Perspectives on the Theory of State-Building

PERSPECTIVES ON THE THEORY OF STATE-BUILDING 1. The problem of gaps in an international system of sovereign states Max Weber (1919) famously defined the state as an organization "that (successfully) claims the monopoly of the legitimate use of physical force within a given territory." Political analysis of international relations has conventionally assumed that the entire habitable world is divided into a "Westphalian" system of independent nations, each of which is governed by a sovereign state that exercises supreme authority for maintaining order and enforcing laws within its territory.1 Economic analysis has regularly assumed that all traded commodities are in the domain of such states which reliably enforce property rights and contractual agreements for everyone. These assumptions have been reasonably accurate in much of the modern world, where they have provided fundamental conditions for peace and economic growth. But these assumptions do not apply in the ill-governed territories of weak or failed states, which can generate conflict and human suffering that can flow across international borders into other states. To understand the problems of such weak states, and to formulate effective policies for mitigating these problems, we must think more fundamentally about how nations develop from anarchy to prosperity and what can be done to promote such political development. This paper offers a general conceptual framework for analyzing the problems of weak states and statebuilding. Under the norms of the Westphalian international system, each state's supreme authority within its recognized national borders should be respected by all other states. Although the exact placement of a national border may depend on generally forgotten accidents of history, people can be motivated to react strongly against even a tiny violation of their nation's border by fear that a weaker response could create an expectation of their willingness to surrender much more territory without a fight. Since World War II, a global norm against the use of military force for territorial expansion has promoted the inviolability of international borders and thus has helped to deter wars of conquest. The expanding membership of the United Nations has helped to strengthen these norms, as new states realized their common interest in diplomatic cooperation to assert and defend the sovereign rights that were promised by Westphalian ideals (Herbst 2000). The Westphalian international system can be understood as a decentralized mechanism 1 See Osiander (2001) for a discussion of how the concept of an international system composed of sovereign states came to be named after the 1648 Peace of Westphalia that ended the Thirty Years' War. 1 for controlling violence in the world. Each state's ability to exercise control in its territory is supposed to be a prerequisite for international recognition of its sovereignty, so that each state should be able to take effective responsibility for controlling violence within its borders. If any state allowed its territory to be used as a base for attacks on other states, then the recipients of these attacks would be expected to make a proportional retaliation against the responsible state. Then threats of such retaliations should sustain an equilibrium where each state ensures that no international violence comes from its territory.2 But this mechanism fails when an internationally recognized state lacks the capacity to control violent actors on their territory. Where government is weak and ineffective, criminal violence may turn residents into international refugees, or militant insurgents may establish bases for international terrorism. In such situations, it would obviously be counterproductive to take punitive retaliation against a national government that is already too weak to control its own territory. To solve the problems of a weak state, other nations of the world should help to strengthen the state, so that it can take effective responsibility for controlling violence in its territory. But an international intervention to promote political development would be a direct violation of the Westphalian norm against interference in the internal political affairs of other countries. So the Westphalian international system does not provide a general mechanism for filling gaps in the system caused by weakness or failure of its member states. Thus, the problem of formulating international responses to the problems of weak states raises fundamental questions in political science. First, what can the international community do that would actually promote positive political development in a country? This question raises issues in the theory of comparative politics, as it demands more understanding about how strong states can develop from weak ones. Second, how should the basic norms of the international system be revised to provide a mechanism for responding effectively to the problems of weak states? This question involves fundamental issues in international relations, which should be guided by principles that are enforceable and help to minimize costs of international conflict. These are the central questions of this paper. Many weak states in the world today depend substantially on the current norms of international relations. A weak state can use international respect for its sovereignty to extract valuable rents from its recognized authority over foreign trade and foreign assistance, even when 2 See Fearon and Laitin, 1996, for an analysis of such an intra-group policing mechanism for deterring intergroup violence. 2 it lacks the ability to exercise effective authority over much else within its national borders. In other historical periods, if a state was too weak to control its territory then it was liable to be invaded and conquered by a stronger state. In this sense, the existence of so many weak or fragile states in the world today can be seen as a result of the norm against wars of conquest. We should not regret this norm, as it can be credited with reducing international violence by deterring invasions for territorial expansion. Furthermore, wars of conquest could also destroy states that were not too weak to maintain internal order, but were just not strong enough to prevent a more powerful state from invading them. In the world today, international interventions to promote political development can also be a primary cause of state weakness and fragility, as a country can be torn apart when different nations intervene to support rival factions for power there. In many cases, the main resistance against a state-building mission has come from groups that also have foreign support, from other nations which favor the establishment of a different kind of state. Interveners on one side may justify their actions as support for a legitimate government, while interveners on the other side may justify their actions as support for freedom fighters, but the overall result of their international rivalry is to exacerbate conflict in a weak state. Thus, any general solution to the problem of weak and failed states in the global order must include some principles for strictly regulating international interventions for political development. An international norm that requires foreign political interventions to get some form of broad prior international approval could help to reduce the possibility of rival interventions that exacerbate conflict in the target state. More generally, standards for state-building interventions should be based on an understanding of how successful states actually develop. Some may question the possibility of benevolent state-building interventions, which would support a nation's political development while respecting the ultimate sovereignty of its people. But to have any hope for planning such interventions, or for holding their planners to account, we need some understanding of what should come first in building a strong state that will effectively serve the interests of its people. 2. Fundamentals of political leadership States are established by political leaders, who organize groups of people and mobilize them for action toward shared goals. An intervention to support the establishment of a state must ultimately enable certain political leaders to take power in the state, but it may also try to create 3 an environment in which new leaders and new forms of leadership can develop. So to get a deeper understanding of state-building, we should begin by thinking more fundamentally about leadership. Ancient social philosophers perceived that the basic foundations of a state may depend on its leaders' reputations for rewarding those who serve them. According to Xenophon's Education of Cyrus, Cyrus the Great founded the Persian Empire with one essential quality of leadership: a reputation for reliably rewarding good service. The captains who flocked to Cyrus's banner were confident that they would be well rewarded for serving him in battle, and their recognition of Cyrus as the reliable paymaster of a great army made him a leader who could build a state which was then the greatest in the world. To compete for power in any society, a political leader needs the active voluntary support of many people in a group or faction, and these supporters must be motivated by some expectation of future reward in the event of their success. But when rivals have been defeated, a leader may be able to enjoy the fruits of power without such broad support, and so an established leader may be tempted to ignore the claims of past supporters. Thus, a successful leader must somehow be credibly committed to reward those who have supported him in the past. This is the central moral-hazard problem of political leadership.3 This central moral-hazard problem can be solved within a leader's own faction of supporters, without relying on any external court. A leader becomes politically accountable to a group of supporters when he could not hold power without their confidence in his promises of rewards for good service. This accountability can be effective when these key supporters can monitor their leader's distribution of rewards and share any evidence of his denying appropriate rewards to any of them. Myerson (2008) showed that, in negotiation-proof equilibria of a simple model of sequential contests for political power, a contender for power would be unable to credibly recruit any supporters if he did not organize them in this way, so that he would fall from power if he failed their trust. To motivate appropriate leadership with such accountability, the status of being a recognized leader must entail privileges and expected long-term rewards that are worth more than any short-term benefits that the leader could get by abusing the powers of his position. Then the standards of behavior which a leader must maintain to keep his supporters' confidence 3 In any transaction where some participants are relying on others to act in some stipulated way, the need to provide incentives for these individuals to act as stipulated is called moral hazard. 4 can effectively become a fundamental political law or personal constitution for the leader. Provisions from a written constitution become enforceable against powerful national leaders when their violation of these provisions would cost them the trust of their essential political supporters. A trusted leader can mobilize the members of a group to work for goals that benefit all of them. But the members of such a group can enjoy the benefits of effective leadership only when they all agree about who is their leader. Thus, the question of whom to recognize as leader will have the form of a coordination game which has multiple equilibria. Under the focal-point effect of Schelling (1960), anything in the culture or history of the group that focuses attention on one particular candidate for leadership can become a self-fulfilling prophecy, as nobody would want to deny a leader who is recognized by everyone else. In effect, the problem of reaching consensus about who leads a group is a coordination problem to solve all other coordination problems for the group. The importance of history and culture in determining who is identified by a group as its leader should warn us that efforts to change the leadership of a group are likely to be frustrated unless they are appropriately rooted in the group's traditions. Even in groups that use formal elections to select their leaders, those who seek a change in leadership generally need to coordinate on one alternative candidate to challenge the incumbent; and in this coordination problem, the identity of the focal challenger may also depend on factors of history and tradition. For example, a group may have a tradition of selecting its leaders from a distinguished family, as long as a candidate can be found there with the essential quality of reliably rewarding good service and support. 3. Moral hazard and foundations of the state Beyond political leadership, a strong state also needs thousands of reliable officials who will implement the state's policies and enforce its laws. But in managing resources for public goods and punishing people for wrongful behavior, state officials exercise powers that could easily be abused for their own personal gain. For every state official, there is a moral-hazard problem of giving the official an incentive to act appropriately according to the laws and policies of the state. The establishment of an effective state depends on solving all these thousands of moral-hazard problems. Thus, in the terms of modern economic theory, state-building is essentially about solving a complex system of moral-hazard problems. 5 Furthermore, a well-functioning state also helps to reduce moral-hazard problems in all kinds of social and economic transactions, as people may rely on the state's system of justice to help enforce contracts and laws that stipulate appropriate behavior in private relationships. So the failure of a state is a national crisis of moral hazard which involves a loss of trust in many kinds of relationships. To solve moral-hazard problems in the state, any state official must expect that the rewards for good service can be better than what the official could get by abuse of power. The rewards from good behavior that must be expected by such an official, to deter any abuse of power, are called the moral-hazard rents for this office. Becker and Stigler (1974) considered the problem of motivating government officials to serve appropriately during a long career in which opportunities for abuse of power can arise every day. Becker and Stigler showed that the cost of moral-hazard rents for such officials can be minimized by an incentive system where large late-career rewards are promised for officials who maintain a record of good service, but any evidence of bad service would be punished by early dismissal without such rewards. In such incentive systems, someone must be actively monitoring observable indicators of an official's actions and judging whether the official's service has been worthy of promotion or dismissal. The straightforward way to do this is to make the state official administratively accountable to a supervisor in a hierarchical chain of command. In a well-functioning hierarchy, a higher official should hold subordinates responsible for (among other things) correctly supervising their own subordinates, so that the discipline to apply laws and policies of the state can be conveyed down the chain of command. So in this structure, incentives throughout the state's hierarchical networks will depend on the officials at the top. An office at the top of a chain of command in the state is a position of leadership. The wide influence of leaders implies wide opportunities for abuse of power and thus requires correspondingly large moral-hazard rents, and so leaders in high office must be highly motivated to keep the rewards and privileges of power. The leaders who occupy these high positions will generally be politically accountable to wider groups of people, who together form the selectorate for their positions. This term selectorate is due to Bueno de Mesquita, Smith, Siverson, and Morrow (2003), and they suggested that most forms of political accountability may be well approximated by a model in which an incumbent leader can retain his office as long as he has the support of some minimal fraction of the selectorate. In a democratic state, high offices are allocated by popular elections, and then the selectorate would include everyone who can vote in 6 these elections. In other political systems, the selectorate from which a leader needs support may include only the members of certain elite groups. But even in a state without any formal constitutional mechanism for electing its highest officials, an individual can exercise authority as a leader of the state only if many others are prepared to recognize and comply with this individual's authority. Thus, even an authoritarian leader who claims an absolute right to power may actually be dependent on a selectorate that includes the key government officials through whom his power is exercised. We have noted that responsible state officials under the top leadership must be motivated by long-term promises of late-career rewards for a record of good service. But promises of deferred rewards become a debt owed by the state, which its leaders could be tempted to repudiate (by falsely alleging bad service late in an official's career). So leaders have a fundamental credibility problem. A leader of the state cannot exercise power unless his subordinates trust his promises of long-term rewards for good service. However, the fact that the leader's power depends on his subordinates' trust can become the basis for maintaining that trust, when subordinates collectively monitor how the leader treats them and would stop trusting the leader if he ever denied anyone's appropriate reward. In this sense, even when the top leader of an authoritarian state is not accountable to anyone outside the state, the leader will be politically accountable to an internal selectorate that consists of state officials who together have the ability to suppress opposition and keep their leader in power. Of course the factional supporters who have helped a leader to win power and the state officials who help the leader to wield power are generally not disjoint groups; many key individuals may belong to both groups. Leaders of states throughout history have recognized the efficiency of rewarding some of their most important political supporters by giving them patronage appointments to powerful offices that are associated with valuable moral-hazard rents. As we have noted, the essential quality that any political leader needs is a reputation for reliably rewarding good service and support both from his administrative subordinates in the state and from his active supporters in the political selectorate. Members of the selectorate to whom high officials are accountable can have greater confidence in the state's commitment to protect their basic rights, including property rights. So a larger selectorate increases the number of people who can have confidence in the state's legal protection for their property, which is essential for them to make private investments that can 7 enrich themselves as well as the local economy.4 Thus, it should not be surprising that most of the richest nations in the world are democracies where high officials are politically accountable to selectorates that include the entire adult population. In an authoritarian state where the selectorate only includes elite government officials, the state's reliable protection of property rights may be limited to the essential protection of rights for those who have been promised moral-hazard rents by the state itself. An alternative perspective on organizational leadership was offered by Alchian and Demsetz (1972), who argued that the chief executive at the top of the chains of command in a firm should be motivated by a stake of ownership in the firm. However, Alchian and Demsetz were writing in the literature of economic theory, where the role of the state in enforcing ownership rights is commonly taken for granted. When there is a well-functioning state, firms and other economic organizations can implicitly rely on the state for enforcement of contracts and property rights, so that political leaders serve as the ultimate guarantors for incentive systems in all kinds of organizations. But when we are analyzing incentives in the state itself, or in political organizations that exist to take power in the state, we cannot simply assume that ownership rights and contracts will be enforced by some external court. An attempt to apply the Alchian-Demsetz theory of the firm to a national state would yield something like a classically conceived monarchy, where the state has a supreme ruler who is ultimately motivated by inherited ownership rights over the entire nation. But even in historical periods when many nations theoretically adhered to principles of hereditary monarchy, the reality was that most monarchs were highly constrained by the possibility that, if they acted in a way that alienated too many of their essential supporters, then their "right" to rule could be contested by other potential claimants to the throne. Thus, even a hereditary monarch may be better understood as a leader with political accountability to some selectorate. The concept of a state that has only one central leader, an autocrat with supreme authority over all officials of the state, may seem appealing as a way to guarantee the consistent coherence of state policies. But such a concentration of all political accountability in one supreme office has some fundamental disadvantages. It implies that many local officials will necessarily be separated by a long chain of command from political accountability at the center; and such a centralized administration may be much less responsive to the needs of the local population than 4 The state's ability to make credible promises of protection to individuals in a large national selectorate may depend on the establishment of an effective judicial system, which can function so that a failure of state officials to uphold the rights of an enfranchised individual would become publicly known to others in the selectorate. 8 a local administration that is headed by a mayor or governor with political accountability to a local selectorate. Furthermore, when a country has only one politically accountable official, then the selectorate cannot compare the incumbent leader's performance to any alternative candidate with a record of political accountability in public service. Competition in national politics is strengthened when mayors and governors prove their qualifications for higher office by providing better public services in local government; but local officials are much less likely to establish themselves as credible challengers to the incumbent national leader when they are appointed by and administratively accountable to the national leader. Thus, we should expect significant differences between centralized states where mayors and governors are administratively accountable to the national leadership and decentralized states where mayors and governors are politically accountable to local voters. More generally, a strong modern state can increase the scope and range of political accountability in the state by having many offices that are politically accountable separately to different selectorates in the country. Then, however, the state has a fundamental problem of minimizing the potential for conflict among such officials with separate political accountability. This problem is solved in successful modern states by constitutional rules that define the limits of power for each politically accountable office. To support the effectiveness of these constitutional rules, people in the selectorates for politically accountable officials should be inclined to distrust any of these officials who violated the constitutional limits of their office. So in the establishment of a strong constitutional state, critical tasks of state-building include formulating a functional system of such limits in the constitutional distribution of powers and cultivating a broad appreciation of these limits by members of the national selectorate. The development of bureaucratic offices and procedures can reduce the costs of moral hazard in many functions of the state, as systems of record-keeping in state offices make many abuses of power harder to conceal.5 Official malfeasance can become harder to conceal also when responsibilities are assigned to committees or working groups, if their members would have difficulty trusting each other in a joint violation of the state's administrative rules. But an effective state bureaucracy requires more than just educated personnel who can manage sophisticated record systems and who bring norms of propriety to their teamwork; it also requires fundamental support for administrative rules from the state's political leaders. As we have seen, 5 Indeed, writing seems to have been invented for this purpose by the earliest states in antiquity (Scott 2017). 9 an essential function of top state leadership is to serve as the ultimate guarantors of administrative incentive systems for lower-level state officials in the hierarchical chain of command. If political leaders do not support the standards for evaluating and rewarding the service of administrators in government bureaucracies, then these standards cannot be maintained against temptations of corruption. In particular, the state's political leaders may be more willing to tolerate corruption in offices that they have allocated as patronage rewards for their political supporters, unless there is some countervailing political pressure from citizens who pay the costs of this corruption. In a well-functioning state, some offices have the function of auditing the operation of other offices in the state. The supplemental information (horizontal accountability) that is provided by such independent auditing agencies can help to reduce the cost of moral hazard in agencies that provide public services. In a weak or fragile state, these independent auditing services may be relatively amenable to foreign assistance, because foreign auditors would be deputized only to provide information about official actions and would not have any direct power over private citizens. However, the services of other kinds of government agencies could be seriously degraded by foreign supervision. When the state has failed in one country, there may be an illusion that order and public services could be restored by an intervention that brings in a team of trained administrative officials from a successful state elsewhere in the world. Certainly, military units are organized to maintain their discipline even in challenging circumstances far from home, and so military forces can be used to restore basic order in a failed state. But while an intervention could try to maintain the lines of administrative accountability for intervening officials who have been sent abroad, the lines of political accountability would fundamentally change in the translation to a foreign country. The leaders who supervise an international intervention are politically accountable in the nation that sent the intervening forces, not in the country that they have occupied. Officials who are charged with maintaining order in a community may behave very differently depending on whether people in the community can vote to elect the political leaders who supervise these officials. So the political deterrents against abusive behavior by government officials can weaken or vanish when they are transferred from a successful democratic state to a country where nobody has any power to elect the intervening officials' political supervisors. 10 4. Questioning the Weberian ideal of a state monopoly on the use of force There was something fundamentally misleading about Hobbes's (1651, ch. 13) view that, in the absence of a state, people would be in a war of everyone against everyone. People have generally lived together in communities with social structures and kinship networks ever since our species evolved, long before the first states were established. After the failure of a state, social networks of trust may become shorter and sparser but they do not vanish. The absence of an effective state means that people have greater need to rely on local groups that can offer some basic security and protection within their community, and these groups must develop some capability for forcefully defending their interests. So the process of state-building is not performed on a blank slate, but must entail a shift in authority from various forms of local leadership to the leaders of the new state. To establish a Weberian monopoly on the use of force, a new state would have to suppress or subordinate all these autonomous local groups on which people have relied for protection. Local leaders of these groups will have a vested interest in preserving their power, and people who have found protection from such a group could have realistic concerns about whether their vital interests will be protected as well by the new state. Thus, an attempt by leaders of the new state to eliminate any group's independent capability for self-defense is likely to elicit widespread resistance from such groups and could indeed be against the interests of large segments of the population. In a new state where political accountability of its leaders has not been tested, people may have realistic concerns about what kinds of abuse they could suffer if they abandon their capability for self-defense and allow the new leadership to take monopolistic control of the use of force. From this perspective, both the feasibility and desirability of the Weberian ideal in statebuilding should be seriously questioned. Indeed, the Second Amendment was added to the United States Constitution in 1791 because many Americans did not want their new federal government to get a monopoly on the use of force. People generally have an interest in minimizing the use of violent force in their society, but a reduction of violence may be achievable without putting all capacity for violent force under the control of one faction. So we may distinguish three general paths for political development from a weak or failed state, depending on whether the Weberian ideal is to be abandoned, or applied with compulsion against the resistance of local groups, or applied only with broad consent from local groups throughout the nation. Abandoning the Weberian ideal means accepting that various local factions will maintain some autonomous capacity for armed action, and then the immediate 11 goals of state-building should be to encourage peaceful power-sharing among these armed factions in an oligarchic political system. Applying the Weberian ideal with compulsion means accepting that the leaders of the new state may forcefully disarm and suppress any group that resists their monopoly on the use of force; but then there may be no impediment to their establishing an autocratic state. Applying the Weberian ideal with consent means requiring that the state's monopoly on the use of force should be established only with political controls that can assure people in all major groups that their vital interests will be represented and protected in the new state, which then may be truly called democratic. The next three sections consider these three paths of political development: oligarchic, autocratic, and democratic. 5. Regulation of violence in an oligarchy North, Wallis, and Weingast (NWW, 2009) have written extensively about the political development of states where many local groups have autonomous capacity for physical force. Indeed, such failures of the Weberian ideal are so common in most of recorded history that NWW use the term "natural state" to denote a state where many groups have autonomous capacity to use violent force. These armed groups have leaders, like Xenophon's Cyrus, whose ability to exercise force depends on their reputations for reliably rewarding the fighters who serve them. These leaders can offer protection to others in their communities through patronclient relationships. The integrity of the oligarchic state is then defined by a mutual recognition among these leaders that they are members of a national oligarchic elite with norms and understandings that help them to co-exist peacefully. Substantial moral-hazard rents are needed to deter these armed groups from using violence to expand their wealth and power at others' expense. That is, to give these armed groups an incentive to help keep the peace, their leaders must have some valuable stake in the national order. Thus, the system may be stabilized by an agreement among the national elite ro not contest each others' privileges of monopolistically exploiting certain public or private resources in the country. In a natural state where social peace depends on such anti-competitive rent-sharing agreements among the leaders of armed groups, reforms that aim to increase economic and political competition by eliminating monopolistic privileges could actually have the perverse effect of increasing political violence, so that people may be worse off than before. To maintain the oligarchic peace, top national officials of the state must be trusted to support this system of elite privilege, and so they must have some effective political 12 accountability to the oligarchic leaders, who in turn may be politically accountable within their respective factions. But when one faction increases its ability to exercise force against others, some corresponding re-allocation of moral-hazard rents in favor of the leaders of this faction may become essential, to keep the peace and deter them from applying their new capability. Thus the stability of an oligarchic political system may require some mechanism for periodically redistributing privileges in proportion to changes in the underlying distribution of force. This dependence of oligarchs' privileges on their capacity for violence can make it difficult or impossible to develop effective nonpolitical security forces for a state that is dominated by such oligarchic factions. (See Matisek and Reno, 2019.) Factional leaders would be drawn into a competition to co-opt officers in the state's security forces, who command a capability for violent force that a faction could use to increase its wealth and power. Then as influential oligarchs demand promotions for officers who have aligned with them, officers without any factional affiliation may be passed over regardless of the quality of their service. Ultimately, an identification of senior military officers with competing political factions can undermine their military effectiveness against external threats, as factions could anticipate some advantage from an external enemy destroying forces that are affiliated with their internal rivals. The stability of an oligarchic natural state depends on maintaining some general agreement among the oligarchic elite about the appropriate distribution of power and privileges among the various factions. Foreign assistance could help to stabilize an oligarchic political system by providing neutral arbitration for disputes between factional leaders. For example, we may consider the role of the podesta in the government of medieval Genoa after 1194, when its internal politics was dominated by powerful clans. (See Greif, 1998.) The podesta was a professional judge and military leader who would preside over the oligarchic Genoese state for a one-year term, supported by twenty soldiers and two judges that he would bring with him. Each year, a council that was broadly representative of all Genoese factions would select a new podesta from outside Genoa. There were restrictions to prevent the podesta or anyone in his extended family from owning property in Genoa or forming any special connections with anyone in Genoa by marriage or commercial transactions. Then the podesta's payment at the end of the year was contingent on his maintaining internal peace, not altering the balance of power among the major factions, and acting impartially in any disputes among them. The point here is that foreigners can play a positive role in an oligarchic state by offering a certifiable neutrality with respect to the state's internal politics. Thus, all Genoese factions 13 could rely on a foreign podesta to help resolve their disputes with impartial arbitration. In this system, sovereignty was vested in the indigenous council that selected the podesta, but it might have been much more difficult for this council to reach any decision if it were not choosing among such certifiably neutral leaders for their state. 6. Centralized autocratic state-building The danger of factional divisions among the state's military officers can be minimized by giving one top official the power to oversee all appointments of senior military officers, as their commander-in-chief. But then, under Weber's assumption that the state should have a monopoly on the legitimate use of physical force, there could not be any legitimate capacity for resistance against the forces under this commander-in-chief. Even if the state includes other senior officials who have separate political accountability to some broader selectorate, they could have no capacity to defend themselves against the commander-in-chief's monopoly of force. So a Weberian state with a coherent unity of command in its security forces would lack any effective protection against the commander-in-chief becoming an authoritarian autocrat. In this sense, autocracy seems almost unavoidable for a coherent Weberian state. As we have noted, the absence of any effective state would tend to make people more dependent on local leaders for basic protection, and then building a coherent Weberian state would require the suppression or subordination of all these autonomous local leaders. Such a broad liquidation of local leadership structures is unlikely to occur without brutal force being applied in at least some parts of the country. There is a straightforward strategy for a foreign intervention to support the establishment of an autocratic state. The interveners just need to identify someone who can be an effective manager of security forces and then provide him the resources necessary to recruit and equip forces that can defeat any opposition. If this intervention is not countered by competing investments in rival contenders for power, then the leadership that it installs may dominate the country for many years. Our concepts of oligarchic and autocratic states can be seen as two ends of an interval of mixed structures that differ in how many state officials hold positions with autonomous political accountability. In a pure autocracy, every state official except the autocrat would be administratively accountable to a superior in a hierarchical chain of command that leads ultimately to the autocrat at the top. Then the only political accountability is that the autocrat 14 must keep the confidence of at least some minimal fraction of a selectorate which includes the key state officials whose responsibilities require some credible promise of long-term moralhazard rents. In contrast, an oligarchic state would distribute power and privileges among the various oligarchic leaders, each of whom has autonomous political accountability to supporters in some faction or clan or constituency. Any new state, to be effective, must build networks of reliable officials who will implement the state's policies and enforce its laws in every part of the country. When one disciplined organization with coherent central leadership develops such a national network by sending out agents to supervise the local implementation of state policies, the result can be an autocracy (perhaps with a central committee in place of an autocrat). The establishment of the Soviet Union may be described in this way (see Fainsod 1958). But a new state can also develop its administrative networks by the inclusion of various forms of local leadership that have been providing security and local public services in parts of the country while a strong state was absent. Then a key question is to what extent will these local leaders continue to hold their positions based on local political recognition within their constituency, and to what extent will their power and privileges be dependent on recognition from the central authorities. Incentives for improving local public services can be sharpened by making local leaders politically accountable within their constituency. But as we have seen, the potential for conflict within the state can be exacerbated when powerful state officials are accountable to separate constituencies, at least until a generally accepted constitutional distribution of powers can be established. The top national leaders in a state may have concerns about losing central control of local officials who develop a personal power base in their province and so become too entrenched to be removed. To prevent such entrenchment, a common policy is to rotate centrally appointed officials among different provinces during their careers, so that local networks of trust do not become focused on the individual official. But in an autocratic state where local investors have negligible influence in the national selectorate, the prospect of such rotation would weaken a local official's incentive to encourage and protect long-term investments that are vital for economic growth in any community. We can understand from agency theory that an efficient way of motivating local officials to cultivate long-term economic development in their community would be by promising them a substantial long-term ownership stake there, or by making them politically accountable to people who have 15 such stakes of ownership in the community. A common way for large autocratic states to steer between these alternative hazards of entrenchment and disinvestment is to distinguish two levels of local government, which we may call districts and provinces, where the lower-level district officials are long-term residents of their districts, but the higher-level provincial officials are centrally appointed and are regularly rotated to different provinces during their careers. By elevating one local notable in each district to be its local leader, with wide local powers, a new state can create a nationwide class of agents who have a stake worth defending in the state and thus can be given basic responsibility for maintaining order in their district. But the privileges expected by such local elites will become debt obligations that constrain the state, which its national leaders could ultimately be tempted to deny. So the longterm credibility of these moral-hazard rents for lower-level district officials may depend on their identification as a class with certain generally recognized privileges, so that a national leader who denied such privileges to one member of the class could lose the confidence of many others. This effect can be strengthened by institutions that facilitate communication among local officials from different parts of the country, so that they can collectively function more effectively as a vital part of the selectorate for national leaders. Thus, the institution of Parliament strengthened the English monarchy after 1300 by assuring local gentry of their essential place in the structure of the state. 7. Challenges of democratic state-building To maintain an international norm against conquest, the international community may appropriately demand that a foreign intervention should not act to promote political change in a nation without some indigenous approval from people in the nation. But the effectiveness of this constraint depends on who is accepted as speaking for the nation. After all, something very much like conquest could be imposed on a nation if a forceful intervention could be justified by an invitation from a puppet ruler, whose position of authority is maintained by the intervention itself. Any state-building intervention could concentrate power in the hands of a small group of elite officials who would then be expected to express their approval of the intervention. Thus, although the word of a high official may be accepted as the sovereign voice of the nation for most questions in international relations, it is not enough when the nature of the state itself is the subject of international negotiations. The establishment of a new state by an international 16 intervention can be justified only by broad approval for the state from some globally recognized national selectorate. In the twenty-first century, it is hard to imagine a global consensus about sovereignty being vested in any selectorate smaller than the nation's entire adult population. Thus, a state-building intervention should be constrained to respect democratic sovereignty in the nation, which can be expressed only by freely contested votes of the population at large, with international monitoring to verify that the intervener is not manipulating the elections. The basic goal of a democratic state-building intervention should be to establish a political system in which different factions will be able to compete democratically for power, and the intervention itself should ideally be neutral among these factions. But from the moment when this new state is established, it must have one national leader or chief executive, to supervise the national government and provide some coherent coordination of its policies. So any financial assistance that interveners provide to the new government will be spent under the direction of this leader, who could then use this assistance to start building a national network of supporters with patronage jobs from the state. Thus, even when a foreign intervention has established an interim provisional government, to exercise power only until elections can be held, foreign assistance to the new government may tend to strengthen the faction of the provisional national leader, and so it may not be politically neutral. This non-neutrality can be reduced, however, if the provisional government functions under a parliamentary system where the national leader is politically accountable to a council or assembly that is broadly representative of all major factions. That is, instead of simply recognizing a provisional leader of the new state, it might be better for international interveners to recognize a provisional assembly that has the power to replace the national leader at any time. With such a parliamentary system, we may expect patronage benefits to be shared among a group of factions that together have a majority in the assembly. The non-neutrality of foreign assistance can also be reduced when some portion of the assistance is distributed to autonomous local councils that can choose local chief executives from different political factions. But this is only one of several reasons why federal decentralization of power may have fundamental importance in democratic state-building. In the aftermath of a weak or failed state, when many people may have had experience of neglect or oppression by their national government, promises of better rule from the new state are just unproven possibilities, while basic protection from local groups may be a proven reality in people's lives. So people in many communities may have reason to believe that their vital 17 interests can be protected better by local armed groups than by whoever is likely to win national leadership in the new state. Then the suppression of popularly trusted local leadership would be against the wishes of that part of the sovereign population that has relied on them. In this practical sense, the principle of intervening for political change only with consent of the people should be applied first and foremost to any reform of local government. People may be willing to trust the political processes of the new state only if some power and responsibility for local government is retained by trusted local leaders in their communities. The identification of such popularly trusted local leaders ultimately requires some form of local elections for offices of autonomous local government. Thus, in practice, democracy with some federal decentralization of power seems essential for state-building when the state's monopoly on force is to be applied with broad consent from local groups throughout the nation. The LSE-Oxford Commission on State Fragility, Growth, and Development (2018, p17) recommended that the first step of developing from a weak or fragile state should be to build checks and balances that restrain powerful officials, so that people can be less fearful of cooperating with an effective state. But adding checks on the state may seem to be a step in the wrong direction when the goal is to build a stronger and more effective state. This paradox may vanish, however, when we see that strong autonomous local governments can constitute a fundamental check on national leaders, especially when some share of the state's monopoly on force is exercised by these locally elected authorities. So the Commission's recommendation could be reinterpreted as suggesting that a first step in political development could be to establish autonomously elected local governments, with a functional division of powers and responsibilities between these local governments and the national government. Any democratic system has the disadvantage that the possibility of incumbent leaders losing the next election reduces their ability to promise long-term moral-hazard rents to their agents and supporters. The countervailing advantage of democracy is that the possibility of taking power democratically in a forthcoming election can also give other groups a stake in the political system. With centralized national democracy, however, only a few national factions can have any realistic hope of exercising power in the near future, and there may be many parts of the country where nobody has any expectation of getting a share of power in the national government. Insurgencies can take root in such regions where local leaders feel alienated from the state. An effective system of democratic local government can ensure that every part of the country has some elected local leaders who have both a proven ability to mobilize political 18 action in their community and a real stake of power in local government. People are more willing to fight for a state when such service is perceived as a defense of the social order in their community, so that their service may earn them honor and status at home. But such a connection between service to the state and status in one's community cannot be expected in communities whose leaders are alienated from the state. Democracy, like oligarchy, involves open political competition among rival factions. However, such factional rivalries may be kept out of the state's armed forces when democracy is established well enough to ensure that the distribution of power in the state will depend on election outcomes and not on factions' capacity for violence. But fully effective democracy may take years to develop in a country. Democratic elections can function correctly only when competing candidates feel constrained to abide by democratic norms, which is generally true in well-established democracies where voters would be shocked by any leader who tried to corrupt the electoral process or used violence against political opponents. In a failed state that has no past experience with democracy, however, the leading contenders for high office would have no prior history of competing fairly in elections, and indeed many political leaders could have a history of using violence in past confrontations with opponents; and so their violations of democratic norms in new elections would be unlikely to shock many of their supporters. Then in high-stakes elections, there would be no one who could deter powerful leaders from stuffing ballot boxes or using violence against their opposition when they have the ability to do so. Thus, the development of effective democracy may depend on the gradual rise of new leaders who have earned voters' trust by respecting democratic norms and so have a reputational commitment to democracy. Local elections can be a vital source of such new democratic leadership. We should understand that effective democracy requires more than just national elections. There have been many cases of autocrats securing confirmation of their rule by an election or plebiscite in which competitive alternatives were suppressed. The public benefits of democratic competition depend on an excess supply of well-qualified democratic candidates as alternatives to incumbent leaders. Thus, successful democratic development requires an ample supply of leaders who have good reputations exercising public power responsibly to benefit the public, not just to provide patronage rewards for a small group of supporters. This supply of trusted democratic leadership can develop best in autonomous local government, where local officials who provide better public services can prove their qualifications to compete for higher 19 office. In effect, democratic local governments make national politics more competitive by reducing entry barriers into the national political arena. But such competition is against the interests of the incumbent national leaders, who may therefore oppose any reform to introduce autonomous democratic local governments where they have not existed. National leaders would prefer to appoint mayors and governors, so that these local offices can be used as patronage rewards for national political service. Thus, when a transition to democracy begins with elections only at the national level, the winners may acquire a compelling interest in centralizing power around themselves. As the only faction to have been entrusted with power by a vote of the people, the new national leaders would have a strong mandate to guide any further process of constitutional reform, and they would be unlikely to promote reforms that devolve power to autonomous local leaders. So if democracy is introduced in the national government before it is introduced in local governments, political decentralization may become much more difficult to introduce thereafter. However, a constitutional system with autonomous local governments can become politically stable once it is established. When governors and mayors have been locally elected, they become local power-brokers from whom competitive candidates for national leadership must regularly seek support, and so it could be very costly for a national leader to threaten the constitutional powers of elected local officials. Dividing powers and responsibilities between independently elected officials of national and local governments inevitably creates a potential for conflict between them, as their policies and ambitions collide. Thus, one of the most important tasks in democratic state-building is to forge some generally accepted rules for dividing resources and responsibilities between local and national governments.6 Principles for defining a federal division of powers and for adjudicating disputes about this division have been developed and applied over many years in successful democracies, many of which devolve between a quarter and half of their public budget to subnational governments. Ghani and Lockhart (2008) have urged state-builders to recognize the fundamental importance of a well-functioning finance ministry for distributing funds reliably and accountably to different units of government. In many weak or fragile states, the trusted local leadership that exists is often associated with tribal or ethnic groups, so political frictions in new democratic states are often along ethnic 6 Abraham Lincoln's 1860 Cooper Union Address offers a powerful testimony on the importance of defining and maintaining a proper line between local and federal authority as a vital issue throughout American political history. 20 lines. Ethnic tensions in local politics can be reduced when the national government constrains local officials to provide equal protection of the law for all citizens, so that minorities' basic rights to vote or to hold property should not become contentious issues of local politics.7 But threats of ethnic conflict may persist until politicians begin to compete successfully with a strategy of earning the trust of people in all ethnic groups. When the goal of an international intervention is to support the development of a sovereign democratic state, the strategic direction of the intervention should depend on regular input from the political leaders who will be competing for power in the new state, both at the local and national levels. Realistic goals for a democratic state-building mission can be determined only in a process of negotiations with the contenders for local and national power in the state. Thus, a democratic state-building mission should take vital strategic direction from a decentralized team of local development coordinators who are sent by the intervention to engage constructively with national and local leaders in every part of the country. For effective support of local political development, broad power over all foreign assistance in each district could be delegated to one local development coordinator, who would oversee the allocation of aid to local groups. Then the coordinator could use this power over foreign assistance to promote the formation of an inclusive coalition for local governance, by directing aid to local leaders who cooperate with each other and with the national authorities. (See Myerson, 2019.) 8. A trade-off between goals of peace and development In the preceding section, our focus on the importance of democracy for a state-building intervention was derived from the necessity of distinguishing state-building assistance from imperialist domination. But we could also argue that a benevolent state-building mission should promote the possibility of democratic development in the future, even if not immediately, because democracy can offer advantages for promoting public and private investments in longterm economic development. Democratic political competition can motivate leaders to use public funds more effectively for greater public benefit. And citizens who have voting rights can have more confidence in the state's legal protection of their property, which is essential for them to make private investments that improve themselves and the local economy. But exclusive privileges of connection with an oligarchic or autocratic state can be 7 Catherine Boone (2009) has shown that customary land tenure rules have exacerbated ethnic tensions in many parts of Africa by opening political questions about which groups have a right to land. 21 instrumental for motivating a smaller elite group to make costly efforts and sacrifices to support the state in consolidating power and maintaining social order. Thus, state-builders should recognize a potential trade-off between the goals of inducing elites to support a peaceful social order and promoting a broader democratic foundation for long-term development. (See Cheng, Goodhand, and Meehan, 2018.) People who have been living in a failed state may value freedom from violence now more than the prospect of greater long-term economic growth, so that a rush to democracy could be against their current interests. A reasonable compromise might then involve trying to support developmental goals even without democracy, or trying to support a state which has at least some potential to become more democratic in the future. North, Wallis, and Weingast (2009) list three "doorstep" conditions that characterize a nondemocratic state that is ready for democratic development: (1) effective rule of law, at least for members of the governing elite; (2) the formation of more perpetual organizations, both public and private; and (3) consolidated political control of the military. Conditions (1) and (3) mean that the capability to apply violent force has become concentrated in a few specialized agencies of the state, so that most of the ruling elite (who may lack personal connections with the specialists in violence) must rely on an effective legal system to protect their rights and their property. Even if such a legal system was originally developed only for protecting the rights of elite officials, its protection may eventually be extended to all citizens. NWW's perpetual-organizations condition (2) includes the development of rules to support the creation of private corporations, so that entrepreneurs can organize new firms that make markets more competitive. But among the perpetual organizations that NWW discuss, none are more important than the institutions of national government and local government. In this regard, a multiplicity of perpetual organizations in government is realized by a federal political system where the national government shares power with autonomous institutions of subnational government. When the national government ensures that people and resources can move freely across the country, local economic activity in a district can be highly elastic with respect to the quality of its local public services; and this elasticity creates an incentive for district leaders to provide better public services, even without democratic accountability. (See Boix 2003.) Thus, the establishment of a stable constitutional division of powers between the national government and autonomous local governments can contribute to better conditions for economic development, even without democracy. For example, a federal division of powers without multiparty democracy can be found in China and in the United Arab Emirates. 22 9. Moral hazard in international interventions As we have seen, the failure of a state is a national crisis of moral hazard which can export violence and suffering to other nations, but international efforts to relieve the situation can introduce new elements of moral hazard. These moral-hazard problems begin at the highest level in the nations that intervene to rebuild the failed state, as the policies of the intervening nations must be directed by their national leaders, whose primary motivation must be based in their own domestic politics. In the absence of standing international forces, some nation must take the lead in supplying and managing an international state-building intervention (Fearon and Laitin 2004). Offers to provide costly support for the intervention will tend to come from nations that have particular interests in the target nation's political development. But any state-building intervention involves decisions about who should be supported as leaders of the target nation, and the interveners' particular interests will influence their choices of whom to support. So as long as the intervention continues, supported leaders may feel an imperative to align themselves with the interveners' interests. This imperative to not displease the interveners can distract the supported leaders from their primary domestic responsibility, which is to earn the trust and support of enough people in their own nation so that they can hold power without external support. In effect, external support from the intervention can become a substitute for internal political support and thus can undermine the healthy system of national politics that the intervention is supposed to be building. This is the basic state-builder's dilemma of Lake (2016): that the political base of the supported state can be distorted and weakened by the foreign interests which motivated the interveners to undertake the state-building mission. But in serious cases of state failure, there could be many nations that are willing to pay at least part of the cost of the intervention just to end the regional violence and enable refugees to go home, even without extracting policy concessions from the reconstructed state. Thus, if a global consensus can be reached about how a politically neutral state-building mission should be managed so as to minimize the interveners' influence on the new state's leadership, then international pressure might be able to compel interveners to act according to these standards. We hope that this paper and others in the literature may ultimately contribute to the understanding that is needed to define such standards for neutral state-building. (See Myerson 2014.) For such standards to be effective, however, if the leader of powerful nation ordered his 23 nation's forces to do state-building in a country which subsequently adopted policies that were unfavorable to his nation's interests, then this leader would have to be able to explain to his domestic voters that international norms prevented him from installing a more favorable government, and that regional peace was the worthy benefit that his nation achieved from its efforts. State-building missions create a complex relationship between the foreign interveners and the leaders of the supported state, and moral-hazard problems can arise on both sides of this relationship. Even an ideal benevolent intervener would find that the leaders who are being empowered by the state-building mission do not necessarily share the mission's goals of establishing a strong effective state that can serve its people without external support. First of all, the expectation that foreign assistance will be reduced when the state has a capacity to stand on its own can effectively reduce the incentive for officials of the supported state to actually achieve this capacity. This incentive to prolong the condition of dependency could be countered by setting a schedule for the gradual reduction of state-building assistance, even if progress toward self-sufficiency has been slow. An intervener might even raise the possibility of assistance being reduced sooner if there is a lack of appropriate progress. Second, although democracy, accountability, and a balanced federal distribution of powers may be vital for the development of an effective state that will serve its people well, the state's governing national elite may feel that their interests would be better served by less democratic competition, less public accountability, and less devolution of power to autonomous local governments. Furthermore, while the goal of state-building should be to establish sustainable institutions, the supreme leader of the state actually has an interest in the personalization of his role in the state, so that his grip on power may be strengthened by a perception that he is irreplaceable. To counter these tendencies, foreign state-builders should consider their indigenous partners as including more than just the top leadership of the state that they are supporting. As we have suggested, a democratic state-building mission should take its general strategic direction from a decentralized team of local development coordinators who are actively engaged with national and local leaders from all parts of the country and from all factions that are prepared to compete for power democratically.8 8 In this regard, a state-building mission may be fundamentally different from a normal diplomatic mission, where respect for national sovereignty is supposed to mean that the national leader is accepted as the ultimate authority in speaking for his nation. 24 We have suggested that these local development coordinators should exercise wide powers over foreign assistance in their district, so that they can effectively promote the formation of a broad inclusive coalition for local governance. But such intervention agents also need supervision. These local development coordinators may be expected to work full-time in their district for a term of one or two years, but then they should be rotated to other districts, so that a local coordinator cannot establish independent personal authority anywhere. For continuity, each local development coordinator could be administratively accountable to a supervisor who would have long-term responsibility for a province which may contain the districts of a few local coordinators. Such provincial supervisors could also play a key role in formulating the longterm strategic direction of the whole state-building mission. (See Myerson 2019.) The establishment of a new state can take many years if it requires people throughout the nation to modify their cultural expectations about how their leaders should be identified and what kinds of behavior would disqualify someone for a position of leadership. A new constitutional system becomes resilient when political leaders have reputations for respecting its norms, but the development of such reputations may take time. And opponents of the new state may be willing to fight for a many years if they think that they could ultimately establish themselves as leaders in a different form of political organization for this nation. So a state-building intervention may require a credible commitment for many years of support from a foreign intervener to fully establish the authority of the new state. But such a commitment would not be credible if it would cost more than people in the intervening nation are willing to pay. Thus, it may be better for an intervening nation to limit its state-building support to a level that it will be willing to sustain over many years. We have seen that interveners may also need to make their assistance conditional on the state adopting some policies or reforms which are essential for the goals of the state-building mission but which might be against the interests of the supported state's leaders. For such conditionality to be credible, the intervening nation must be willing to reduce or withdraw their support if leaders of the supported state refuse to do what is required. At the very least, the interveners should not be so committed to supporting the state's current leadership that there is no realistic alternative. In this regard, parliamentary forms of government in which top leaders can be replaced at any time may have some advantage for state-building missions. To support the establishment of a constitutional state in which power is distributed across different levels or branches of government, the interveners must encourage officials at all levels 25 to accept constitutional limits of their authority and to work cooperatively with officials from other parts of the state. (When the state is well established, a state official should normally get such encouragement from the supervisor to whom the official is administratively accountable or from the selectorate to whom the official is politically accountable.) Agents of the intervention can provide this encouragement by promising officials at every level that they can expect benefits of foreign assistance if they work to fulfill their responsibilities in the new constitutional system, and such promises can be credible when these officials have long-term working relationships with authorized agents of the intervention. Each responsible official needs to know that his cooperation and good service to the state today will be recognized by someone who knows him and will be able to help him in the future, and while the state is being formed such encouragement can initially be provided by agents of the intervention. Thus, the success of a state-building mission may depend on personal relationships of trust between officials of the supported state and agents of the intervention, whose job is to encourage these officials to fulfill their roles in the new constitutional system. At this point we should consider the question of whether a state-building intervention could accomplish its goals with just economic assistance, without bringing any military force of its own. One might hope that an unarmed intervention could avoid any responsibility for violent conflict and instead could concentrate just on helping people in the country; but it is not so easy for state-building assistance to be disassociated from violence. Even if agents of the intervention came only with economic assistance, their goal of transforming the nation's political system would make them a serious threat to groups whose capacity for violence gives them local power in the absence of an effective state; and so state-building agents cannot do their job without some armed protection against the violence of such groups. Shipments of food and other valuable assistance also require protection against banditry. When an intervention does not include any military forces of its own, then local armed groups must be paid to protect the intervention's agents and supplies. These protection payments can significantly increase the wealth and power of the armed groups that receive them and thus can actually become one of the principal ways that the intervention affects the local political situation. By subsidizing armed militias in this way, even humanitarian missions can do more harm than good. In Somalia, for instance, foreign assistance organizations regularly paid warlords for protection and thereby increased the resources and power of these violent actors relative to others in the local political arena. (See Lake 2016, p172). 26 The issues of state-building have generally been considered here as if a nation would have at most one foreign intervention to support the establishment of a new state. When international relations are guided by a norm that requires state-building interventions to get broad multi-national approval, then indeed we may expect no nation would ever be the target of more than one internationally approved state-building mission. But such a norm has rarely prevailed in world history. In fact, competition between rival interveners has been a primary cause of state failure in many countries. As we have noted, the main resistance against a state-building mission has often come from groups that also have foreign support, from other nations which favor the establishment of a different kind of state. In cases of such rival interventions, much of the discussion above remains relevant to the question what each side could do to most effectively promote the establishment of its favored political system. But any hope for state-building interventions to be a mechanism for improving the lives of people in weak or failed states will depend critically on some international consensus about when an intervention is justified. Ultimately we must recognize that the state-builders' goal of helping a nation to establish a sovereign independent state means that, after the intervention ends, the target nation's politics will continue to develop in ways that foreign agents cannot control. As long as its national history continues, new groups will compete for power in the country, its diplomats will negotiate new kinds of international relationships, and its form of government will continue to evolve through conflict and reform. So nothing that a foreign state-building mission can achieve should be considered final for the sovereign independent state that it is supporting. Interveners must understand that, in their efforts to reconstruct the political system of the target nation, neither their best hopes nor their worst fears can ever be permanently accomplished. During the period of the intervention, foreign state-building agents may strive to reduce violence and increase the effectiveness and accountability of government in the target nation. But the duration of the statebuilding intervention must be strictly limited, and thereafter it must yield all its achievements and failures to the process of endless change that is human history. 10. Making the world safe for state-building After surveying complex and interconnected problems of political development and international relations, we should return to our basic question: What kinds of international responses can be recommended to help solve the problems of weak and failed states, without admitting new forms of imperialism into global politics? 27 Even without formal interventions, norms of the international community can affect political development in weak states. The basic principle of international respect for the sovereignty of a recognized national government enables national leaders to extract valuable rents from regulatory control over foreign trade and foreign assistance. Such rents from foreign assistance can actually affect the distribution of power in a poor country, strengthening the national leader against potential rivals. International assistance might be more neutral with respect to the crucial distribution of power between national and local authorities if foreign assistance donors could give some appropriate fraction of their aid through autonomous local governments without national supervision. In terms of its effect on the distribution of power in the target country, the least intrusive form of international intervention that we discussed was that of a neutral foreign arbitrator or podesta, who would help to settle disputes among the various oligarchic leaders whose factions dominate the country in the absence of a strong state. A neutral international agency could perhaps facilitate the use of such podesta peace-keepers by certifying the qualifications and neutrality of individuals (such as retired diplomats or former heads of state) whom the agency could recommend to countries where such peace-keeping arbitration is needed. But because neutral peace-keepers are not supposed to change the distribution of power in country, they could do little or nothing for building an effective state where one does not exist. We have argued that, for an international intervention that aims to build a state, some form of democratic political accountability should be required for at least some officials in the new state, so that the intervention can distinguish itself from conquest or thinly disguised imperialism. But we know that democracy is hard to introduce in a failed state that has never been democratic. With no prior history of democracy, it may be difficult or impossible to prevent prominent national leaders from corrupting a high-stakes national election with violence and fraud. So even if an intervention sponsors formal elections to choose national leaders, there may be serious doubt about whether the electoral outcomes actually depended on voters' choices. One way to address international doubts about democratic state-building is to have broad multi-national supervision of the process of establishing a provisional government and holding the first elections. Then, even if elections are flawed, supervisory agents from different nations can act as independent observers to provide the world with some credible evidence about the extent to which elections were or were not manipulated by the foreign interveners. Another important way to address the problems of a new democracy is to have local 28 elections for responsible local councils along with or before the first national elections. When different parties win power in different areas of the country, the world can see that the interveners have not simply given all power to their favored party. Local elections can also help to ensure broader acceptance of the new state, as trusted local leaders gain a stake of power in local governments. And there are likely to be at least some districts or municipalities where local democracy actually functions appropriately, and then the country will have some local leaders who have actually gained power by competing fairly in elections. This vital supply of leaders who have good reputations for respecting democratic norms and exercising public power responsibly can eventually expand from local politics into national politics, as successful local leaders become credible democratic contenders for higher office. The elements that we have discussed here could be combined in a practical template for effective international responses to the problems of failed states. The United Nations Security Council is the primary international institution for authorizing and regulating an intervention that could violate the sovereignty of a country with a failed state, so as to bring order and lay the foundations for a new state there. Even when the Security Council designates one nation to lead the intervention militarily, the Security Council could direct the Secretary General to ask other nations to join a supervisory coalition that would support the intervention and ensure that the process of political reconstruction is as inclusive as possible. Then leaders of the major factions which have been contending for power in the country could be invited to join in forming provisional national government, which would be formed as an oligarchic coalition, distributing incentives for cooperation to as many factions as possible. As in past peacekeeping operations, the Secretary General could send a Special Representative to serve as a neutral arbitrator for negotiating a distribution of state offices and public funds among the cooperating factions. But the Special Representative should also bring an international team of local development coordinators who would have responsibility for overseeing the formation of a local council in each district or municipality. Ideally, the local councils should be democratically elected; but in any case, the development coordinator in each district would be expected to promote the formation of a local council in which cooperating factions and significant local groups can get representation that is appropriately commensurate with their local popular support. In this plan, the Special Representative should have general authority to direct the distribution of all foreign financial assistance for the new provisional government. The larger share (perhaps 2/3 of all assistance funds) would go to the national government. But a 29 substantial fraction of the assistance funds (perhaps 1/3) should be distributed to the local councils, where the budgeted use of all foreign assistance funds would require approval by the local coordinator. The Special Representative might also name a commission of constitutional experts to help delimit the powers and responsibilities of each branch of government, until the country can ratify a new national constitution. During this period, the intervention team's ability to withhold assistance funds could help in motivating officials to accept these limits. Any state-building intervention needs a clear exit strategy. Within a clearly limited number of years, the interveners' goal of supporting political development must give way to the normal principle of international respect for national political independence. Then, during a period of transition, the amount of foreign assistance that is directed by the Special Representative and the team of local development coordinators should be reduced gradually to zero, and international donors should be encouraged to fill in with independent offers of assistance wherever needs are identified by the new state's national and local authorities. Such an intervention would require at least two resources that are currently in short supply. One is an international cadre of foreign-service officers with the training in local government administration that would qualify them for professional service as local development coordinators. The nations of the world maintain a more-than-adequate supply of trained soldiers, but future state-building missions could have better outcomes if some nations or international organizations also maintained a ready corps of trained coordinators for local political reconstruction. The other scarce resource is a broad international consensus about when such violations of national sovereignty should be sanctioned. All our hopes for constructive interventions that actually help to solve the problems of weak and fragile states depend on international agreements to support cooperative solutions, and such agreements become more difficult to negotiate when international tensions are high. International rivalry among powerful nations can induce them to sponsor rival interventions in a weak state, as they each try to install a government under the control of their own clients. Such rival interventions can tear a country apart, actually causing the failure of its state. Thus, diplomacy that reduces great-power rivalries and increases the recognition of shared interests among the most powerful nations of the world also has the potential to benefit people in the weakest and most fragile states. 30 References: Armen A. Alchian and Harold Demsetz, "Production, information costs, and economic organization," American Economic Review 62:777-795 (1972). Gary Becker and George Stigler, "Law enforcement, malfeasance, and compensation of enforcers," Journal of Legal Studies 3:1-18 (1974). Carles Boix, Democracy and Redistribution (Cambridge U Press, 2003). Catherine Boone, "Electoral populism where property rights are weak," Comparative Politics 41(2):183-201 (2009). Bruce Bueno de Mesquita, Alastair Smith, Randolph Siverson, and James Morrow, The Logic of Political Survival (MIT Press, 2003). Christine Cheng, Jonathan Goodhand, Patrick Meehan, “Synthesis Paper: Securing and Sustaining Elite Bargains that Reduce Violent Conflict,” UK Stabilisation Unit, 2018. https://eprints.soas.ac.uk/30510/1/Elite%20Bargains%20and%20Political%20Deals%20P roject%20-%20Synthesis%20Paper.pdf Merle Fainsod, Smolensk under Soviet Rule (Harvard U Press, 1958). James D. Fearon and David D. Laitin, “Neotrusteeship and the Problem of Weak States,” International Security, Vol. 28, No. 4 (Spring, 2004), pp. 5-43. Ashraf Ghani, Clare Lockhart, Fixing Failed States: A Framework for Rebuilding a Fractured World (Oxford University Press, 2009). Avner Greif, "Self-enforcing political systems and economic growth: late medieval Genoa," in Analytic Narratives, editors Robert H. Bates, Avner Greif, Margaret Levi, Jean-Laurent Rosenthal, and Barry R. Weingast (Princeton U Press, 1998), pp. 25-64. Thomas Hobbes, Leviathan (London, 1651). Jeffrey Herbst, States and Power in Africa: Comparative Lessons in Authority and Control Second Edition (Princeton University Press, 2014). David A. Lake, The Statebuilder's Dilemma: On the Limits of Foreign Intervention (Cornell U Press, 2016). LSE-Oxford Commission on State Fragility, Growth, and Development, "Escaping the fragility trap," International Growth Centre, LSE and Oxford (April 2018). https://www.theigc.org/wp-content/uploads/2018/04/Escaping-the-fragility-trap.pdf Jahara Matisek and William Reno, "Getting American security force assistance right: political context matters," Joint Forces Quarterly 92:65-73 (Jan 2019). 31 Roger Myerson, "The autocrat's credibility problem and foundations of the constitutional state," American Political Science Review 102:125-139 (2008). Roger Myerson, "Standards for state-building interventions," in Economics for the Curious, edited by Robert Solow and Janice Murray, pp 174-188 (2014). https://home.uchicago.edu/~rmyerson/research/std4sb.pdf Roger Myerson, "State-building lessons from the British Empire" (2019). http://home.uchicago.edu/~rmyerson/research/sblessons.pdf Douglas North, John Wallis, and Barry Weingast, Violence and Social Orders (Cambridge, 2009). Andreas Osiander, “Sovereignty, International Relations,and the Westphalian Myth,” International Organization, July 2003. Thomas C. Schelling, The Strategy of Conflict (Harvard University Press, 1960). James C. Scott, Against the Grain: A Deep History of the Earliest States (Yale U Press, 2017). Max Weber, "Politik als Beruf" (Munich: Duncker & Humblot, 1919). Xenophon, The Education of Cyrus, translated by Wayne Ambler (Cornell U Press, 2001). 32