1Control of access to land is central to power relations in rural Afghanistan and conflict over land has been exacerbated by the instability of the last three decades (Alden Wily, 2003; Alden Wily, 2004; Patterson, 2004). However, Afghanistan is an enormously complex country. Consequently, generalised statements about how things do or do not work and normative positions on what should be done have run aground on the sands of Afghanistan’s history and contemporary social order (Mielke et al., 2011).
- 1 It is not possible to give precise figures for Afghanistan’s population levels, their evolution thr (...)
2This paper engages with the theme of conflict over land by focussing on three interrelated dimensions of this complexity, which help to bring some clarity to our understanding of the diverse linkages between land, power and conflict. Land relations are not static, and current conflicts have to be seen in historical context. Underlying structural inequalities in the access to land of different social groups and economic classes have been brought to the surface not only by identity-based conflicts and unaccountable power, but also by increasing rural populations,1 shrinking farm sizes, drought and a rural economy that has failed to transform itself in the ways in which other agrarian-based economies in Asia have. Markets have failed to deliver prosperity and, given the failure of the state to ensure welfare, the growing rural population is trapped in rural villages, drawing on the only two durable institutions from which a degree of welfare can be obtained – the household and the village.
- 2 Muhummad Ibn Khaldun, The Muqaddimah: an introduction to history, ed N.J. Darwood, trans Fanz Rosen (...)
3Any discussion of land and its role in the economy of Afghan households and villages has to take into account the key structural variables that underlie the patterning and complexity of Afghan rural livelihoods at a regional and a more local level. One dimension is geographical, which, at the highest level of simplification, contrasts the mountains with the plains. As noted by the medieval Arab social historian Muhummad Ibn Khaldun2 , there exist deep structural contrasts between the geographically marginal areas constituted by mountains (and the interstitial arid plains), on the one hand, and the intensively irrigated, densely populated plains, on the other. The latter are characterised by agricultural surplus and high levels of social inequality, based on the ownership of land and strong social hierarchies. The more marginal, mountainous areas characteristically have subsistence-based economies, relatively low population densities and more equitable, less hierarchical social structures with strong group solidarities. Thus, the Pashtuns of the mountains in eastern Afghanistan have relatively egalitarian tribal structures, within which traditional leaders are held accountable through local councils or jirgas (Barth, 1959). In contrast, the plains have a more hierarchical social structure with strong social patron – client relations based on inequalities in land ownership, as typified by the Durranis living close to cities such as Kandahar (Guistozzi and Ullah, 2006).
4A second dimension to be taken into account is the distinction between privately held land (amlaki shakhsi) and the other four classes of land, defined as government land, public land, communal land and land belonging to religious institutions (Alden Wily, 2003:4). This paper primarily focuses on privately held land. A third and related dimension to be taken into consideration concerns the difference between irrigated and rainfed land, and their relative proportion in household and village economies.
5The first theme addressed by this paper relates to the effects of water, rather than land, scarcity, and examines conflict over resources between villages, driven by spatial inequalities inherent to a large, plain-based irrigation system overlaid with patterns of settlement based on social identity and driven by political motives. The second theme draws on contrasts between plain and mountain villages to examine the significance of land inequalities within villages and the effects of these on elite behaviour and power relations, and the way in which this might relate to the provision of public goods. The third theme concerns a brief examination of the role of land in contrasting household economies over the last decade, given the high degree of functional landlessness in rural Afghanistan, and the implications this might have for the future. This leads to a final discussion of the links between land, village contexts and social welfare. All three themes draw on primary data from two long-term studies relating respectively to opium poppy cultivation (Pain, 2007) and to household livelihood trajectories (Kantor and Pain, 2011).
6Conflict over access to water has been widely reported in Afghanistan at multiple levels: at a system or river basin level, as in the Hari Rud system of Herat (Chokkakula, 2009), and between and within districts, as in Faryab Province (Pain, 2004). Underlying all of these water-access conflicts is the breakdown of customary practices of water distribution driven by water scarcity and wider conflicts.
7It is in Balkh Province and in the Balkh Ab River and Hazhda Nahr canal network – an irrigation system that dates back more than a thousand years – that the incremental breakdown of historical practices of water management can be traced in detail. The following account points to both the deeper, underlying structures and causes of social inequality, and to the more recent drivers which have brought these inequalities to play in the conflict over water. These drivers have included shrinking farm sizes, which have in turn driven the need for agricultural intensification to support a farm-based economy and an increased demand for water to support that intensification, occurring at a time of increasing water scarcity, particularly from the 1990s onwards. Over the last three decades rising conflict based on identity and location has fuelled the use of water control as an instrument of power.
- 3 I am grateful to Haji Sayed Gawhary for detailed discussions in relation to the historical dimensio (...)
8An understanding of this conflict needs to draw on the history of the area and the complex patterns of its settlement since the end of the nineteenth century (Tapper, 1973; Tapper, 1991).3 In the late nineteenth century, the Turkman plains had become depopulated with the collapse of the northern Khanates and the rise to power of Abdur Rahman, a Pashtun from the Durrani tribal confederation. As a key part of a political strategy to gain and maintain political control of the north, a systematic process of settlement of the north by mainly Pashtun people from the south was implemented. During the first part of the twentieth century, settlers were primarily Durrani Pashtuns from the west and the south (Kandahar) and were mostly of nomadic origin. From the mid-1930s, under the influence of Minster Muhammad Gul Khan Mohmand, the assertion of Durrani dominance was balanced by the systematic settlement (particularly in Aqcha and Balkh) of Pashtuns of Ghilzai origin from eastern Afghanistan and this process of settlement has continued since that time. In addition, Hazara people were also forcibly resettled. As Tapper notes (1991) and evidence from the field confirms (Pain, 2007), settlement by the Durrani Pashtuns in particular often (but not always) took place in the most favourable locations, primarily upstream from the existing settlements of Uzbek, Aymak, Arab and Turkman people.
9Up to the 1950s, population levels were relatively low and land was abundant, so that competition for land and water did not represent a cause of conflict. Tapper makes clear that the settlers from the south were quick to gain political and economic control over existing populations, and local authorities tacitly supported a central government policy of political and cultural discrimination against non-Pashtuns (Tapper, 1991:34). While the settlement history is different for each of the irrigation canals of the system, in two case-study canals in Charbolaq District (Pain, 2007), the upper reaches of the canals have been settled almost exclusively by people of Pashtun origin, although each village traces its origins to different parts of the south. As one travels downstream, the proportion of non-Pashtun villages rises, particularly at the bottom end of the irrigation system. In Chimtal District on the Imam Sahib Canal, there are more Pashtun villages downstream, but these are largely of Gilzai origin.
- 4 A paikal is a relational measure of water flow or volume to area, and is approximately 400 jeribs o (...)
10Through a network of eighteen canals, the Balkh irrigation system historically provided water to four districts in Balkh, as well as to Jawzjan Province. It has two distinct sections. The first is upstream and located within the Sholgar Valley (predominantly settled by people of Pashtun origin), where water from seven canals was customarily allocated to irrigate 5 000 ha (Asian Development Bank, 2004). The downstream section in the Hazdha Nahr irrigation network fed an estimated 424 880 ha of irrigated land through a system of 11 separate canals with a total length of 475 km across Mazaar, Balkh, Aqcha and Jawzjan. The principle of the water allocation system was historically linked to an agricultural taxation system and based on water allocation rights expressed in a unit called a paikal.4 Water distribution worked through a traditional system with a water bailiff (mirab bashi), assisted by mirabs who were responsible for the major secondary canal intakes.
11The creation of new settlements in an irrigation system has implications for the management of water distribution between upstream and downstream villages. At a time of water and land surplus, with an incremental process of settlement over a long period of time, it is unlikely that consideration of the implication of settlement for water distribution were taken into account. The primary concern of government was to develop a system of taxation around water distribution.
- 5 One kabuli ser is 7 kilograms, and 1 jerib is one-fifth of a hectare.
- 6 A bawra is the proportion of land that can be irrigated each year; in the case of 1 bawra, all land (...)
12Based on detailed discussions with the mirab bashis of three districts who have had a long history of engagement in water distribution, it is evident that structural inequalities are built into the irrigation system and water distribution within the canals. Downstream villages have not been allocated longer irrigation times, for example to compensate for the effects of reduced water flow downstream. The existence of such inequalities is supported by consistent reports of higher yields for crops upstream, compared to those downstream. For example, one informant commented that villages at the top end of the irrigation system in Charbolaq could easily produce 40 sers per jerib5 of wheat, while those at the bottom would only achieve 18–20 sers per jerib in a good year. In this informant’s view, a wealthy landowner at the top end of the irrigation system would have between 40–50 jeribs of land, while a wealthy landowner at the bottom of the system would have to own at least 100 jeribs, if not more. These inequalities are officially recognised and formed the basis of taxation classes introduced during King Amanullah’s rule in the 1920s. Thus, land at the top end of the canals, which is termed 1 or 2 bawra,6 has been taxed at the rate of 5 Afs per jerib, and land that is 3 or 4 bawra at the bottom end has been taxed at the rate of 2 Afs per jerib.
13Under the previous conditions of water sufficiency, water distribution inequalities appear not to have constituted a point of friction between upstream and downstream villages. Villages at the bottom of canals such as Chimtal’s Imam Sahib Canal, reported that there was sufficient water for them to grow wheat during the winter period to meet household requirements and a cash crop of either cotton or sesame during the summer to meet cash requirements. For Turkman villages at the bottom of the Charbolaq canals, large livestock holdings combined with a secure winter wheat crop provided a sufficient economy.
14Since the 1950s, there have been two interrelated processes at play which have increasingly put pressure on the irrigation system. The first has been the steady growth of the rural population, declining farm size and increasing intensification of agriculture as household have sought to intensify production to respond to the reduced land areas per household. One eight-year-old informant recalled that his father and two brothers together had cultivated 1 000 jeribs of land. He reported that eighty households now utilized the land. The second change has been the increased demand for water, driven by processes of intensification and increasing urban demand from Mazaar city, particularly since 2001. Since the drought of 1998, river flows and water supplies have also been in decline.
15Moreover, the growing imbalance between demand and supply of water has been exacerbated by a breakdown in the customary practices governing the water distribution system. During the 1990s, farmers in the upstream, largely Pashtun Sholgora District illegitimately built a further five canals to support an expansion of rice cultivation. Since that time these farmers have been extracting a greater proportion of irrigation water than allowed for by the original, traditional agreement. A pattern of upstream farmers taking more than their customary allowance has been repeated at each point of water extraction down the irrigation system. One direct measurement of water outtake in 2003 found that three upstream main canals with an allowable allocation of 17.6 % of the water were taking nearly 35.6 % of the water flow, almost double their allocation (Asian Development Bank, 2004: 30). The top seven canals of the Hazdha Nahr system were estimated to be taking ‘over half of the water out of the river, while having only the right to a quarter’ (Asian Development Bank, 2004:30).
- 7 As Alden Wily notes, there is no robust data on landlessness (Alden Wily, 2003:18). National values (...)
16The implications of this pattern of water over-extraction can be traced through the study of water distribution along the irrigation canals flowing into two districts of Balkh (Pain, 2006). The Imam Sahib Canal in the Chimtal District has 20 main villages along its length and is divided into an upper, middle and bottom section. The elders from one village at the head of the district system reported that in the past the village had supported a cropping system based on intensive double cropping of wheat, barley, peas, maize and cotton. An estimated 80 % of households were landless (high levels of landlessness is a theme that runs through these accounts and is discussed below),7 but the intensive cropping system combined with landholdings had provided sufficient production and employment (through sharecropping arrangements or farm labour) to support households within the village.
17A second village in the middle of the irrigation canal, in which 80 % of households were also landless, had the same intensive double cropping system as the first village. However the third village, situated at the bottom end of the canal, had a 3 bowra system, with a third of the land being irrigated each year. About 50 % of the households were reported to be landless. The cropping system was less intensive than the in the other two villages, with mainly wheat and cotton, along with irrigated lucerne for the village’s significant livestock population.
18All three villages had an account of the cause of the decline in the availability of irrigation water and its consequences for their village’s economy (Pain, 2006: 13–14). For the Tajik Village at the head of the system, the decline in water availability stopped cotton production and confined the village to a single winter crop of wheat. Water shortages were reported to have started at the beginning of the drought in 1998, and the combination of subsequent years of insecurity and drought had led to considerable migration and movement of landless individuals and families out of the village. Explanations for the decline in water availability cited the increasing use of water in the Sholgara District and its expanded area of rice cultivation.
19The Hazara Village in the middle of the system had a more historical view of the reduction in water supplies, the beginning of which predated 1990. In this village, people attributed the decline in the water supply to increasing water demand upstream, due to intensification processes driven by land division through successive generations. The consequences for their village, it was reported, had not only been a reduction from double to single cropping, but the area of single cropping under winter wheat had also been reduced and a fallowing system introduced. There has also been extensive migration of landless people away from the village.
20The Ghilzai Pashtun village at the bottom end of the canal also reported a drastic reduction in irrigation water and a sharp decline in the area used for wheat production. Informants argued that this was more to do with the distribution of water within the District. They referred to powerful people at the office of the woliswal8 and attached significance to the location of that office at the upper end of the irrigation system. They argued that a previous mirab, who had been unable to resist the pressure of powerful people within the District, was the reason for reduced water flows to them. This has led to a reduction in the area of wheat cultivation and the loss of household self-provisioning.
21Similar accounts of inequalities in water distribution were found along the Charbolak Canal in Charbolak District. Field observations and respondent comments support a picture of differential water delivery between the top and bottom ends of the irrigation system. This is in part historical and by design, with villages at the top end (Pashtun communities) reporting economies based on intensive double cropping and horticulture, while those at the bottom end (almost exclusively Turkmen populations) reporting a more livestock-based economy with a single wheat crop. The issue of water scarcity emerged from discussions with elders at the woliswal’s office, with the mirab and with other informants. Not only were irrigation water constraints stated to be severely limiting water availability in all parts of the District, but at the bottom end of the irrigation system the villages had severe problems associated with a lack of drinking water. As on the Imam Sahib Canal, villages upstream saw the cause of the problem as resulting from districts upstream of Charbolaq taking more water than they were allowed. However, the interpretation of the villages in the middle and the bottom of the system is centred more around water distribution within the District, as confirmed by discussions with the mirabs. Thus, one mirab affirmed:
I am not in a position to distribute water easily, it is out of our hands. There are power issues and sensitivities (Pain, 2006:18).
22In the bottom villages, respondents noted that they had not been able to cultivate even a tenth of a bowra, and this was restricted to winter wheat. Significant numbers of men had left the village to find work in Mazaar, or had migrated to Iran.
23The causes and effects of conflict over water are closely related to spatial position on the irrigation system; this in turn often maps closely with the ethnic identity of villages, which relates to settlement history and the political drivers thereof. The power relations that underpin the conflict are most clearly revealed by the dynamics of opium cultivation in these two districts since the 1990s. As argued in detail elsewhere (Pain, 2007), the rise and fall of opium poppy cultivation between 1996 and 2006 in these two districts (in which opium poppy cultivation came to be concentrated in Balkh Province) provides an insight into the underlying power structures of these districts, how these link in to the control of water and, in turn, how these are linked to the informal regulation of markets. An analysis of the opium poppy economy offers a lens to analyse power relations around resources and the links between upstream villages, the opium poppy trade and provincial authorities, represented in the districts by both the woliswal and the chief of police. It also reveals the transformative effects that a dynamic rural economy can have on conflict (Pain, 2012).
24In summary, four distinct phases of opium poppy cultivation can be identified, each with clear spatial dimensions (Pain, 2007). Before 1994, there was limited and localised production of opium by Turkmen people, mainly settled at the bottom end of the irrigation system. This production was for self-consumption. After 1994 and associated with the rise of the Taliban to power, a new phase of commercial cultivation began, linked to new opium varieties developed among Durrani Pashtun villages located upstream in these districts. Through their control of the market, cultivation of the crop did not spread beyond these villages. However, with the fall of the Taliban and the loss by these villages of absolute control of the market, cultivation spread throughout these two districts, even though it remained more intensive upstream, in the double cropped areas. In the case of the Imam Sahib Canal in Chimtal District, cultivation took place right along the canal, with the water-scarce bottom villages receiving sufficient water to allow for an opium crop, bringing them food security. In contrast, at the bottom end of the Charbolaq Canal, absolute water scarcity continued and opium cultivation only spread to the villages in the middle of the Canal. However, the expansion of the opium poppy crop brought with it an enormous demand for labour which was mobilised across village divides and consequently contributed to a reduction in the level of conflict between villages. Access to the main traders nevertheless continued to be regulated by social identity. The rise in prosperity of the rural economy during this period was notable, and it mitigated many of the tensions over water distribution, although unequal access to water and the role of informal power in regulating its distribution were not significantly moderated.
25However, in 2006 coercive power deployed informally by the provincial authorities (in a manner similar to that exerted by the Taliban when they banned the cultivation of opium in 2001) abruptly brought opium poppy cultivation to a halt in the Province (Pain, 2007), leading to a rapid collapse in the health of the rural economy and the re-emergence of conflict over water, which remains unresolved to this date. With the rise in insecurity in Afghanistan and an associated rise in the price of opium, cultivation of opium poppy had re-emerged in the core villages of Chimtal by 2011 (Pain, 2011).
26This account of conflicts between villages over water brings out the spatial dimensions of resource conflict linked to history, settlement patterns and social identity. It focuses on differences between villages within a given resource environment. However, contrasts between villages based on resource distribution within villages can also be examined for what they reveal about the nature of power and its links to land and conflict within villages.
27The second body of evidence on which this paper draws is derived from a study of livelihood trajectories of a group of study villages and households between 2003 and 2009. An initial study (Grace and Pain, 2004) monitored the livelihoods of 390 households across 21 villages in seven districts in seven provinces of Afghanistan, seeking to build an understanding of what people in rural areas do and how this varies by context. In 2008, a re-study of a sub-sample (64 households in 8 villages in three provinces) investigated household trajectories with a focus on how household circumstances had changed and what factors had brought about these changes (Kantor and Pain, 2011). The discussion here draws on evidence from the three villages in Badakhshan and two villages in Kandahar, so as to contrast villages of the mountains with those of the plains – a central point of diversity for this paper.
28The two study villages in Kandahar (Pain, 2010b) provide evidence of major inequalities in land holdings similar to those of the Balkh villages discussed above, underlining the fact that the irrigated plains contain significant populations of landless households. Both are characterised (if in different ways) by the presence of a strong social hierarchy. Those at the apex of this hierarchy own most of the land and have diversified into non-farm income sources and salaried urban employment through powerful political connections, including the late Provincial Governor, Ahmad Wali Karzai, and the President Karzai. In the most extreme case, Lalakai village in Dand District close to Kandahar city, a proximity that has largely ensured security, the village lands are owned by just three landlords. As observed by a group of village elders:
In the villages of Dand you will not find such a village where the whole land belongs to three families …in other villages where it is the malik system but in our village the main landowner is the landlord and head of shura (Pain:2010b:10).
29The main landowner, a Kandahari Pashtun, owns half of the village (500 jeribs). He reportedly did not inherit the land but bought it during the 1990s from Kandahari Hindu traders. Land is one source of his power. A second source flows from the fact that all of this land is cultivated by sharecroppers, who remain in a dependent position in relation to him, many being immigrants to the village, including a group of Jats (by origin pastoralists from Pakistan), and not of the same tribe as him with the obligations that that might carry. A third source is his strong external connections, ranging from key members of the Provincial Council all the way up to the top of Government in Kabul. The evidence demonstrates clearly that this powerful landowner runs the village for his own benefit (Pain, 2010b:10-11).
30The second village, Julan in Dand District, has slightly less-concentrated land ownership, but well over 80 % of households are landless. Much of the estimated 2 000–3 000 jeribs of irrigated land in the village was apparently previously owned by Hindus, but came under the control of key landlords during the Mujahidin period when many Hindus left. Several informants reported that these lands were taken illegally and the deeds of transfer and land ownership forged. The land is now largely owned by the six maliks of the village, with one head malik (whose father was a malik before him) and his assistant, along with four sub-maliks, being responsible for the remaining parts of the village. According to informants, the village shura is made up primarily of these six maliks and four other villagers, but the decision-making and power lies firmly with the former group.
31The head malik in Julan is ruthless in his dealings with others, including even with his own relatives. Three of the case-study households were related to him and each had an account of ways in which he had used his power to seize assets that belonged to them. In one case, he reportedly took over a shop that had been left to his nephews; in another, he was reported to have taken land from the widow of a deceased brother. In a third case he simply ignored the needs of a poor young relative struggling to survive.
32The village elite are well connected with provincial power holders and have diversified, as with the landlord of Lalakai village, into non-farm income sources and salaried employment for family members. Other households in the village remain locked either in positions of dependent security through insecure sharecropping arrangements, or have moved to find income sources in the margins of Kandahar’s urban economy.
33What these two villages have in common is the way in which the landed village elite acquired land during the conflict period from departing Hindu land owners, most probably through seizure. Moreover, in both cases the elite have enriched themselves through their strong connections to the political elite of the Province and through connections to the higher reaches of Kandahar’s booming reconstruction economy, as well as through gaining salaried employment for their close family members. Public good provision in both villages is minimal. The choice available to the landless has been either to remain under conditions of uncertain security and dependency on powerful landlords, or to secure an income from non-farm sources outside the village and at the margins of Kandahar’s urban economy. In the case of those still tied to landlords by patron-client relations, they had either simply maintained their household economy or it had declined.
34The three study Badakhshan villages (Pain, 2010a), located between 1 200 and 2 000 metres above sea level, provide a contrast to those of Kandahar. Moreover, there is greater variability between these three mountain villages. All three benefited from the opium economy from 2000 to 2006, although the extent of this benefit varied according to the land available to the village. All three have subsequently suffered serious economic decline. It is the differences between the villages in terms of their social order and the behaviour of the social elite that provide the key points of contrast.
35Of these three villages, the lowest, Toghloq, is the most similar to the Kandahar villages, having relatively high levels of land inequalities and resource wealth, although the degree of land inequality and the extent of irrigation still do not compare with the two Kandahar villages. In the past, Toghloq was self-sufficient in grain and there was sufficient land to provide sharecropping and labouring opportunities for the landless. Prior to 1978, the village had a school, but it produced few graduates and no one from the village went to university or into government service. During the war with the Soviets, the village was the site of intense resistance to the Communist Government, and there was considerable conflict. The Mujahiddin destroyed the school and killed teachers in the valley, stopping education until after 2001. There are accounts of old customary structures existing prior to 1978, but with the onset of the conflict these customary structures appear to have been overtaken by the rise of a new village elite based on fighting capacity. While this might have provided a degree of protection against external forces, there are accounts of insecurity within the village. One informant described how a commander had married his daughter by force.
36With reasonable irrigation resources the village was less affected by drought than other Badakhshani villages, and it profited considerably from opium cultivation. Attempts to destroy the opium crop in 2005 were met with armed resistance, which would have been unlikely in the other two Badakhshani villages. However, the decline of opium has badly affected the village economy and an increasing number of men are leaving the village to join the army or the police force, often gaining this employment through former commanders living in Kabul. While farming still supports some households, for many, this is no longer possible. Although a school has been re-established in the village, an NGO working locally reported that the village was one of the more difficult locations in which to operate due to opposition from key power-holders. A Community Development Committee (CDC) was formed, but internal disputes led to its collapse. A new committee has been set up, but it does not appear to be functioning well. Because of its geographical position and resources, as well as its relatively powerful elite, this village has enjoyed a degree of physical security and independence from the outside world. Within the village, however, the provision of public goods, including education, has been limited and the village elite have acted in a largely self-interested manner.
37The second village, Khilar, is a marginal village and the smallest of the three. It is located on a small plateau up a narrow valley that runs off the main river and has had to establish and maintain external protective relations in order to ensure its physical security. In part, this is the result of its being small and poor. It is also partly the result of its being an Ishmaili village. During the first phase of the war, this village was subjected to considerable hostility on account of its religious-minority status. Violence was used against households and land was taken by force by powerful people in the neighbouring valleys. The village was eventually able to establish dependent relations with a key valley commander, giving it some measure of physical security, but the drought led to many households falling into debt and many had to mortgage and ultimately sell their land – about a third of the village’s irrigated land now belongs to non-residents. While the opium business offered some a means of economic recovery, villagers largely worked as labourers on other people’s land.
38Since 2001, the village has maintained dependent relations with the valley commander mentioned above. This commander’s authority is greater than that of the village shura, and he is responsible for dispute resolution within the village. His relations with key political figures in the District’s main valley are uneasy, which means that the village has limited political influence. Many villagers are related to each other, so internal security and social support is strong. However, the village’s small size and dependence on external support render problematic and difficult its access to wider resources. It only recently obtained road access and, while more children are now attending school in the valley, because of the distances involved many girls do not attend beyond grade 8.
39The third village, Shur Gul, is both the largest and the furthest from Jurm, being located in a narrow plain in the Kokcha Valley. Land inequalities are relatively low and only about 30 % of households are landless. However, few households are grain self-sufficient. During the 1950s, its location close to the Badakhshan lapis lazuli mines resulted in the village arbob coming into contact with educated government officials, prompting the village elite to establish a school. By 1978, a significant number of men had graduated from tertiary education and established wider contacts. During the period of conflict from 1978, many returned and supported the village, keeping the school open, managing security and, through wider social connections, gaining access to economic resources in times of need. In the 1990s, for example, the lapis lazuli mine provided employment for village households during the drought. In the 1990s, a girls’ school was also established. Although the village derived some benefits from opium, by reason of its altitude and limited land it did not benefit to the same extent as Toqhloq. After 2001, the educated elite have been active in bringing NGOs to the village, expanding the provision of education, improving roads and providing clean water and electricity through a micro-hydro scheme. With the decline of agriculture after 2006, the village has suffered economically and there has been a rise in outgoing migration to other provinces and to Iran. Poorer households have also begun to ration consumption. However, even these poorer households are trying to keep children in school, efforts which are helped by food for education support.
40The evidence presented above points to significant differences between villages in terms of their behaviour in relation to the provision of village public goods.
41The most fundamental public good that a village can provide is that of security against an insecure outer world. This can be achieved through military means, as was the case in Toghloq in the past. But it can also be achieved through building strong external connections with more powerful or influential people, as has been possible in Kandahar and in one of the Badakhshan villages –Shur Gul, which managed external security extremely effectively during the Mujahiddin period and after (Pain, 2010a: 16–19). Alternatively, village security can be achieved, as in the case of Khilar, through obtaining protectorate status, when the village does not have the resources for its own auto-defence.
42However, external security does not mean internal security – the villages of Kandahar, and Toghloq, provide examples of villages which lack internal security and in which powerful people are abusing their power.
43Beyond security, villages can also provide other public goods. Welfare constitutes one of these, and the existence of charity and a moral economy (Scott, 1976) represent one dimension of this. A second public good is dispute resolution, the existence of which is supported by a wide body of evidence (see, for example, Gang, 2010). A third level of public good provision relates to investments in education and health, as is found in Shur Gul, both through village efforts but also through building connections to external sources of provision, primarily including NGOs.
44The case-study villages discussed above demonstrate that the external relationships a village can build with more powerful figures at the district and provincial levels are critical. Villages are not autonomous, but should be seen as part of a wider web of relationships and connections. Some are better at building these relationships and connections through the connections of their elites, a process that is facilitated by a degree of political stability in the region. Where such stability does not exist, this process becomes more difficult, and the critical question then becomes when, where and why is it in the interests of village elites to prioritise the interests of the village over their own individual interests. Why, for example, do the Kandahar elite act in one way and the elite of Shur Gul behave in a completely different manner?
45The argument has been made that villages have distinctive social orders that underlie the ways in which they behave. There is no wider, systematic evidence of the extent to which this might be true, although a wide range of observational evidence (see Pain and Kantor, 2010: 32–33) supports the view that this variability is not unique to the study villages. Many NGO field staff talk of how villages differ with respect to the ease or difficulty they have in working with them. There is evidence of the capacity of villages to provide dispute resolution and other public goods through customary structures (Brick, 2008; Smith, 2009). But what is the connection between elite behaviour, the degree of public goods provision in the village and the nature and quality of social relationships that can be drawn upon to provide security?
46The available evidence points to the fact that the resources available to the village, the distribution of their ownership and the presence and performance of village customary institutions need to be seen as interrelated dimensions of the village, rather than separate, independent aspects and factors. In the example of Shur Gul, the village’s limited irrigated area, minimal land inequalities, grain-deficit economy and relatively flat social hierarchy can be identified as conditions likely to favour, or even to maximise, efforts to secure the village’s provision of public goods, as well as being related to high-quality social relationships on which people can draw. Moreover, this remains the case regardless of the fact that the level of public goods provision is limited by the overall poverty of the village. In contrast, the Kandahar villages, with their wealth of resources and major land and social inequalities, have the lowest provision of public goods and the poorest quality of social relationships on which people can draw, despite the fact that, for those who can access reciprocal relations, the relative prosperity of the area produces opportunities for them to prosper.
47Drawing on Brick (2008), it appears that the key factor is the way in which village customary structures function. Three key customary structures can be found in most Afghan villages – the village council, the village leaders (maliks or arbobs) and the village clerics (mullahs). When these structures work well, they can produce conditions that are favourable to the provision of public goods. Four key features of the organisation and behaviour of these customary structures can drive such public goods provision when they are present and strong. The first feature is present when each of the three customary structures has distinct and non-overlapping areas of authority, and their authority and legitimacy is derived from different sources. In the case of the village council, membership is based on reputation, performance and acceptance by village households, and this structure manages dispute resolution. The village headman is the key interlocutor between the village and Government, representing the village. The cleric’s authority is derived from religion and he speaks on matters determined by Sharia law. The evidence from the case-study households and villages discussed above is consistent with this pattern. The evidence also confirms the second key feature of these customary structures – their capacity to independently raise revenue from within the village. For example, the school in Shur Gul was financed by a levy instigated by the shura. There were also several examples of payments being made to the mullahs.
48The third feature is constituted by the checks and balances that can prevent members of the three customary structures abusing their authority. Even in Kandahar, clerics were able to impose constraints on the behaviour of the elite in relation to matters determined by Sharia law (Pain, 2010b: 96). However, the ways in which these checks and balances function can be variable, so that they might be effective in constraining the abuse of authority in Badakhshan but not in Kandahar. The reason for this variability relates to the fourth factor – the need for there to be sufficient actors who have the ability to stop potential abuses of power and act as veto players. Where land distribution is relatively equal there are more landowners who have an interest and power is consequently dispersed. In contrast, where land ownership is concentrated, so is power. Under such conditions there are few constraints on elite behaviour and powerful individuals are more likely to act in their own interests.
49It is the relation between social solidarity and inequality that is likely to be a critical issue. Where land inequalities are high and the elite are economically secure, they have few incentives to expand the provision of public goods and are largely immune to social sanctions, although ties created by social identity, which were absent in Lalakai Village, might challenge such power. Where the elite are economically insecure, they are likely to have a shared interest in supporting social solidarity and promoting the provision of public goods. But what does this mean for the role of land in the economic life of households?
50There is no doubt that for most rural male Afghans the idea of being a farmer is central to his identity. However, as the data gathered from the three Chimtal villages makes clear, a majority of households in these villages are functionally landless and a farm-based economy for these households requires working either as a labourer or as a sharecropper. Furthermore, when the health of the rural economy declines, as it did after the decline of opium cultivation in Balkh or in dry years, there is no work, resulting in labour migration outflows to urban centres in Afghanistan and beyond. The question consequently remains: how central is land to the economic life of rural households?
51The key findings from one study (Kantor and Pain, 2011) point to a long-term decline in the health of the rural economy across the study sites, as well as to the limits of land in providing even a portion of household food requirements, either directly from on-farm production or indirectly through labouring on or sharecropping land. The contrasts between the three provincial settings are illuminating. Of the 64 households (16 in Kandahar, 24 in Badakhshan and 24 in Sar-i-Pul) considered, only 13 had prospered since 2003.
52Ten of these 13 households were located in Kandahar and none of them had prospered directly through agriculture. Two of these ten households had leveraged existing assets of land and external social connections to powerful people, so as to diversify into multiple income streams, including land, off-farm work and non-farm sources such as overseas trade or employment with key provincial or national figures. A further four had used social connections to move into the urban economy. The four without land in Kandahar had either found a niche on the margins of Kandahar’s urban economy or through an adequate supply of male labour in their joint households were able to command sufficient labour to improve their circumstances.
53It was through the deployment of labour resources that the only example of a household prospering through agriculture was found, and this occurred in one of the Badakshani villages. The improvement in the circumstances of this case-study household had been built on the growth of the opium economy and the use of the resulting profits to acquire land. The other two prospering households in Badakhshan had achieved that result primarily through salaried employment, in one case, as a teacher, and in the other, as a driver for an NGO.
54Of the remaining 51 case-study households, six had managed to maintain their economic position and 45 had seen their livelihood security decline. For the declining households in Badakhshan and Sar-i-Pul, resource constraints – both of land and labour – and household-specific events had contributed to the erosion of their well being. This had taken place in a broader context of decline in the village, district and provincial economies. These households clearly demonstrate the limits of what an agrarian economy in a resource-poor and market-distant area has been able to provide since 2002.
55This paper has examined three dimensions of the dynamic links between power and land relations under conditions of conflict in Afghanistan: access to water under conditions of increasing scarcity, structural contrasts between mountain and plain villages and the declining economic role of land in rural livelihoods. These are not the only sources of variability, but each in turn provides a degree of understanding of some of the factors that might underlie such variability. Each dimension discussed points to some of the structural inequalities that underlie Afghanistan’s rural economy in terms of access to resources. On the other hand, all three dimensions also point to various factors and dynamics that can mitigate potential conflict over resources: the rise of the opium economy in Balkh, for example, or collective action at the village level to promote the provision of public goods. However, they all also point to the need for historically-based and contextualized empirical studies to ground our understanding of land and conflict in Afghanistan.