can lead to policy paralysis; populist measures can lead to show projects and spectacle rather than hard decisions and long-term investments being made. Democratic governments in poor countries are often severely underfunded, having difficulty raising taxes or enforcing them; their fiscal weakness makes it difficult for them to recruit and keep talented 7 administrators or conscientious law-enforcement and security personnel, and to supply needed public investments and services to their populations. Poor countries often have suffered as much economically from disarray and disorder under democratic regimes as under dictatorship: Nepal, the Philippines, and Nicaragua gained little in legitimate or effective government from ending their monarchy or the Marcos or Somoza dictatorships. In short, the period of transition to democracy in a poor, war-torn, or badly-governed country is fraught with risks. Even if donors can agree on targets for elections and other political and social changes to support democracy, local sources of legitimacy may compete with electoral institutions for authority; elections may intensify and polarize existing social cleavages; anxiety about the conduct and outcome of elections may precipitate illicit or violent attempts to control the elections’ outcome; these may in turn give rise to more widespread violence and instability, fomenting a return to dictatorship, or even civil war or genocide. Even success in implementing democracy may not produce much improvement in governance; regimes in new democracies have commonly been prone to corruption, indecision, and ineffective rule. Representational Models and Democratizing Trajectories Given the issues raised above, one may wonder why donors still make democracy promotion part of their goals. Yet while the hopes that democratization would provide a swift and easy solution to state failure or fragility were naïve, that does not mean that democracy never yields benefits. Indeed, if only a portion of the hoped-for benefits can be realized, even if the time and costs of doing so are greater than expected, the net gains would still be worthwhile. South Africa, Indonesia, Northern Ireland, Mozambique, Peru, Colombia and Turkey are a few of the places where democratic solutions to political conflicts have been both politically and economically successful, and make a democratic solution more appealing than an authoritarian one. In all these cases democratic movements developed within these countries, which then received external support and assistance in resolving conflicts and institutionalizing democratic practices. These cases suggest that where indigenous actors lead the process, and external donors are sensitive to specific local needs, successful outcomes can arise (Kaplan 2008, Ghani and Lockhart 2008)). We can point to two particularly important issues that domestic and international democracy reformers must face: the timing or sequencing of elections in relation to other political and social changes, and the choice made among different electoral systems. Timing and Sequencing: Donors frequently ask when a country is ‘ready’ for a transition to stable democracy. In fact, there is a contentious academic controversy over what Thomas Carothers has called the “sequencing fallacy” (Carothers 2007, Carothers et al. 2007). Some scholars argue that before a stable and liberal democracy can develop, there must exist a competent state, some progress toward rule of law, and reasonable security. Ideally, a moderate level of economic development has taken place, and the state is not unduly empowered by the existence of oil revenues or other secure and autonomous sources of funding that allow it to disregard the needs and support of the populace. However, in response Carothers has argued that this form of thinking can lead to embracing autocrats who build strong states, but then are reluctant to share power and themselves become obstacles to democracy and sources of instability. He contends that the correct course is to build as many elements of democracy as possible, if only gradually – rule of law, civil society, representative institutions, political parties, free media – to limit executive authority and build expectations of popular participation as soon as possible, and not to wait for certain levels of economic development or state effectiveness to 8 be reached prior to any efforts at democratization. Rather than ‘sequencing’ one should consider ‘prioritizing’ various elements of change in making progress toward democracy, where priorities reflect local opportunities and needs and no single sequence dominates. In sorting through this debate, several points of consensus emerge. First, it is pointless to try to hold elections until a reasonable degree of security has been achieved for most of the country. A lack of security can mean ongoing conflicts, such that the government does not control key areas of its territory, and cannot assure safety against attack to its citizens. Opposition forces may then disrupt, intimidate, and seek to control election results in areas that the government does not control. However, a lack of security can also mean that government military and police forces do not function mainly to protect citizens, but instead to protect leaders against any opposition (Wiatrowski and Goldstone 2010). Opposition politicians and voters are then not secure against threats of government-sponsored violence. Under such conditions, elections will not produce legitimate outcomes, nor have any effect in resolving conflicts. Prior to holding elections, then, it is crucial that the vast majority of regions enjoy conditions in which politicians and voters of all parties can engage in campaigning and voting with sufficient security that the legitimacy of the election and the participation of the electorate are not severely undermined. This does not mean that a country has to be completely at peace to hold elections; but it does mean that international and local security forces are capable of providing protected spaces for political activity that will allow reasonably free and fair elections to occur. Second, progress toward democracy is a process, and not a sequence of events. That process involves shifting attitudes and behaviors, not merely attaining deadlines in regard to passing a constitution or holding elections. People must learn to trust political parties to reasonably represent them, political leaders must learn to trust each other and their opponents in order to participate effectively in elections and democratic governance, and security forces must be trusted to provide equal protection and follow the law rather than foster the dominance of particular elites. Corruption must be gradually curbed from being endemic to a degree that undermines all government efforts to provide services and effective administration to tolerable and diminishing levels, so that the government wins support for its actions and can carry out policies. These goals require support for diverse measures, ranging from civic education and support for civil society, parties, and media operations to trust-building diplomatic exercises, security-forces training, and training and guidance for legislators, jurists, and executive personnel. Structures of incentives must be arranged that reward elites for progress toward democracy and rule of law and hinder corruption and arbitrary extensions of executive power. These goals rarely are reached by a ‘once-and-done’ approach; rather, they require gradual learning of new attitudes and behaviors, which must be reinforced by a series of modest successes and supported through inevitable setbacks. Third, attitudes and patterns of interaction among elites are a major factor in whether democratic transitions will help resolve conflicts, or exacerbate them. Where hostility and mistrust across groups and their leaders mean that uncertainty is likely to lead to defensive and pre-emptive moves to secure position, the inherent uncertainty of democratic institutions and procedures is likely to precipitate conflicts. In this situation, which is extremely common in war- torn and multi-ethnic societies, steps to help build elite cooperation and mutual confidence should precede actions to invest national power by electoral competition. Such steps could include constitutional conventions, interim governments formed to include leaders from diverse groups, and the election of local and provincial governments that provide opportunities for experience with election rules and allow electorally-savvy leaders to emerge. Unless political 9 leaders are comfortable submitting major policy decisions and the choice of national leaders to democratic processes, such processes are likely to break down. Fourth, popular support for democracy is prone to virtuous or vicious cycles. If new democratic governments are able to provide some immediate benefits, such as improved security, economic recovery, reduced corruption, some social services (medical or educational), or responsible justice and law enforcement, they will win support that allows them to function more effectively. They can then increase their provision of such benefits, winning more support, and building stability. Conversely, if new democratic governments appear ineffective by offering no perceived improvements or benefits, or show themselves to be as corrupt and unresponsive as prior governments, the population can quickly become disaffected and turn to local strongmen or opposition groups that do offer such benefits. It is thus vital that some immediate benefits or services that improve people’s lives follow very quickly as part of the movement toward democratic institutions, to build confidence in the new order. These initial benefits can then buy time to further improve governance and increase benefits later. As one example, efforts to install democracy in Haiti were ineffective when donors simply restored Aristide to power then declined to follow-through with security and economic assistance; later efforts to strengthen democracy have been more successful because they focused on immediately restoring security to the most lawless areas as part of aiding the new government. In sum, there is no shortcut or optimal pattern of assistance that will quickly bring about democratic change, nor that is certain to succeed. The best chances of success seem to lie in working with close attention to local conditions, the motivations of local leaders and politically- relevant groups, and the resources and opportunities available, and patiently pursuing progress on multiple dimensions of democracy. Often, it will take years of support for pro-democratic forces and incremental institutional changes until a sudden breakthrough (as in South Africa in 1994 or in the Ukraine in 2004) occurs. Representation and Electoral Systems: One important tool in building democratic systems that are attentive to local conditions is the design of representation, in regard to both electoral systems and government institutions. It is clear from history that a wide variety of electoral systems and government institutions are compatible with good governance and stable democracy (see Reynolds, Reilly and Ellis 2005). OECD countries exhibit both presidential and parliamentary systems of executive authority (the former with a strong separation between a president with executive authority and the legislature, the latter in which executive authority is vested in the parliamentary leadership). OECD countries also exhibit both centralized and federal systems of authority, and employ a wide variety of electoral systems for translating voter choices into the selection of election winners. These include first-past-the-post single-delegate districts (FPP), federal-unit based voting (as in America’s electoral college system), proportional representation (PR), and mixed systems such as Japan’s, in which some legislative delegates are chosen by winning elections in local districts while others are chosen based on their party’s share of the national vote. The choice of system often depends more on a nation’s history and particular compromises made among leaders at the time of democracy’s inception than on any effort to optimize representation or outcomes. Nonetheless, so long as elites and voters find the system works for them, stable democratic outcomes have persisted under diverse systems. In developing democracies in fragile and war-torn states, the choice of electoral systems offers opportunities. As Donald Horowitz (2003) has argued, electoral systems do not simply passively translate votes into election outcomes, they actively shape the political environment 10 and strongly affect voter preferences. Yet they must be chosen with clear awareness that every option involves trade-offs. For example, in a country with strong regional cleavages, federal systems are often more effective than central systems in providing checks on the authority of the largest or traditionally dominant group. In a country with strong and asymmetric ethnic divisions, proportional representation can help assure that even minority groups obtain representation. Yet PR also runs the risk of intensifying ethnic cleavages by increasing the political importance of communal identification and translating ethnic identity differences into struggles for political power. Simple PR systems are often politically demanded in divided societies. Yet they are a particularly poor choice in countries with strong religious or ethnic groupings and where one communal group forms a large majority of the population, as that group’s demographic weight can make it impossible for other groups to ever hope to lead a government. Much the same is true of FPP single delegate systems, which can amplify social cleavages by turning small majorities in certain regions into political dominance by a single party. Both systems are best suited to countries where political divisions are shifting and based on policy views rather than on much less flexible ethnic or religious identities. Multi-stage elections, in which a run-off between leading candidates is required if no candidate obtains a clear majority in the initial round, are both more expensive and time- consuming than plurality-victory elections, but can help produce more legitimate victors and allow supporters of minor candidates to have a decisive role in the final choice of leader. Similarly mixed systems, which require political parties to win seats by different means (some awarded by district victory, some by national proportion of votes), can give opportunities to diverse constituencies to win representation, but are less likely to fracture society along a single set of divisions. As we noted earlier, Iraq’s initial national elections in January 2005 turned out poorly. They are in fact an example of a poor choice of electoral system to fit local conditions. These elections used a system of simple proportional representation, treating the entire country as one district, with the legislature’s 275 seats allotted to competing parties according to the percentages they won of the total national vote. It was hoped that this system would not only provide broad participation but also favor representation by women and minority groups, who could pool their votes for parties that would represent them. In fact, these hopes were disappointed. Announcement of proportional representation intensified the tendency for political parties to form along ethnic lines, with Sunni, Shi’a, and Kurdish parties seeking votes from those communal groups. As a result, the initial election became less a debate on policy than a contest for group dominance, with a predictable victory for the Shi’a community as by far the largest population group. Sunni’s were so disaffected with the outcome that they withdrew their support for the government, and sectarian conflict increased rather than diminished. Only in the last couple of years, with deliberate efforts by the U.S. to have Sunnis play a greater role in local defense and encourage cross-group political alliances, has the prospect improved of Iraqi democracy working to bridge communal differences rather than increase them. And as of yet, the crucial issue of how voting will take place in the region of Kirkuk – which is contested by both Kurdish and Sunni groups – remains unresolved. An alternative voting system that might have worked better would have combined proportional representation with the stipulation that parties needed to win a certain percentage of votes in a minimum number of provinces or regions in order to qualify for their seats. This would have required parties to build cross-regional and multi-ethnic support bases from the 11 outset. Electoral systems requiring candidates to obtain a certain proportion of votes in diverse regions have been implemented in Nigeria and Indonesia and have been fairly successful in breaking down previously sharp regional differences that had produced violent conflicts. It should be clear that the goals in choosing an electoral system are often in conflict. For example, the goal of representation is often best served by proportional representation with low thresholds for parties to obtain legislative seats. However, this choice can reduce political stability, as many small parties can lead to a fractured, indecisive government in which splinter parties hold vetoes or can bring down weak coalition governments. A strong two-party system, which often emerges from first past the post systems, can produce greater stability and favor outcomes with true majority support; yet such systems almost invariably deny representation to minority groups that are widely dispersed. For these reasons, the recent trend in democratic systems is to create more complex hybrid systems, combining national and federal characteristics, and proportional and local- district methods of choosing national officials. Thus Indonesia adopted a mixed system for its 2009 legislative elections, in which candidates could win office either by winning district elections or by their party winning a significant share of the residual votes not going to the winning candidate. The hazard here is that too much complexity hurts transparency, and the understanding and acceptance of the legitimacy of outcomes by voters. However, the Indonesia elections were considered the most successful to date in its democracy, with voters feeling they had more choice and more say in this dual system (formally, an open-list proportional system with popular vote), and considerable success in regard to overcoming the separatist violence in Aceh (ANFREL 2009). Perhaps the most logical statement to be made regarding electoral systems is that there are an immense variety of approaches available to constitution and election designers, which allow a great deal of flexibility in choosing electoral systems and state institutions that best fit the needs and obstacles to democratization existing in specific countries. Attention to historical inheritances, existing sources of legitimacy, the major political cleavages, and the demands of local and national leaders all need to be taken into account in order to design institutions and election systems that will be most workable in specific conditions. The ‘best’ should never be the enemy of the good. After all, the United States Constitution – the world’s longest enduring and to many the most successful – was cobbled together through a series of compromises that resulted in outcomes that were not foreseen by anyone. The method of aggregating votes for the U.S. president – by a majority of electors chosen by majority vote in each state, rather than by a majority of the national vote or of votes in the legislature – is not one that would be chosen today. But it reflected concerns at the time of inception and has worked well enough in practice for two centuries that there are not overwhelming pressures to change it. At the same time, while the U.S. constitution was widely imitated in Latin America, most of those emulations proved unsuccessful. The U.S. constitution – both at home and in emulations abroad – has shown itself prone to the steady shift of power from the legislature, originally envisioned as the most important branch, to the executive; and in many countries Presidential systems have devolved into authoritarian rule. To a significant degree, any electoral and democratic system is capable of being diverted or ineffective, depending on the actions of elected leaders. Especially in presidential systems, leaders who are determined to assert their authority can whittle away at checks and balances in a constitution, increase control over voters and voting processes, strengthen the military, and hem in the opposition. Conversely, in a parliamentary system, selfish leaders of non-majority parties 12 can hobble coalitions, create recurrent instability, and block legislation. It takes very strong and independent judicial and federal systems with their own leaders to maintain checks on executive authority in presidential systems; and it takes cooperative and flexible political leaders to make a multi-party parliamentary system effective and avoid deadlock. These considerations suggest that just as important as the electoral and political institutions chosen to advance democracy are the qualities and commitments of democratic leaders. The most successful transitions to democracy in poor and war-torn states have been led by leaders who exhibited a rare combination of qualities: charisma and leadership capacity combined with a self-deprecating desire to be known more for leading their countries to stable and inclusive democracy than for advancing their personal power or that of their political group. Nelson Mandela in South Africa and George Washington in the United States are the prototypical examples. In the absence of strong and effective leaders who are committed to democracy, transitions are often less successful, succumbing to corruption or authoritarian leanings. Conclusion: Building Democracy in Fragile and Failed States Scholarship and history in regard to democratic transitions in poor and war-torn states offer few blueprints for success, but do offer several principles that can usefully guide efforts to promote stable democratic rule: (1) Democratization offers many potential benefits, but also many substantial risks. It should therefore be approached not as a quick or simple solution to the problems of fragile/failed states, but as a strategy that requires patience, care, and close attention to local conditions over a sustained period of time. (2) Prior to national elections, it is essential to create a climate of sufficient security and cooperation that parties and voters can engage in campaigning and voting without severe risks, and that political leaders will accept the results of free and fair elections regardless of the outcome. (3) Democratic institutions and electoral systems should be developed to take advantage of existing sources of legitimacy, rather than collide with them. They should also be designed in anticipation of likely outcomes, with choices of institutions and electoral systems aimed to minimize the greatest obstacles to democracy and build on historical and local traditions promoting cooperation. Electoral systems and institutions should be designed with particular care to avoid outcomes that are likely to concentrate power in the hands of a particular group or individual, and to encourage cross-group cooperation. (4) Efforts to promote democracy will often need to be gradual and occur in steps, pursuing democratization on a wide range of dimensions, including institutional reform, supporting parties, free media, and civil society, confidence-building, and skill training. (5) Efforts to promote democracy need to occur in conjunction with efforts that enable the new democratic government to offer immediate and visible improvements to the population – whether these occur in provision of justice, services, jobs, security, or reduction of corruption. These are vital to building short-term credibility and support for new democratic institutions, otherwise non-democratic actors may step in to provide desired but missing services and undermine the nascent democracy. (6) In the longer term, efforts should be designed to improve government legitimacy and capacity. This includes the capacity to acquire revenue and project authority and provide 13 security over the whole of national territory; improvements in providing impartial and timely justice; reduction of corruption; improved relations between military and police forces and the population; broader provision of infrastructure, health, and medical services; and a secure legal and enforcement framework for expanding employment and economic growth. Even in the experience of the OECD countries themselves, democracies tend to improve slowly over time, rather than arrive fully complete and competent. It is more important to help countries keep moving in the right direction than to rush or grow impatient regarding the creation of democratic rule. References Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2005. “Institutions as the Fundamental Cause of Long-Run Growth.” In Philippe Aghion and Steven N. Durlauf, eds. The Handbook of Economic Growth. San Diego: Elsevier, pp. 385-473. ANFREL. 2009. A Decade of Democracy in Indonesia: The 2009 Legislative Election. Report of the International Election Observation Mission, Asian Network for Free Elections. http://www.anfrel.org/report/indonesia/2009/Indoesia_General_Election_2009.pdf Carothers, Thomas. 2007. “The Sequencing Fallacy.” Journal of Democracy 18:12-27. Collier, Paul. 2008. The Bottom Billion. Oxford: Oxford University Press. Diamond, Larry. 2005. Assessing the Quality of Democracy. Baltimore, MD: Johns Hopkins University Press. Diamond, Larry. 2008. “The Democratic Rollback: The Resurgence of the Predatory State.” Foreign Affairs 87(2): 36-48. Doyle, Michael. 1983. “Kant, Liberal Legacies, and Foreign Affairs.” Philosophy and Public Affairs 12(3): 205–235 De Soto, Hernando. 2000. The Mystery of Capital. New York: Basic. Epstein, David, Robert Bates, Jack A. Goldstone, Ida Kristensen, and Sharyn O’Halloran. 2006. “Democratic Transitions.” American Journal of Political Science 50(3):551-569. Fearon, James and David Laitin. 2003. “Ethnicity, Insurgency, and Civil War.” American Political Science Review 97(1): 75-90. Freedom House. 2008. Freedom in the World 2008. Charts and Tables: Historical Country Ratings http://www.freedomhouse.org/uploads/fiw08launch/FIW08Tables.pdf Fukuyama, Francis. 1992. The End of History and the Last Man. New York: Free Press. Fukuyama, Francis. 2007. “Development and the Limits of Institutional Design.” In Natalia Dinello and Vladimir Popov, eds., Political Institutions and Development. Cheltenham, UK: Edward Elgar, pp. 21-42. Fukuyama, Francis, Thomas Carothers, Edward D. Mansfield and Jack Snyder, and Sheri Berman. 2007. “The Debate on ‘Sequencing.’ Journal of Democracy 18:5-22. Ghani, Ashraf and Clare Lockhart. 2008. Fixing Failed States. New York: Oxford University Press. Goldstone, Jack A., Robert Bates, David Epstein, Michael Lustik, Monty Marshall, Jay Ulfelder and Mark Woodward. 2010. “A Global Forecasting Model for Forecasting Political Instability.” American Journal of Political Science, 54(1):190-208. Kaplan, Seth. 2008. Fixing Fragile States. Westport, CT: Praeger. Mansfield, Edward D. and Jack Snyder. 2005. Electing to Fight: Why Emerging Democracies Go to War .Cambridge, MA: MIT Press. 14 Marshall, Monty G. and Keith Jaggers. 2009. Polity IV Project: Political Regime Characteristics and Transitions 1800-2008. www.Systemicpeace .org/ polity/polity4.htm. Munck, Gerardo L. and Jay Verkuilen. 2002. “Conceptualizing and Measuring Democracy: Evaluating Alternative Indices.” Comparative Political Studies 35(February):5-34. North, Douglass. 1990. Institutions, Institutional Change and Economic Performance. Cambridge: Cambridge University Press. Polterovich, Victor and Vladimir Popov. 2007. “Democratization, Institutional Quality and Economic Growth.” In Natalia Dinello and Vladimir Popov, eds., Political Institutions and Development. Cheltenham, UK: Edward Elgar, pp. 72-100. Puddington, Arch. 2009. “The Freedom House Survey for 2009: The Erosion Accelerates.” Journal of Democracy 21(2): 136-150. Przeworski, Adam, Michael E. Alvarez, Jose Antonio Cheibub, and Fernando Limongi. 2000. Democracy and Development. Cambridge: Cambridge University Press. Reynolds, Andrew, Ben Reilly, and Andrew Ellis. 2005. Electoral System Design: the New International IDEA Handbook. Stockholm: International IDEA. Rodrik, Dani, Arvind Subramanian, and Francesco Trebbi. 2004. “Institutions Rule: The Primacy of Institutions over Geography and Integration in Economic Development.” Journal of Economic Growth 9(2):131-165. Russett, Bruce. 1993. Grasping the Democratic Peace. Princeton: Princeton University Press. Sen, Amartya. 1982. Poverty and Famines. Oxford: Clarendon Press. Sen, Amartya. 1999. Development as Freedom. New York: Alfred A. Knopf. United Nations, n.d. UN Iraqi Electoral Fact Sheet. http://www.un.org/news/dh/infocus/iraq/election-fact-sht.htm U.S. Agency for International Development. 2002. Foreign Aid in the National Interest: Promoting Freedom, Security, and Opportunity. Washington, D.C.: USAID. Vreeland, James. 2008. “The Effect of Political Regime on Civil War: Unpacking Anocracy.” Journal of Conflict Resolution 52(June):401-424. Wiatroski, Michael D. and Jack A. Goldstone. 2010. “Democratic Policing and Democratic Development.” Journal of Democracy, forthcoming April. Zakaria, Fareed. 2007. The Future of Freedom: Illiberal Democracy at Home and Abroad. New York: W.W. Norton. 15