Anarchy, self-governance, and legal titling

Abstract

Scholars and policymakers prescribe legal titling to improve prospects for economic development and political order. However, a public choice literature exists that has long recognized that self-governance often works well and that the state may not be able to improve upon local economic institutions at reasonable cost. Although the implication that legal titling should proceed with caution is seemingly straightforward, the literature on legal titling does not take anarchy seriously as a policy option. In addition, there is a public choice literature that presumes the state is the most important source of property rights. This essay fills this gap in the property rights literature by applying the concept of “efficient anarchy” to legal titling in Afghanistan. Original fieldwork evidence from rural Afghanistan suggests that anarchy of land governance is a better option than legal titling. The essay concludes by opening up the black box of state building by explaining why it often makes sense to sequence improvements in political capacity and political constraints prior to investing in legal titling.

1 Introduction

Scholars and policymakers prescribe legal titling in the hope of increasing access to credit, improving private investment incentives, and increasing investment in public goods as the government enlarges its revenue base (Deininger and Feder 2009; Field 2005; Galiani and Schargrodsky 2010). Legal titling is also thought to improve political order by reducing the economic return to violence (Albertus and Kaplan 2013; Fearon and Laitin 2011). Yet the case for legal titling is not as obvious as it might seem. Many legal titling programs have been subject to capture by interested parties or fostered conflict over land (Boone 2013; Lund 2008). Moreover, much of the land in the developing world is not sufficiently valuable to justify the cost of issuing legal titles (Arruñada 2012; Bromley 2009; Bromley and Anderson 2012; Herbst 2000). More fundamentally, self-governing property arrangements are often quite effective, which implies that formal registration of ownership may not be necessary (Ostrom 1990, 2005).

These criticisms of legal titling notwithstanding, this essay contends that the literature on legal titling does not provide a satisfactory framework to consider when and where legal titling is efficient because it neglects anarchy as a serious policy option.¹ To alleviate this shortcoming, we use the concept of efficient anarchy, which explicitly considers whether state intervention in community economic relations is wealth maximizing (Ellickson 1991; Leeson 2006, 2014), to analyze the efficiency of legal titling. The literature on efficient anarchy, rather than presuming the inevitability or efficiency of state intervention, is explicit about the possibility that doing nothing may be efficient in an economic sense.

Applying the concept of efficient anarchy to legal titling yields three primary benefits. First, the literature on efficient anarchy clarifies that legal titling, because it is a profound extension of the state in local affairs, invites a host of potential incentive and implementation problems that come with this sort of ambitious projection of state power. Despite these foreseeable costs of state intervention in informal land relations, the literature on legal titling does not take anarchy seriously. For example, De Soto (2000) does not explicitly consider the possibility of anarchy as an alternative to legal titling. Even public choice studies critical of legal titling, such as Kerekes and Williamson (2008, 2010), do not explicitly consider anarchy as a policy choice.

Second, the concept of efficient anarchy responds to a variant of the public choice literature that focuses perhaps disproportionately on the state as a source of property rights. An important strand of public choice literature presumes that a powerful state is necessary to provide private property rights (Olson 1993; Riker and Sened 1991; Sened 1997). We explain why the presumption that property rights come from the state is particularly troubling in analyzing the efficiency of legal titling.

Third, the literature on efficient anarchy explicitly considers “governance costs.” We build upon earlier insights by suggesting that governance costs consist of several broad categories that include administrative capacity, adjudicative capacity, and costs of establishing constraints on the state. Our reading of the public choice literature is that it focuses overwhelmingly on constraints (Leeson 2007a). Yet as Fukuyama (2013, 2014) explains, administrative capacity is a central component of high-quality governance. We explain why it is important to consider state capacity to record land relations, enforce property rights, and to facilitate dispute resolution, in addition to constraints on the state.

We use evidence from Afghanistan to understand the extent to which legal titling is efficient. Afghanistan is an important case because legal titling has become part of the state-building process even though self-governance is a potentially important option. The evidence includes original fieldwork conducted by the authors in 30 villages between 2006 and 2008, as well as a nationally representative survey the authors helped to design that was implemented in 2011.

There are two main reasons why a broad range of original evidence is necessary to draw credible conclusions regarding the efficiency of legal titling. First, the broad range of evidence permits us to make generalizable statements about the extent to which self-governance is a serious policy option in a country. In a study of Peru, Kerekes and Williamson (2010) conclude that legal title might not lead to more security based on fieldwork that consists of interviews with 20 people in a single community. However, as we are not typically interested necessarily in whether legal titling works or does not work in a single community alone, we require fieldwork designs that are more likely to yield generalizable conclusions about self-governance. Since legal titling it generally implemented in many communities or even in an entire country, it is important to understand the extent to which self-governance works throughout the area of implementation. Second, information about informal property institutions at the local level is necessary to conclude whether or not legal titling is efficient. Williamson and Kerekes (2011), Williamson and Mathers (2011) and Williamson (2009) consider how informal institutions influence the quality of formal property rights. However, these cross-sectional studies measure informal institutions in terms of trust, obedience, and individual self-determination, rather than measuring the extent of informal property rights protection. As such, these cross-sectional studies do not provide much guidance regarding the efficiency of legal titling since policy conclusions regarding legal titling require that we understand the quality of informal private property institutions.

Finally, we open up the “black box” of state building by clarifying when it makes sense to sequence legal titling before, during, or after generalized investments in state capacity and constraints on it. State-building efforts typically include a large number of costly, simultaneous reforms that present serious challenges to implementation and evaluation of those reforms (Andrews 2013; Coyne 2008a, b; Pritchett et al. 2013). We show why it often makes more sense to implement legal titling after improving state capacity and successfully establishing political constraints that lend credibility to legal titles.

This essay is organized as follows. Section 2 situates our study in the literature on anarchy and social order. Section 3 applies insights from the concept of efficient anarchy to legal titling. Section 4 describes the fieldwork and survey research, as well explains why a broad range of evidence is critical to assessing the efficiency of legal titling. Section 5 considers empirically whether anarchy is efficient in rural Afghanistan. The conclusion addresses the question of sequence between legal titling and investment in capacity and constraints on the government.

2 Efficient anarchy: the political economy of doing nothing

The Hobbesian view that there can be no “mine and thine” without an all-powerful Leviathan to establish property rights remains compelling in many circles. The “old” institutional economics, for example, emphasizes the role of the state in emergence of property rights (Bromley 2006; Commons 1924; Hodgson 2009). As Bromley (2006) explains, claims to ownership without the backing of the state are merely aspirational since the concept of a right—including a property right—implies that individuals can call on the state to enforce it.²

There is also an important strand of thought in public choice and new institutional economics that views the state as the fundamental source of property rights. Riker and Sened (1991) and Sened (1997) presume that property rights are public goods provided by the state. Similarly, Olson (1993), North and Thomas (1973) and North (1981) rationalize the state as an efficient response to property insecurity arising under alternative political organizations.

The argument that the state is an efficient solution to the problem of property insecurity as economic activities extended beyond the boundaries of small communities is uncontroversial. However, “law” need not arise within the confines of a state (Hadfield and Weingast 2012, 2013). Individuals often cooperate or coordinate in the absence of a state, or in spite of it (Greif 2005, 2006; Landa 1981; Milgrom and North 1990; Powell and Stringham 2009; Skarbek 2011, 2012; Sugden 1989). In some situations, these informal rules may even have characteristics similar to laws enforced by the state.

The experience on the American frontier provides ample examples of order in the pale shadow of the state. Anderson and Hill (1990, 2002, 2004) depicted how frontier settlers cooperated without relying on the American government in an “anarcho-capitalist” environment. During the California gold rush that commenced in 1848, individuals established informal systems of property rights to allocate access to gold deposits (Umbeck 1977).³ In fact, individuals in each of the major frontier sectors jumped the gun on land settlement by establishing government-like organizations to specify and enforce property rights during times when they had no formal rights to do so (Murtazashvili 2013).

The usual policy implication when confronted with a robust system of informal property institutions is to formalize them. The reason is that recognition of informal institutions promises to reduce the transaction costs of establishing formal property rights (Barzel 1989; Libecap 1989, 2007). Yet even formalization of de facto property rights is a form of state intervention. In the presence of effective informal systems of governance, the best option may not be formalization; it may be to do nothing at all.

The literature on anarchy is explicit about doing nothing as a policy option. Anarchy, as Leeson (2006, 2014) defines it, refers to the absence of state control of economic, social, and political activities. Anarchy may be socially beneficial when the state is exploitative or simply incapable of governing (Leeson 2007b; Scott 2009, 2012). Although the state may seem inevitable as an enforcer of property rights, it is not clear that the state is functional enough to provide property rights protection in much of the developing world. Nor is it is obvious that governments necessarily have the right incentives to respect the property rights embodied in legal titles. The following section explicitly considers the implications of anarchy for legal titling.

3 Efficient legal titling

3.1 The benefits of legal titling

Efficiency in the process of institutional change requires that the benefits of institutional change exceed the costs (Knight and North 1997; Riker and Weimer 1995). Several broad benefit and cost categories together provide a framework for judging when legal titling is efficient. The list that follows is not necessarily exhaustive, although we believe the categories below are important ones to include in any such list.

The first benefit category that determines the efficiency of legal titling is the extent to which legal titles improve land-tenure security. Security from trespass is one of the fundamental dimensions of a property right, along with clarity of allocation, alienability, and credibility of persistence (Riker and Weimer 1993, 1995; Weimer 1997). Security from trespass refers to the extent that one is able to exclude non-authorized users from land.

Determining the gains in security requires comparisons of property security of those who have legal titles versus those who do not have formal recognition of their ownership. The self-governance literature finds that in some situations, individuals are able to establish informal property rights that provide claimants with a sense of security. As the state’s legal titles may not be associated with security, it is necessary to show that people without legal titles suffer property insecurity before investing in legal titling.

The second category of benefits of legal titling is the improvement in prospects for adjudication through formalization of legal titles. By providing legal titles, the state implicitly promises to provide opportunities to resolve land conflicts. Issuing legal titles without simultaneously providing a forum to resolve land conflicts is a hollow promise since legal titles are worth less when disputes must be resolved without the aid of a formal legal apparatus.

The reason why improvements in prospects for adjudication are an important benefit category is that such forums influence clarity of allocation of private property rights, which refers to how clearly ownership is delineated. When disputes are unresolved, clarity of allocation declines because it is unclear who actually owns land. Although it is often presumed that security is the fundamental outcome of interest associated with a legal title, the quality of a property right also depends on clarity of allocation. Consequently, analyzing prospects for dispute resolution will be important in assessing the efficiency gains from legal titling.

The third source of benefits of legal titling is the extent to which economic interactions are impersonal. Leeson (2006) focused on impersonal exchange as a fundamental justification for expanding the role of the state in governance of economic activities. In the context of property rights, when interactions are impersonal, individuals are more likely to require a legal system of property ownership since it is less likely that contracting parties will possess requisite knowledge of local property arrangements. In contrast, when exchange is primarily local, there will be fewer benefits from formalization through issuance of state-backed legal titles since informal norms are often quite able to allocate ownership rights in such situations (Alston et al. 1999).

3.2 The costs of legal titling

The first and most obvious cost of legal titling is surveying, documenting, and registering land. Establishing a system of land registration requires cadastral surveys, generating images of landholding, and recording of deeds. In addition to registration, it is critical that repositories of information are accessible in order for individuals to utilize information regarding land ownership (Arruñada 2012). In a fragile state, where officials may be unable to conduct a census, let alone a cadaster, documentation typically will be a major challenge, and quite costly.⁴

A second cost of legal tilting is extending the adjudicative capacity of the state. As noted above, a legal title is an implicit promise that the state will help individuals resolve disputes. Consequently, the state generally will have to invest in additional capacity to resolve disputes alongside investments in recording land ownership.

The third cost category is improving enforcement capacity. It is possible for individuals to invest in self-protection to secure their possessions (Onoma 2009; Sjaastad and Bromley 1997). However, when the state issues legal titles, it must also have the capacity to enforce those rights or the ability to extend its capacity to enforce property rights.

Enforcement is an especially important issue in fragile states. The reason is that such states typically have very limited abilities to implement and enforce reform of any sort (Andrews 2013; Pritchett et al. 2013). For this reason, enforcement generally is a major cost the state must bear as it issues legal titles.

A fourth cost of legal titling is establishing the institutions that govern the credibility of commitment of the state to property rights. Economic theories of the state illustrate that rulers have self-interested reasons to provide property protection to increase their revenues (Olson 1993). However, rulers may not be able to commit to respecting those property rights once investments are made (Gehlbach and Keefer 2011; Leeson 2007a; North and Weingast 1989; Riker and Weimer 1993). Political institutions that constrain the state, such as federalism or another scheme of decentralized governance, allow the state overcome its commitment problems (Leeson 2011; Weingast 1995). Fragile states typically are afflicted by an inability to commit credibly to reform (Coyne and Boettke 2009; Flores and Nooruddin 2012). It is therefore likely that the state will have to invest in establishing credibility-enhancing political institutions as it issues legal titles.

Table 1 summarizes each of the dimensions of benefits and costs, as well as different measures of each category.

Table 1 Benefits and costs of legal titling

The case for legal titling is strongest when anticipated benefits—measured across these specific categories—are large and the costs of establishing legal titles are small. When there are benefits from state-backed legal titling but substantial costs, anarchy may be an appropriate policy solution. In such situations of benefits alongside prohibitive costs, efficiency requires sequencing generalized improvements in state building, such as investments in policing and a more competent court system, prior to establishing legal titles. Finally, when there are few benefits and substantial costs of legal titling, the case for anarchy is strongest.

4 The evidence

Our empirical study considers land governance in rural Afghanistan. We focus on rural parts of the country because 80 % of Afghans live in rural areas, typically in villages of less than 500 people.⁵ In order to understand when legal titling is efficient in rural Afghanistan, as opposed to whether it is efficient in a single community or two, we conducted fieldwork in many villages throughout the country, as well as helped to design a nationally representative survey that included a number of land-related questions.

Most of the village-level fieldwork for this study between 2006 and 2008. One of the authors, with support from the Afghanistan Research and Evaluation Unit, an Afghan research organization headquartered in Kabul, led a team of male and female Afghan researchers to conduct qualitative research, including semi-structured interviews, focus-group discussions, and participant observation in 32 villages in 17 districts across six provinces in rural Afghanistan.⁶ In each village, dozens of interviews were conducted with informants that included community leaders, local government officials, religious officials, participants in donor-sponsored development projects, as well as randomly selected villagers. Two villages were chosen in each district for field visits, one that was close to the district center and one that was farther away. Villages were selected to ensure inclusion of all major ethnic groups. Members of the research team conducted interviews in three local languages: Pashto, Dari, and Uzbek, with effort made to select an equal number of men and women. In addition, the authors conducted interviews at district and provincial capitals with governors and their staffs, NGO officials, and journalists. This fieldwork is much broader in its empirical scope than existing ethnographic studies of economic institutions.⁷ This empirical perspective also complements historical studies of economic institutions in Afghanistan by providing original evidence of informal economic institutions.⁸

In addition to the qualitative evidence, this essay draws on an original, nationally representative public opinion survey the authors helped to design that included 8,620 fully completed face-to-face survey interviews across all 34 provinces of Afghanistan. Democracy International, an American non-government organization, with support by the United States Agency for International Development, conducted the survey. The survey employed a five-stage sampling plan that included randomized district selection and a random selection of within-district settlements. Interviewers selected households by having the interviewer start at a recognizable starting point (e.g., mosque, school, or bazaar) and then heading in a randomly determined direction. In addition, the analysis also relies on the Asia Foundation Survey of the Afghan People, another nationally representative public opinion survey conducted annually since 2006 that provides insight into dispute resolution. In presenting our findings, we focus on the rural subpopulations of these surveys.

A broad range of evidence helps in understanding just how well self-governance works in a fragile state. For example, Leeson (2007b) suggests self-governance is an option in Somalia. However, it is not exactly clear how well self-governing arrangements work throughout the country (Menkhaus 2007). Extensive fieldwork helps to alleviate concerns that examples of self-governance may not generalize. The surveys are also helpful for understanding whether the findings from our sample of villages selected for fieldwork generalize. Although fieldwork in dozens of villages is more likely to provide an accurate picture of the country, Afghanistan has tens of thousands of villages, and so survey evidence helps understand the extent of self-governing property arrangements. More generally, fieldwork and survey evidence are both necessary because there is very little systematic evidence of informal governance and informal economic institutions in post-Taliban Afghanistan.

5 Is legal titling efficient in Afghanistan?

5.1 The formal property regime

Afghanistan has had a highly centralized system of government since Abdur Rahman attempted to consolidate political authority during a brutal reign from 1880 to 1901. Previously, Afghanistan was more like a loose confederation of quasi-independent tribes than the centralized states of Europe (Newell 1972). As part of this strategy of consolidation, Abdur Rahman declared all land to be Crown property, subject to reallocation by sovereign decree, as well as prohibited the development of railroads (Barfield 2010; Kakar 1979). The Iron Amir, as Abdur Rahman came to be known, believed that a poor country would be less attractive to foreign meddling (Rahman 1900). Although subsequent leaders may have eschewed the brutality of the Iron Amir, they generally aspired to the ideal of a centralized state.

Land governance remains highly centralized under the Land Management Law, adopted in 1965, which provides the legal framework for land relations. Under Afghan land laws, which various governments have revised since 1965, district courts are the main source of authority over land governance and legal titles are the only recognized ownership documents. Nor did Afghanistan’s communist regime, which governed from April 1978 to 1992, recognize informal ownership documents or customary systems of governance. Rather, Afghan communists viewed customary governance and social norms as backward and “feudal,” electing instead to coopt and subjugate customary authority, as well as to redistribute land (Edwards 2002).⁹ In contrast, the Taliban Land Law of 2000 provided a framework for restitution of past expropriations of land and recognized customary deeds. However, the Taliban, which ruled the country from 1996 to 2001, generally was unable to implement its laws (Rashid 2010).

Since the fall of the Taliban government in 2001, the central government reverted to its old ways under the regime of Hamid Karzai, shunning customary authority and centralizing government control over property rights. Some reforms to the land administration system, such as the creation of an Independent Land Authority, were implemented, but those reforms maintained the centralized system of land adjudication, with authority centered on primary state courts, without any formal role for customary governance (Alden Wily 2013). In addition, the current Land Management Law, revised once again in 2008, does not provide a clear role for customary deeds or customary adjudication. The incoherence of the formal property regime has led to a festering of land disputes in many areas that in some situations may even strengthen local support for the Taliban to resolve land conflicts (Giustozzi et al. 2012).

5.2 Self-governance in rural Afghanistan

Although formally centralized, governance in Afghanistan is polycentric in practice (Murtazashvili 2015). Customary authority in Afghan villages usually is divided between three distinct organizations: village leaders (maliks), village councils (shuras or jirgas), and religious arbiters (mullahs). Maliks historically have been selected through consent of the governed or descent (Roy 1990).¹⁰ Although the literature in political economy increasingly recognizes the role of traditional “chiefs” in development (Acemoglu et al. 2014; Baldwin 2013), and maliks are sometimes described as a kind of feudal chief or lord (Beath et al. 2013), our evidence suggests maliks are better thought of as a first among equals, rather than as chiefs. Villagers often select maliks through a deliberative process, oftentimes choosing as the malik one of the more educated members of a village.

Shuras (sometimes referred to simply as “elders” or “white beards”) are the fundamental deliberative organizations in rural Afghanistan. These ad hoc village-level councils usually are comprised of “elders,” which is actually a misnomer since many members of a shura are in their thirties or forties. Rather, an elder is a respected member of the community. These elders, who usually are men, consider issues of collective importance such as land disputes, local conflicts, criminal matters, and relations with the government. However, shuras do not necessarily deal with all land issues because many disputes do not require community-wide resolution, and so maliks or religious leaders often resolve them.

Mullahs are local spiritual leaders who provide religious guidance to community members. They also play a role in resolving conflicts within families over matters such as inheritance, divorce, or other issues concerning Islamic Law. The Taliban elevated the role of mullahs in local politics, administering villages through its own mullah network. In post-2001 Afghanistan, mullahs continue to participate in local politics, frequently participating in shura processes involving contentious issues within and across communities.

Although governments of all ideological stripes—monarchs, communists, and, more recently, religious autocrats such as the Taliban—have attempted to stamp them out, customary forms of governance remain common across all Afghan ethnic groups. According to the Democracy International survey, conducted in 2011, almost all individuals in rural Afghanistan have access to a malik (over 99 %), while the vast majority have access to shura (around 87 %) and mullah (around 77 %).¹¹ Although some scholars maintained that customary authority was decimated during decades of war (Rubin 2002), one of the contributions of our fieldwork and survey evidence is to show that self-governing organizations endured decades of war.

Self-governance is especially important when it comes to property relations. According to our survey, customary organizations are also the main source of property rights in rural Afghanistan. Around 85 % of rural Afghans indicated that they own some land. However, only 19 % of these respondents have a state-issued legal title, while 94 % have customary deeds to their land. In addition, as we explain more fully below, these customary organizations are the main source of dispute resolution in the country.

Despite the presence of robust systems of customary land governance, the state-building process has proceeded with legal titling in urban parts of the country. There have also been several limited-scale legal titling projects in rural parts of the country (Stanfield et al. 2013). The remaining sections consider whether such exercises in legal titling are efficient, as well as whether it would be efficient to roll out such projects more generally in rural Afghanistan.

5.3 The benefits from legal titling in Afghanistan

The first category of benefits from legal titling is the expected improvement in security of property rights by issuing state-backed ownership documents. Our fieldwork suggests that few gains in terms of security would materialize from formalization. During field interviews in each of the 32 villages visited, we asked informants about pressing problems in their community. As these were field interviews, we were able to ask open-ended questions, probing informants for additional information. It turns out that informants rarely mentioned that they suffered from land-tenure insecurity. They typically explained that the problem in their community was a lack of capital to provide basic public goods or threats from insurgents such as the Taliban, rather than ownership insecurity. Nor did field interviews produce evidence that customary representatives or elders meddle in property rights or are complicit in land grabbing. The exceptions were when in some situations, outsiders—warlords, Taliban, or commanders—came into communities and in the process pushed aside customary representatives. Yet informants contrasted the behavior of these interlopers with the behavior of customary representatives and elders.¹² Rather, villages tended to describe the latter group as responsive to citizens’ demands. As we asked about important problems facing communities, and almost none of the informants mentioned land-tenure insecurity, we take this as evidence that customary deeds convey a sense of land-tenure security.

The exception to the finding that rural Afghans generally feel secure in their property possession was in regions where nomads and settled communities staked a claim to the same land. Around a million Afghan nomads, known as Kuchis, continue to migrate seasonally from their traditional homes in search of winter pasture (Alden Wily 2003, 2009).¹³ Violent conflict has led nomads to alter traditional migration patterns, resulting in interactions between groups that have little prior history. In addition, many nomads have very old legal titles—sometimes more than a century old—that contribute to land conflict. Some of these titles originated during the times of Abdur Rahman, who promised Pashtun nomads land rights in exchange for support in efforts to mollify opposition to his rule in the central parts of the country (Barfield 2010), while others originated with the redistributive campaigns of the communist government in the late 1970s and early 1980s.

These conflicts weighed heavily on the minds of Hazara informants in Panjab District, Bamiyan Province, which neighbors Behsud. When asked in field interviews to discuss the security situation in his district, the district governor mentioned the threat of such migration to peace and security:

Another problem we face in this area is the Kuchi issue. The Kuchis claim that they have a lot of land in this area. If hundreds of Kuchis come to this area, they will destroy the agriculture of our district….The second problem is that if the Kuchis return, then terrorists will accompany them—like Al Qaeda and the Taliban.¹⁴

A focus group discussion with women in a village in this district also betrayed the fear associated with nomadic return:

Security is good now. Sometimes some looting happens, but our main concern is about the Kuchis. If the Kuchis come to our village, they will destroy our agriculture by driving their animals here. Our second problem with Kuchis is it that we are afraid that the Taliban will come with the Kuchis and they will create instability in our area.¹⁵

In situations where nomadic return undermines self-governing property arrangements, the case for state intervention in rural land relations is stronger (Stanfield et al. 2010, 2013). However, rural Afghans generally are less concerned with land-tenure insecurity than other problems, and so we conclude that there would likely be little gained through issuance of state-backed legal titles, except in a few parts of the country.

The second category of benefits of legal titling is the improvement in quality of adjudication through formalization of private property rights. As anticipated by the work of Ellickson (1991) and Leeson and Coyne (2012), who have shown that rules-in-use in communities which are agreed upon by community members are often more efficient than formal legal venues in resolving “legal” conflicts, customary modes of land governance are often effective in resolving disputes over land. As an informants from a village in Anjil District, Herat Province, explained:

For example, two brothers fought in the village. One brother sold land without permission of the other brother. When the brother found out that his brother sold his land, he beat his brother. After this, the malik gathered the people in the village together and the malik took these two sparring brothers to the khalifa [religious leader]. The brother who sold the land was told to host a party for the people in the village so that he could formally apologize to his brother and the people in the village for what he did. The malik found a resolution that satisfied both of the brothers.¹⁶

A malik from Anjil provided an example of how he resolved a land conflict:

Recently, there was a fight between cousins over land. Both were fighting over water and irrigation. One of the cousins accidentally killed the other in a fight. We (the families and the malik) discussed and we resolved that problem. It was just [an accident]. But according to custom in Afghanistan we should take money from the murderer’s family and give it to the victim’s. So you can see, we had many problems but we resolved all of them.¹⁷

The shura provides a complementary source of dispute resolution. An informant from a village in Bamiyan Center, Bamiyan Province explained how they resolved a conflict over agricultural land between two brothers:

[T]he white beards from the village sat and distributed the land between two brothers and they calmed down. Even when people approach the government to resolve disputes, the government refers these disputes back to the white beards in the village for resolution.¹⁸

In contrast to the faith placed in customary bodies, villagers expressed little confidence in formal authorities to resolve land conflict. When villagers petitioned the state to resolve disputes, the venues were either the woluswal (a district governor whom officials in Kabul appoint) or courts. In about half of the interviews where the performance of a woluswal came up, informants expressed trust in their woluswal. However, the informants who mentioned the woluswal as a source of governance also explained that in most cases, the woluswal was more like an elder in that he applied customary rather than formal law to resolve land conflicts. The confidence in these government officials therefore derived from their competence in resolving disputes.

Survey evidence provides a picture of dispute resolution similar to our fieldwork. Alden Wily (2013) reviews the survey evidence regarding customary resolution of land disputes. According to the Norwegian Refugee Council, “poor people” found land disputes “to be more meaningfully resolved outside the courts in local shuras.” An Oxfam survey in 2008 concluded that only “eight percent of respondents said they would use courts to resolve land disputes,” while the Asia Foundation reported that up to 90 % of disputes—many involving land—were resolved outside the formal system. Indeed, the majority of Afghans “have never been near a court” (Alden Wily 2013, p. 91). When asked whether they were able to resolve a dispute by venue, the Asia Foundation (2011) found 70 % who took the dispute to a mullah had resolved it, 64 % who took it to a malik were able to resolve it, and 59 % who took it to a shura/jirga were able to resolve their dispute. Government agencies (52 %) and district authorities (51 %), which include courts, have lower rates of resolving disputes than any of the customary channels. Thus, the evidence suggests few gains from legal representation.

Finally, the small scale of village activities also suggests there may be few gains from legal titling. Most Afghans in rural areas live in small villages of around 500 people and often stay there for most of their lives.¹⁹ The vast majority interviewed said that they and their families had lived in the same village for a very long time (in fact, several mentioned that their families initially came to the village because of repopulation campaigns conducted by Abdur Rahman). The accounts from individuals in nearly every village visited suggest that economic interactions are mainly local. The personal nature of economic relations suggests there would be few gains from legal representation, in particular in rural parts of the country.

5.4 The costs of legal titling in Afghanistan

The costs of registering land in Afghanistan are enormous. The government last conducted a cadastral survey in 1965; civil war cut short the next attempt in the late 1970s. Against challenging odds, the land registry offices throughout rural Afghanistan have tried to maintain as many land titles as possible during decades of conflict. Unfortunately, the variation in the condition of land records throughout the country is considerable, and many records have been destroyed (Safar and Stanfield 2007).

The government has not conducted a national cadaster since the fall of the Taliban in 2001. Growing insecurity means that a cadaster is unlikely even in the post-Taliban period. It is possible, as some have argued, that a cadaster would improve prospects for security under the presumption that better registration could reduce land conflicts (Batson 2008, 2013). However, to conduct a cadaster, substantial improvement in the security situation is required so that the government is able to demarcate land ownership.

The costs of extending formal adjudicative and enforcement capacity are also substantial. Despite tremendous investment to strengthen courts, including investments in personnel and physical infrastructure, the court system remains plagued by a sense of illegitimacy and corruption (Barfield 2008; Barfield et al. 2011). During fieldwork, several of the woluswals we interviewed said they used the threat of adjudication in state courts to compel people to resolve conflicts over land through customary channels.

The role of the Taliban in adjudication of land conflicts also indicates the weakness of formal courts. The Taliban have taken advantage of the perceptions of government corruption as well as weak capacity to adjudicate and enforce dispute resolution in an effort to “win hearts and minds” of Afghans by providing justice for Afghans, including by attempting to redress land grievances created by government-backed land-grabbing (Malkasian 2013). In an impressive series of interviews with Taliban officials as well as those who used Taliban courts, Giustozzi et al. (2012) found that many Afghans find that the Taliban provide effective and swift adjudication. These competencies of the Taliban suggest just how much investment is necessary to improve the state’s capacity to adjudicate land conflicts.

There are also tremendous costs of extending the role of the state as an enforcer of property rights. According to the Democracy International survey, people have little confidence in police: 50 % expressed “no confidence” in the police while 33 % have “some confidence.” Only 18 % have “a lot of confidence” in police. Distrust of police likely reflects a more general problem with illegitimacy of the state, including distrust of those who are supposed to enforce its laws, as well as substantial costs of increasing the capacity of police to enforce property rights.

Finally, establishing credibility-enhancing institutions is costly. Land grabbing illustrates the government’s credibility problem. Journalistic accounts show that the government is often complicit in land-grabbing, and that misuse of state power is one of the fundamental explanations for land-tenure insecurity in the country (Foschini 2013; Peter 2011). In fact, a study commissioned by the Afghan government in 2013 found that found that most land grabbing has occurred with the blessing of the state (Sediqi 2013).

In theory, several rounds of elections and a commitment to democracy in the 2004 Afghan constitution should improve credibility of the government to commit to formal property rights. However, elections at the village level have yet to be held and district and provincial officials are accountable to Kabul, rather than to citizens (Murtazashvili 2014). Consequently, the government has a long way to go to establish the kinds of institutional constraints that are associated with credible commitments to respect for private property rights. Table 2 summarizes the benefits and costs of legal titling on these various dimension.

Table 2 Benefits and costs of legal titling in rural Afghanistan

6 Conclusion

All is not well in Afghanistan as far as land relations are concerned. The formal legal system is highly centralized and out of step with the reality of informal governance. However, legal titling is not necessarily the best option in Afghanistan given the current state of political and legal institutions in the country. The presence of self-governing organizations and the substantial costs of increasing state capacity and establishing constraints on politicians suggest that legal titling will produce few benefits and substantial costs. Despite this seemingly commonsense conclusion, anarchy of land governance was never a serious option in the state-buildings process. This essay suggests it is time to rethink anarchy as a policy option in Afghanistan and similar fragile states.

State-building processes typically involve many simultaneous reforms. Issues of sequence are often lost in the plethora of reforms. One of the implications of our study for state-building processes more generally is that the sequence of improvements in state capacity and constraints and legal titling matters. The reason why sequence matters for state-building efforts is that getting the right mix of capacity and constraints will ultimately determine the extent to which investment in legal titling is efficient.

One of the challenges confronting dispassionate analysis of legal titling is that providing people with state-issued ownership documents is firmly entrenched now in development policy. Accordingly, adopting a more moderate and cautious approach requires changing the mental models of many actors in the international community, as well as many domestic policymakers. Unfortunately, as the public choice literature has long recognized, even if doing nothing is optimal, there are few reasons to presume that bureaucratic processes anywhere, let alone in fragile states, will choose efficient policies. Nonetheless, we are hopeful that insights from public choice will at least encourage some reflection regarding the legal titling enterprise in Afghanistan and elsewhere.