The international community, led by the United States, has invested trillions of dollars in state-building efforts during the past two decades. Yet despite this commitment of substantial resources, conflict and violence remain a challenge in fragile states. It therefore seems especially important to consider the reasons why state-building has not lived up to its expectations.
Past state-building efforts were predicated on the belief that a centralized government would improve prospects for political order and economic development. These efforts therefore have typically emphasized powerful national governments and centralized bureaucratic administration as the keys to generating improvements in the state’s provision of public goods, including rule of law and collective security.
This article challenges the underlying assumptions to that approach, arguing that centralization actually undermines efforts to stabilize and rebuild fragile states. It describes several risks centralization poses for effective state-building. For example, many highly centralized governments prey on their own citizens and are therefore prone to civil unrest, conflict, and collapse.1 Most of the countries that have experienced prolonged civil conflict over the past several decades—including Afghanistan, Libya, Myanmar, Somalia, Syria, and Yemen—had extremely centralized governments prior to the outbreak of conflict. As Roger Myerson points out, centralization may also alienate local elites, resulting in a return to conflict and violence in states subjected to foreign state-building.2 Another risk is that centralization undermines the quality of public administration by making it largely unresponsive to local demands.
The obvious alternative to trying to build powerful central governments is to encourage polycentrism. Polycentrism allows multiple, coexisting centers of power, which provides for local autonomy in collective decisionmaking.3 Political decentralization promises several benefits. A more robust system of local governance can reduce the incentives to rebel against the central government and provides resources for those regions to defend themselves from outsiders.4 Polycentrism can also provide a foundation for local self-governance and reduce the state’s administrative burden by allowing communities to govern their own affairs. By decentralizing political decisionmaking, the state can also improve the responsiveness of the government to local demands, and in the process improve the efficiency of public goods provision.
I use evidence from Afghanistan to illustrate my theory of the pathologies of current models of centralized state-building. Despite vast investments, the government in Kabul has limited legitimacy and struggles to provide public goods. One plausible explanation for the failure of state-building in Afghanistan is that the government remains extremely centralized in all critical dimensions, including the power of the executive, subnational governance, judicial institutions, public budgeting and finance, and the national security forces. Of these, only the Afghan National Army has implemented meaningful reforms. In the other areas, almost no reform has occurred compared to the institutional status quo before 2001. A consequence is that most Afghans continue to experience the same type of centralized, predatory state that they endured prior to 2001. Paradoxically, by resurrecting the centralized, predatory state, the stabilization effort continues to give rise to an antigovernment insurgency across the country.
The Case for Decentralization
It is not a mystery that policymakers, both local and international, would seek to build strong central governments in states riven by conflict. Historically, states emerged to facilitate commerce and provide collective defense at scale, as societies grew and became increasingly complex and older, more traditional political forms, such as tribes, clans, or fiefdoms, were insufficient for the evolving political challenges. Political leaders seeking monopolistic power over their territory are incentivized to provide public goods and encourage production to generate more revenue via taxation.5 As institutions of wealth creation are established, the government must invest in collective defense to defend commercial interests from external predators.6 Centralized state capacity thus contributes to economic development and political order.
A government powerful enough to provide these public goods, however, can also use its power for destructive purposes or to misappropriate public goods.7 To prevent that unwelcome outcome, state authorities must be constrained by counterbalancing political institutions that empower citizens to participate in collective decisionmaking.8 This is especially important in conflict-affected states where insurgencies began as revolts against predatory, unconstrained central authorities.
Decentralization is a form of polycentrism and can be found in both de jure and de facto manifestations. De jure decentralization provides for constitutional rights of local units in a political system to determine their collective futures. Such decentralization is the basis for formal, law-based self-governance. De facto decentralization occurs when the central government institutions are unable to impose their will on local jurisdictions, thus yielding authority by default.
One justification for decentralization is that it provides a political architecture for checks and balances in government. Local authorities can constrain central governments in ways that prevent a drift toward predatory practices. Decentralization also provides institutional mechanisms for the government to acquire knowledge of local conditions, needs, and demands. The government requires this information to provide public goods efficiently. The grant of local autonomy may even in itself provide legitimacy to the central government.
Local autonomy and the experience of self-governance are valuable in conflict-affected states where communities must often rely on local problem-solving skills when the central government is unable or unwilling to address local needs. Most places plagued by long-term conflict are home to a plethora of powerful “warlords” or local commanders. Stabilization efforts typically seek to bring these individuals under the umbrella of a centralized state, often neglecting the role such individuals and groups play in crafting order based on local preferences.
Decentralized governance is not a panacea. Politicians and leaders do not become angels simply because their power extends to the local context. There are indeed risks associated with decentralization, such as local corruption. Nonetheless, the risk from local rulers behaving badly pales in comparison to grand corruption at the national level, where leaders often have few ties to local communities. In stabilization contexts, this situation is exacerbated as many view those appointed by the international community as puppets rather than legitimate leaders. Another fear is that decentralized systems may fail to maintain cohesion. Such systems may be more prone to separatist movements.
The Political Problem of State-Building
Despite the benefits of political decentralization in state-building, international and domestic actors typically prefer the centralization model. Consider the international actors who play an important role in state-building efforts. First, in unstable environments, outside actors—both military and civilian—prefer to interact with entities that exercise unity of command. They seek one interlocutor upon whom they can count to implement policies. The more actors involved, the more diffuse the decisionmaking, which tends to slow down the policymaking process.9 Second, it is easier for external actors to exert influence over a unified single entity rather than a group of individuals; external actors prefer to have control centralized in the capital in order to exercise oversight and have a single accessible partner to hold accountable. Centralized government reflects the desires of the international community in postconflict situations because they have a primary relationship with the national figures they believe they can control.
Newly elected domestic leaders also prefer highly centralized power, especially when this authority comes with international support. In most conflict environments, power tends to diffuse across a range of actors at subnational levels who may have recently challenged a newly installed leader. During the immediate postconflict period, new leaders desire centralized authority, enabling them to exert greater control over regions and increasing their national authority. Unlike states not affected by conflict, the decentralization issue presents very real threats to the endurance of a polity, as subnational regions may be controlled by or under the heavy influence of armed groups that may pose or have recently posed a serious challenge to the central government.
These demands for political and administrative centralization often conflict with the reality of political power in such contexts. During conflict, power necessarily diffuses as central governments lose their monopoly of violence. It is the primary goal of stabilization and state-building efforts to help nascent states assemble or reassemble disparate pieces of power into a cohesive unit. But during this process, policymakers often eschew informal rules, norms, and other practices that emerge locally during periods of conflict.
New governments seeking to assert authority typically view informal practices as impediments to the uniformity of authority they seek to establish. These rules may be the result of longstanding custom or newly developed practices that emerged as coping mechanisms during conflict. They may be rules imposed by warlords or local commanders who were parties to the conflict. Regardless of their origin, those international and domestic forces seeking to build and stabilize states see them as a source of fragmentation rather than contributing to the coherence and legitimacy of a new government.
The desire to centralize authority and impose uniform rule over fragmented and weakly institutionalized environments widens the gap between de facto practices and de jure laws and regulations. This growing gap challenges both development and stability. Typically, stabilization processes do not seek to modify their rules and procedures to account for local practices but instead attempt to encourage citizens to turn to the state. By imposing highly centralized forms of governance in fragmented environments, stabilization efforts inadvertently undermine their own mission by exacerbating development challenges in such contexts.
In addition to centralization, the international community frequently prioritizes national elections, rather than the quality of governance, as the measure of success. In the “golden hour” after the cloud of war clears, state-builders rush to congratulate themselves for organizing successful elections that they believe consolidate political transformation. The distribution of power—especially to subnational units—and the way that power is executed through the bureaucracy receive comparatively little attention. However, national elections are often a poor measure of the quality of government. They have in many instances little bearing on the provision of public goods or implementation of beneficial policies. Indeed, in some instances, non-democracies may be better at governance, in terms of delivery of services, predictability, and order, than democracies.10
The Afghan Context
The desire for greater political centralization is perhaps the fundamental political theme in Afghan history. From its inception with the Durrani empire in 1747, the Afghan state was more like a loose confederation of tribes than a centralized state, with tribal and local leaders as well as blood relatives of the monarch exerting substantial autonomy. Political power was exercised through indirect rule, in which monarchs and other political leaders exerted authority through a network of autonomous princes while maintaining a role for tribal and customary authority.
Abdur Rahman sought to centralize government during his brutal reign from 1880 to 1901. He took it upon himself to stamp out the “middlemen”—tribal and customary leaders, as well as religious authorities—whom he viewed as a source of political disorder. Subsequent regimes continued these efforts at centralization, although the tactics were less brutal. Amanullah (1919–29) attempted to increase the role of the state in Afghan society and economy through a series of centralized directives. He admired the strong leadership of Ataturk in Turkey, who used executive authority to promote an expansive agenda of social transformation. However, most of Amanullah’s reforms were met with opposition, and he was ultimately ousted in a peasant-led rebellion that briefly unseated the Afghan monarchy.
After several years of instability, Afghanistan experienced a long peace under the rule of the Musahiban, a Durrani Pashtun sub-tribe from which its leaders descended. These leaders included Zahir Shah, who ruled from 1933 until his overthrow in a bloodless palace coup by his cousin, Daud, in 1973. Daud ruled from 1973 until 1978, when he was ousted by a Soviet-supported faction of the Afghan communist party.
In general, the Musahiban pursued an economic development strategy that relied heavily on foreign subsidies from both the United States and the Soviet Union, leading to continuation of the rentier state dynamics begun in the 19th century when monarchs began accepting British aid in turn for quiescence. They also relied on centralized economic planning. Despite a constitutional reform in 1964 that created a brief flirtation with democracy under a constitutional monarchy, the monarchy remained committed to principles of centralism, with most important governance decisions emanating from Kabul.
The growing influence of the Soviet Union in Afghanistan after World War II reinforced the historical struggle toward increasing centralization. Beginning with Daud’s reign as Prime Minister in the 1950s (he would later become president in 1973), the Afghan government adopted many Soviet features. Daud famously fell out with the Americans and grew to mistrust them after his unconditional requests for weapons systems to facilitate the creation of a modern army were rejected by Washington, which he perceived to support newly independent Pakistan. Labelled the “Red Prince,” Daud also imported bureaucratic centralization and five-year plans along with Soviet loans. The organization of the central government, the system of public financial management, and the bureaucracy were very similar in design to their Soviet counterparts. Such influence continued to grow over time and escalated in 1978 with the Saur revolution that brought Afghan communists to power for the first time and marked the end of the Afghan monarchy.
The Saur revolution of April 1978 brought a divisive period in which the Khalq and Parcham factions of the People’s Democratic Party of Afghanistan (PDPA) vied for control. In late 1979, the Soviets invaded partly in response to instability brought about by local opposition to Khalq policies. Soviet leaders were wary that the Afghan government had moved too quickly in its efforts to achieve “communism” and that its actions would continue to provoke unrest in the countryside (which they did, as the mujahideen emerged in direct response to communist policies). They were also increasingly wary of the infighting among the PDPA factions that fueled instability. Once the Soviets invaded, the centralized Soviet imprint on Afghan bureaucracy and government became heavier. Although the Soviets did move to reverse some of the more drastic reforms initially imposed by the Khalqis, they continued to support centralized state control.
The PDPA government fell in 1992. Although there have been few studies of governance during the civil war period (1992–96), formal institutions had very little influence. Once the Taliban government (1996–2001) came to power, it began to reform; however, the logic of Taliban governance was also one of centralization. Instead of relying on old networks of bureaucrats, the Taliban relied on a network of its own local mullahs to administer authority. But the character of public administration was characterized by centralized governance. While the Taliban were never fully able to provide effective central administration, they did little to encourage local autonomy.
After the fall of the Taliban regime in 2001, U.S. and North Atlantic Treaty Organization partners, along with several factions of the Afghan resistance and the former monarchy, gathered to sign the Bonn Accords, which established the 1964 constitution as the interim source of law for the country. The agreement appointed an interim president, Hamid Karzai, and planned for elections. Thus, the first move of the “new” Afghan state was to adopt a decades-old constitution. Yet over the course of a decade, the 1964 constitution had failed to establish a stable government, let alone a government that could implement reforms. In fact, many Afghans met the overthrow of the 1964 constitution by Daud with relief, believing it to be inherently unstable and preferring authoritarian rule.11 Nonetheless, the constitution of 2004 would copy many elements directly from the 1964 constitution. Indeed, in most major areas of Afghan public administration—the national government, the degree of centralization of the government, judicial institutions, public finance, the security forces, and administration of land—state-building efforts have reinforced old, inefficient institutions. Interestingly, one exception is the Afghan National Army.
The 2004 Afghan constitution established a presidential system with a bicameral parliament. Although it provides for a separation of powers between the president, the legislative branch, and the judiciary, the executive branch has remained by far the strongest. Not only does it have the authority to appoint members of the judiciary, it is also responsible for appointing one-third of the upper house of the National Assembly. The strength of the executive mimics the overwhelming authority of the 1964 constitution. Elections notwithstanding, the executive has nearly the same powers that Afghan kings exercised.
During the hotly contested 2014 presidential election between Abdullah and Ashraf Ghani, tainted by allegations of corruption and fraud, a growing number of doubters began to criticize the strong presidential system. To resolve the disputed electoral result, the United States brokered a deal between the two candidates that created a new chief executive officer (CEO) who would sit alongside the president. Cabinet positions would be jointly appointed by the two leaders. While the creation of the new CEO position facilitated a power-sharing agreement, it is unclear the degree to which such an extraconstitutional system will endure. The creation of a new powerful position atop a strongly centralized presidential system is inherently unstable.
Ironically, those involved in drafting the 2004 constitution considered a parliamentary system featuring both a president and a prime minister, but the drafting commission ultimately shifted to a strong presidential system, creating a powerful executive with complete authority to appoint the cabinet, subject to the approval of the National Assembly. The rationale for this shift was that under conditions of state weakness, creating a parliamentary system characterized by factions would yield a “fragmented body dominated by warlords, local factions, and even drug traffickers.”12 Both Afghan and foreign policymakers involved in the state-building process feared that a weak executive would be incapable of generating state capacity, as infighting would dominate politics. A strong executive, however, could help rapidly steward preferred policies so a presidential system without substantial checks and balances was ratified.
The executive authority of the president was not based on a model of separated powers. Rather, its origins lie in a pre-democratic constitution with powers that were designed to be wielded by a monarch. It should therefore come as no surprise that since 2001, the National Assembly has not played a sustained role challenging executive authority on public policy issues.
The 2004 constitution created a centralized political system and public administration that were identical in form and function to the monarchy and communist governments. All subnational government officials—including provincial and district governors—are to be appointed by the central government in Kabul without constituent input. The design of subnational government bears a very heavy Soviet imprint. The appointment system and organization of the bureaucracy at the subnational level in Afghanistan are virtually identical to the Soviet model upon which it was based.
The notion that there is even formal local government in Afghanistan is deceptive, as subnational officials, including both provincial and district governors, are appointed by and beholden to Kabul. Although the appointment process is technically controlled by the Independent Directorate for Local Government in Kabul (an executive agency under the president), the appointment process is highly political and has emerged as a system of patronage for the central government. Individuals have no say in who runs their local government. There is no representation at the local level. In a 2016 survey, just 43 percent of Afghans said that they have influence over local government officials.13 This heavily centralized de jure system of government stands in stark contrast to—and disregards—the robust system of de facto informal governance that exists in communities around the country.
Not only are provincial and district governors appointed from Kabul, the authority to assign local government bureaucrats—representatives of the various national ministries—is also reserved for the line ministries in Kabul. This does not mean, however, that powerful local figures cannot emerge as both provincial and district governors. Some effective provincial governors have emerged, many of whom are or were powerful warlords affiliated with mujahedin parties. Their effectiveness is due to their independence from Kabul with respect to resources or legitimacy. Instead, they govern by almost completely disregarding the formal laws and rules meant to constrain their local authority.
The 2004 constitution called for the creation of elected provincial, district, and village councils. By early 2019, there had been three rounds of provincial council elections (2005, 2009, 2014), but these elections have not yielded local councils that play an important role in checking the authority of provincial governors. Provincial councils remain extremely weak as governing and oversight bodies and have no formal authority to draft local legislation. They are supposed to oversee the work of provincial governors, but their oversight responsibilities are not clearly defined. As governors are appointed by Kabul, they have no incentive to respond to the wishes of provincial councils. Elections for district and village councils have not yet been held.
The centralized system of local governance has resulted in a subnational administration that is incapable of reflecting constituent needs because it cannot aggregate the preferences of citizens. Most provincial and district governors are not from the areas they serve. Instead, they are rotated from one district or province to another. As a result, customary systems of governance that resolve disputes and provide small-scale public goods remain quite important in rural areas, and effective district officials employ customary rule rather than state law.14 As local government officials are not formally accountable to citizens at the local level, this system has bred vast corruption, causing further disillusionment of citizens with their government.
Delivering the rule of law is one of the most basic functions of a state yet is one of the most difficult to achieve. In Afghanistan, establishment of a coherent justice system in many ways remains as remote as it was in 2001, largely the result of the perceived illegitimacy of the formal, state-backed justice system. The judicial system remains highly centralized, with authority to select and oversee judges and the court system firmly entrenched in Kabul. According to a nationally representative survey, Afghans reported paying more bribes to the judiciary than to any other public institution, and informal dispute resolution mechanisms are far more popular and trusted than government courts.15
During the second half of the 19th century, Afghan monarchs relied upon the court system to extend the reach of the state into communities throughout the country.16 Thus, the courts played an important role in extending the writ of the state, under highly centralized control. Consistent with this historical legacy, the Afghan judiciary remains under the control of the executive branch. This was true throughout most of the 20th century, including the PDPA government that sought to use courts as a means to impose social and political agendas.17
The current court system in Afghanistan is almost identical to the legacy system. The supreme court oversees most local courts. Judges in provincial and district courts are appointed by the supreme court in Kabul. Although the National Assembly does play a role in affirming appointments to the supreme court, justices have enormous control not just in terms of interpreting the law but also over administration of the entire justice system, which—given the supreme court’s subordination to the executive—becomes an extension of executive political control.
Beginning in the 1950s and accelerating during PDPA rule, government institutions, particularly the judiciary, reflected Soviet models. The PDPA abolished the supreme court and replaced it with a special revolutionary court. Soviet advisors helped set up a new internal security agency modeled on the KGB, known as the KhAD.18 In practice, the KhAD “exercised full judicial power through summary arrests, detentions, and executions.”19 The supreme court was eventually reinstated toward the end of PDPA rule, but the legacy of the courts as tools of politics is one that has not been quickly forgotten by most Afghans, who continue to shun the formal courts.
Not only is the court system highly centralized in Kabul, but so too is the public prosecutor’s system and the bodies responsible for drafting laws. In 1981, the PDPA created an attorney general (Loya Saranwal) office that serves as a general public prosecutorial body, with prosecutors appointed by authorities in Kabul.20 The attorney general’s office operates under a similar logic today as it did under the PDPA period. Similarly, the ministry of justice, rather than legislators, drafts most laws that go before the National Assembly.
Some heralded the 2004 constitution as a break with the past because judicial bodies were to have some degree of oversight of constitutional and legal interpretation, but this authority has been hotly contested. The National Assembly has turned to the Independent Commission on the Supervision of Implementation of the Constitution, while President Karzai maintained that the supreme court should fulfill this role (the 2004 constitution established both bodies but did not clearly specify their duties).21 Although observers have lamented the absence of judicial review, the assertion of the National Assembly vis-à-vis the executive-controlled supreme court represents a step by the legislative branch to assert some authority.
Public Budgeting and Finance
There is perhaps no area where Soviet influence has been more felt in Afghanistan than in the system of public finance—in raising revenue, budgeting, and budget execution.22 Although donors have tried to affect some reform of the system to increase efficiency, it is unlikely that any amount of reform or capacity-building within the current system could yield policies that reflect citizen interests.
The system of public finance remains one of the most centralized in the world. Local governments do not have the right to tax and spend funds that they raise. Most local revenue must be sent back to the central government in Kabul, where it is then redistributed to local subunits based on the national budget plan. Although state-builders tried to make some changes to the budgeting system and incorporate some local input into the process, these reforms have been little more than window dressing as more than 50 line ministries and executive agencies make their budgets in Kabul for all provinces and districts of the country. In 2016, just 34 percent of citizens reported having confidence that government ministries are meant to deliver services to communities.23
The public finance system together with lack of local self-governance means there are few opportunities for citizen preferences to be translated into the budget. Budgets reflect the priorities of the central government in Kabul. Because local officials are not involved in the budget plans, they rarely have the staff or personnel required to execute the budget once it has been drafted. Execution of the development budget (which does not include salaries) has hovered around 40 percent during the past decade, a signal that the government cannot spend the meager funds it collects. Although some small modifications have been made to the system of public financial management (such as the introduction of program budgeting and some moves toward more localized decisionmaking over implementation), such changes have been cosmetic.
In several areas, such as education, rural development, and public health, the government has made some claims to success in service delivery. This is because these programs have been funded and managed by donor partners who then contract out implementation to international and domestic nongovernmental organizations and contractors. In this parallel system, donor funds go to a specific ministry (such as health or rural development) to fund a “national priority program.” Donor funds are transferred to the specific ministry but then are quickly contracted out to third parties with donor oversight. While this system may have been more effective than relying upon the government of Afghanistan, it has further undermined the state. As donor assistance decreased, the increases in “capacity” generated by such a parallel system quickly evaporated. In this aspect, rentier parallel structures emerge because the formal system of public financial management and policy execution is so weak that donors cannot rely on it to deliver the results citizens expect.
Donor resources have encouraged centralization rather than reform.24 As a result, the government remains unable to deliver many basic public goods and services to its citizens. While some donors have encouraged the use of contractors and other third parties to deliver services on behalf of the government, citizens rarely recognize these as government programs because they are delivered by foreign organizations. The inability of the state to generate and execute a budget has cost the government its legitimacy.
National Security Forces
The Afghan National Army (ANA) and the Afghan National Police (ANP) embarked on two very different paths of institutional reform after the fall of the Taliban regime. Previously, both the ANA and the ANP evolved as highly centralized security forces that by the beginning of the civil war reflected heavy Soviet influence. Daud began centralizing the army during his first reign as prime minister in the 1950s. He asserted stronger centralized control over its organization after he reclaimed power in 1973. At first, he partnered with factions of the Afghan Communist Party. For instance, Hafizullah Amin, who became leader of Afghanistan under communist rule, began recruiting Khalq members into the army beginning in the 1970s. During the 1980s, the Soviets brought in advisors and imposed the Soviet political-bureaucratic model of management to strengthen the army, which even began employing Soviet agitprop models to win the hearts and minds of local populations. By 1990, 70 to 80 percent of army officers were PDPA members.25 Although many officers relied on patrimonialism, the army was designed to maximize political control by the PDPA. It remained so until the outbreak of the civil war in 1992.
The police had always been a much weaker organization than the army as its development had received far less attention from Kabul during the monarchy. This changed after the 1978 coup, when the PDPA began to impose Soviet organizational models on the ministry of interior (MoI), which was responsible for managing the police. Although Germany had played a role training the police prior to the Saur revolution, the police grew as an organization under Soviet tutelage.
In the first few years after 2001, donors were committed to a “light footprint” approach in Afghanistan and did not develop long-term strategies to develop the security forces. As the Taliban insurgency grew, the United States took increased responsibility for reforming what became the Afghan National Defense and Security Forces (ANDSF). Efforts to rebuild the army and the police focused on strategies to integrate and disarm militias, as initially the United States had no plans for security force development. It was not until 2005 that the United States assumed the lead role to develop the ANDSF.
Strategies to develop the ANA and ANP diverged. After developing a set of interim solutions, in 2002 the international community decided to rebuild the organizational structure and the recruitment of the ANA from scratch. This decision was based on several factors, including concern that the nascent army had come under the control of mujahideen factions, but also an acknowledgment that the Soviet model that was designed for political control and that featured forced conscription remained as a sticky institutional structure. The decision to rebuild the ANA from the ground up meant that it would have an entirely new structure and political management and the Soviet political model would be dismantled.
In contrast, initially little attention was paid to the structure of the police. Police forces remained in the MoI, which, like other ministries, relied on central planning. By 2009, it was clear that the MoI lacked the “ability to perform basic management functions, particularly in personnel, procurement, and logistics; and an overall strategy for police operations and development. . . . Institutional reforms, which began in earnest in 2005, were routinely resisted or thwarted by political interference, often from the highest levels of government.”26 Unlike the ANA, the ANP sought to reintegrate demobilized combatants into the forces, which undermined development by allowing militia commanders to fill police ranks with patrons. By 2010, the ANP was so untenable that the United States encouraged the creation of village-based protection forces, such as the Afghan Local Police, which were based on customary models of self-defense as an alternative.
The most serious obstacle facing the creation of both the army and the police related to force size planning. When the ANA was initially established, planners imagined a force of 70,000 and a police force of 60,000. These relatively modest numbers ballooned during the period of the military surge when the projected size of the ANDSF skyrocketed to 196,000 ANA and 162,000 ANP.27 Analysts argued that it would be impossible to sustain such a large fighting force in a population that had such low levels of literacy without returning to a conscription-based model.
The outcomes in the ANP and the ANA could not be more striking. Although there are reports of corruption and ghost soldiers, the ANA has become “a symbol of nationhood and factional nonalignment.”28 A 2017 report on corruption in the security forces said that the decision to build the ANA from scratch as a national-multiethnic body protected it from capture by factional and criminal elements as had been the case with the ANP. A senior U.S. official was quoted in the report saying, “In the Ministry of Defense, the problem is contracts. . . . In the MoI it is everything.” In 2017, 44 percent of the population said the ANP was honest and fair. In contrast, 60 percent said the same of the ANA.29 A former interior minister stated that organizational reforms led to a professionalization of the army that reduced its casualty rates and the perception of its corruption compared to the police, which remained largely unreformed.30
Improving Prospects for Successful Stabilization Efforts
One way to improve prospects for successful stabilization and state-building is to recognize that democracy is not an end in itself. Rather, state-building looks to improve the ability of a government to provide public goods and services. Elections at the national level are not an indicator of successful state-building unless the state is also able to provide such goods and services.
Second, political decentralization ought to be considered carefully as a strategic approach for reorganizing failed or failing governments. Centralization may be associated with public goods provision in some now-wealthy democracies. However, conflict and violence have typically resulted in diffused power in fragile states, including power asserted by customary, tribal, and insurgent organizations that dominate politics outside capital cities. It is necessary to consider establishing political institutions that fit the local context, including the existing power structures rather than creating rules that conflict with existing realities.
Third, reform of the bureaucracy is often more directly responsible for improvements in public goods provision than national elections. It is also important to consider the prospects for self-governance as a solution to administrative challenges. Self-governing communities reduce the need for state intervention to provide public goods and can provide a framework for collaborative or shared governance.
What remains is the political problem introduced earlier. Politicians may not want to give up power; nor do bureaucrats who enjoy authority in the centralized status quo. The international community may not want to move away from relying on the old administrative structures. Nor might it want to invest resources in mundane public administration reforms when the world seems to care more about elections. Regardless, we need to more carefully explore the links between centralization of power and the failures of state-building efforts. PRISM
1 Bates, When Things Fell Apart (New York: Cambridge University Press, 2008).
2 Roger B. Myerson, “Rethinking the Fundamentals of State-Building,” PRISM 2, no. 2 (2011): 91–100.
3 Elinor Ostrom, Understanding Institutional Diversity (Princeton, NJ: Princeton University Press, 2005).
4 Roger Myerson, “Constitutional Structures for a Strong Democracy: Considerations on the Government of Pakistan,” World Development 53 (2014): 46–54.
5 Mancur Olson, “Dictatorship, Democracy, and Development,” American Political Science Review 87, no. 3 (1993): 567–576.
6 Ronald W. Batchelder and Herman Freudenberger, “On the Rational Origins of the Modern Centralized State,” Explorations in Economic History 20, no. 1 (1983): 1–13.
7 Barry R. Weingast, “The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development,” Journal of Law, Economics, and Organization 11, no. 1 (1995): 1–31.
8 Jennifer Murtazashvili and Ilia Murtazashvili, “Wealth-Destroying States,” Public Choice, 2019.
9 George Tsebelis, Veto Players: How Political Institutions Work (Princeton, NJ: Princeton University Press, 2002).
10 Francis Fukuyama, “What Is Governance?” Governance 26, no. 3 (2013): 347–368.
11 Barnett R. Rubin, “Crafting a Constitution for Afghanistan,” Journal of Democracy 15, no. 3 (2004): 5–19.
12 Rubin, 12.
13 Asia Foundation, A Survey of the Afghan People: Afghanistan in 2017 (Washington, DC: The Asia Foundation, 2017), 111.
14 Jennifer Murtazashvili, Informal Order and the State in Afghanistan (New York: Cambridge University Press, 2016).
15 Asia Foundation, A Survey of the Afghan People.
16 Amin Tarzi, “The Judicial State: Evolution and Centralization of the Courts in Afghanistan, 1883–1896” (doctoral dissertation, New York University, 2003).
17 M.G. Weinbaum, “Legal Elites in Afghan Society,” International Journal of Middle East Studies 12, no. 1 (August 1980): 39–57.
18 International Crisis Group (ICG), “Reforming Afghanistan’s Broken Judiciary” (Kabul: International Crisis Group, 2010), 6.
19 ICG, 6.
20 ICG, “Afghanistan: Judicial Reform and Transitional Justice” (Brussels/Kabul: International Crisis Group, January 2003), 9.
21 Scott Worden and Sylvia Q. Sinha, “Constitutional Interpretation and the Continuing Crisis in Afghanistan” (Washington, DC: United States Institute of Peace, November 7, 2011).
22 Michael Carnahan, Reforming Fiscal and Economic Management in Afghanistan (Washington, DC: World Bank Publications, 2004).
23 Asia Foundation, A Survey of the Afghan People.
24 Nematullah Bizhan, Aid Paradoxes in Afghanistan: Building and Undermining the State (London: Routledge, 2017).
25 Antonio Giustozzi, The Army of Afghanistan: A Political History of a Fragile Institution (London: Hurst, 2016), 44.
26 Robert M. Perito, “Afghanistan’s Police” (Washington, DC: United States Institute of Peace, 2009), 12.
27 SIGAR, “Reconstructing the Afghan National Defense and Security Forces: Lessons from the U.S. Experience in Afghanistan” (Washington, DC: Special Inspector General on Afghanistan Reconstruction, September 2017).
28 Kenneth Katzman and Clayton Thomas, “Afghanistan: Post-Taliban Governance, Security, and U.S. Policy” (Washington, DC: Congressional Research Service, 2017).
29 Asia Foundation, A Survey of the Afghan People, 63.
30 Ali A. Jalali, “Afghanistan: The Challenge of State Building,” in Afghanistan: Transition under Threat, ed. Geoffrey Hayes and Mark Sedra (Waterloo, ON: Wilfrid Laurier University Press, 2009).