Heterogeneity is not necessarily bad
Back to contents - Previous file - Next file
There are three main sources of heterogeneity which bear upon the capability or the motivation of resource users to participate in collective action. First, heterogeneity may result from ethnic, race, or other kinds of cultural divisions. Second, it may arise from differences in the nature of the interests various individuals may have in a particular collective action. Third, it may originate in inter-individual variations in some critical endowments, that are reflected in varying intensities of interest. While the first two causes of heterogeneity are a strong impediment (especially the second one) to collective action, the same cannot be said of the third cause. The discussion below substantiates this proposition.
Heterogeneity of interests or objectives is particularly threatening for collective action aimed at resource management when a fraction of the resource users have alternative income-earning opportunities and, more seriously still, when they also reside outside the area where the resource is located. In Mali, for example, the predominance of absentee herdownership which resulted from the great Sahelian droughts in the 1970s (when pastoralists were forced to sell their livestock to farmers or, more generally, to town-dwellers like traders and civil servants) appears to be a major stumblingblock on the way towards pastoral institution-building for sustainable rangeland management. According to a recent evaluation study of the Mopti Area Development Project:
Absentee herd owners favour open access rangelands so that their herds can graze anywhere. They may even use their political influence to prevent pastoral associations receiving legally defensible land rights. Their herders, on the other hand, working for a wage, are fully engaged in herding. The pressure of work, their low social status, and the fact that the animals they are herding are not their own demotivate them from becoming involved in natural resource management. (Shanmugaratnam et al., 1992: 20)
Similar difficulties are encountered in Mauritania where, although recently authorized by law to deny outsiders access to their demarcated rangelands when pasture is scarce and overgrazed, pastoral associations seem to continue to apply Islamic law and to follow local customs, implying that the stock of absentee herd-owners still graze on their commons (Shanmugaratnam et al., 1992: 25). Given that the extent of herd dispossession is considerable (like in Mali), the problems created by absentee ownership for pastoral management are not to be underestimated:
Both Moor and Peul pastoralists have lost large shares of their livestock through sale to absentee herd owners.... The abolition of animal taxes since the droughts of 1984-85 and the introduction of property taxes have given further impetus to absentee ownership. Many richer people in the towns prefer to invest in cattle instead of in real estate. Absentee herd owners in Mauritania are mainly concentrated in Nouakchott and a few other towns. They prefer to keep their herds as close as possible to their towns; for instance, the bulk of the herds belonging to owners living in Nouakchott are found within a radius of 100 kilometers around the city. Such concentration of herds in limited areas contributes to over-grazing and exacerbates land-use conflicts around the towns. (Shanmugaratnam et al., 1992: 25-6)
Note in passing that the above description also fits well with other country experiences in SubSaharan Africa, for instance in Burundi (many cattle herds grazing around the capital city of Bujumbura belong to city-dwellers).
Marine fishing is another sector where resource management is seriously undermined by the existence of different interests, most notably between small-scale (artisanal) fishermen and industrial fishing companies. Contrary to the former, the latter have indeed many exit possibilities, not the least because they can easily move their fleets to other, possibly distant fishing-grounds or even switch to other economic sectors if fish resources have become too degraded to make their investments in fisheries economically attractive. An immediate consequence of this situation is that industrial operators feel much less concerned about conservation of fish resources in a given harvesting ground, or perhaps even in general, than artisanal fishermen whose subsistence crucially depends upon the state of these resources owing to a lack of alternative income opportunities.
A vivid illustration of this contradiction of interests is the conflict which arose during recent decades around mackerel fishing between hook-and-line artisanal fishermen and industrial seiners in the coastal waters of Katsuura in Japan (Chiba prefecture). While the former have been taught from their childhood that catching young fish is a bad thing that jeopardizes the future of the resource (a fisherman found guilty of such an act gets a bad name), the latter have systematically caught all the mackerel regardless of their size. This indiscriminate fishing on the part of industrial companies eventually caused the depletion of mackerel stocks not only in Chiba prefecture but in the whole coastal area where these fish migrate (from Izu islands south of Tokyo to the offshore of Hokkaido Island, northern Japan) resulting in tremendous hardship for small-scale fishermen. The same story has been actually repeated for sardines and tuna and, today, it seems to hold true for skipjacks as well (oral communication of Masao Suzuki and Yoshiaki Aziki at the 10th Anniversary Conference of the International Collective for Support of Fishworkers (ICSF), Cebu, Philippines, 2-7 June 1994). Unfortunately, examples such as these are very easy to find all over the developing world.
Conflicts between men and women regarding the use of village commons can also be regarded as conflicts of interests. For instance, in many parts of Africa, men want to clear forests to open new fields for cultivation while women want to preserve them as a permanent source of firewood (Freudenberger and Mathieu, 1993: 14). Furthermore, socio-economic differentiation can easily lead to collective action problems if the pressure of survival constraints among the poor translates into high discount rates of future incomes while the rich have low discount rates (see above the crisis situation in Kottapalle mentioned by Wade).
It must be stressed that the difficulties associated with heterogeneity of interests are often compounded by the problem of cultural heterogeneity. An example which immediately springs to mind is the well-documented failures of herders and agriculturalists to reach joint agreements about how to regulate the use of common lands (see, for example, Bassett, 1993 about the conflicts between the Senufo peasants and the Fulani herders in Côte d'Ivoire). Another illustration is provided by the conflicts existing in many parts of the world between fishermen and agriculturalists or aquaculturists when the latter encroach upon mangrove areas that are so crucial for prawn reproduction.
When heterogeneity of interests is present, a negotiated agreement is clearly required to the effect that each interest group enjoys well-delimited rights and faces well-defined constraints. This often implies that the physical domain of the CPR is subdivided into several sections corresponding to the different interest groups. Thus, for example, in Valença (Brazil), following the introduction of new harvesting techniques which created heterogeneous interests and consequently gave rise to bitter conflicts (involving many acts of physical violence), the local fishermen 'divided their estuary into various areas and assigned a different technology to each area' (Cordell, 1972, cited from Schlager, 1994: 248). More precisely:
The mangrove fence and barricade net are always located highest on the shore, succeeded by the dragged nets, encircling nets, and tidal flat fish corrals. Finally, moving out to the channel are positioned the fish traps, trotlines and gillnets. In any case, the distribution of techniques in a wedge of water is always such that they do not overlap. (Cordell, 1972: 42, cited from Schlager, 1994: 249)
Unfortunately, there is no guarantee that different interest groups will be actually able to reach an agreement. Just to quote one example, in the Bay of Izmir (Turkey), there are many co-operatives yet each co-operative seems to represent the narrow interest of a particular gear type competing for access to the resource. None is apparently able 'to tackle the larger problems of managing a crowded fishery in the proximity of a large urban centre, or protecting the collective interests of fishermen against resource degradation' (Berkes, 1986: 223).
It is equally possible to think of situations in which people disadvantaged with respect to some endowments are unlikely to contribute to resource regulation. This happens when they are excluded from its benefits because they do not have access to some key factor. (Note incidentally that, in such circumstances, heterogeneity in endowments is tantamount to an heterogeneity of interest.) For instance, behind the failure of a grazing association established in Lesotho within the framework of a pilot range-management project apparently lay a considerable diversity of endowments in livestock production and livestock management practices in the local population (Lawry, 1989b: 10-11). In particular, smallholders and households short on herding labour and management skills (households headed by women or by men absent for work, mainly in South Africa) were less able to adopt many of the more intensive rangemanagement and livestock-production practices promoted by the project's management plan. Furthermore, households which did not own cattle posts in the mountain pastures sought to avoid incurring the cost of gaining access to the posts belonging to others by keeping their livestock in the village year-round, in clear violation of the plan.
The idea that economic inequality is not necessarily an obstacle to resource regulation emerges from the already quoted study by Wade. In the South Indian villages he studied, there are indeed marked inter-household differences in landholdings and wealth. Yet, big landowners who also form the village's elite are not powerful enough to control all the land located near the irrigation channel. Holdings are typically scattered about the village area in small parcels, partly to diversify production risks and partly as a result of inheritance practices. Landowners, whether big or small, may have a plot close to one irrigation outlet and another plot close to the tail-end of a block fed from another outlet. For this reason, all landowners have a common interest in establishing and enforcing a system regulating access to water (Wade, 1988a: 185). In the words of Kanbur, fragmentation of holdings vis-à-vis the irrigation channel under conditions of inequality of landholding has caused 'a "homogenization" of the returns from cooperation when compared to the non-co-operative outcome' (Kanbur, 1992: 20).
In the Zanjera irrigation communities of the Philippines, to take another example, land along a lateral canal is divided into several blocks or sections perpendicular to the source of water so that blocks are at different distances from this source. Each of the blocks is further divided into several parcels and each share in the irrigation system corresponds to one parcel in each block, so that each shareholder has to cultivate parcels of various distances from the water source. As a result, all the farmers have some land in the most advantageous location near the head of the system, and some near the tail. This arrangement has the effect of equalizing income-earning opportunities and providing insurance to the farmers. It has also the important advantage of motivating them to deliver water throughout the entire watercourse and 'to contribute many hours of physically exhausting labor in times of emergency, when control structures have been washed out, and for routine maintenance' (Ostrom, 1990: 83).
A similar system has been found operating in Sri Lanka and is known as the Bethna system. According to Quiggin, the rice areas are divided into a number of sections at varying distances from the dam, and each of the villagers is entitled to cultivate a portion of each section proportionate to their overall landholding (Quiggin, 1993: 1133-4). A positive aspect of this system where total landholdings are divided into a number of separate parts is that in periods of water shortage participants in the water-sharing scheme may decide to restrict the area irrigated to the sections nearest to the dam. In this way, output per unit of water is higher than if water were spread over the entire field (ibid.).
It is interesting to note that the scattering of landholdings which 'may arise from a desire to reduce intragroup conflict' also characterized European agriculture in early medieval times, as attested by the open-field system found in thirteenth-century England (Homans, 1970: 90; see also Quiggin, 1988).
When economic inequality does not thus prevent uniformity of interest in a collective agreement, it can even be a favourable factor if the rich can assume a leadership role. Such a situation in which economic inequality generates the necessary incentives for playing a leadership role has been apparently encountered by Wade in the above study. The reasoning is as follows: since benefits of unified action are positively related to land area, the elite have a disproportionately great interest in the effective regulation of water resources and this helps ensure that the required corporate organization is started and effectively run. In the words of Wade, 'the effectiveness of a council depends on its councillors all having a substantial private interest in seeing that it works, and that interest is greater the larger a person's landholding (provided holdings are in scattered parcels)' (Wade, 1987: 230). The claims that big landowners can make 'are sufficiently large for some of them to be motivated to pay a major share of the organisational costs' (Wade, 1988a: 190). To tie up this observation with our discussion in Chapter 5, it must be noted that the game of irrigation implicit in Wade's example typically raises a co-ordination problem and, given their high stake in proper co-ordination, the elite have a particular interest in initiating it.
The same explanation underlies the story of the success of a village grazing scheme in Lesotho. In Ha Nchele, a lowland village in this country, rotational grazing has been introduced on village-grazing lands as an alternative to taking animals to a cattle post in the mountains. According to Sharp (1987), success was obtained in spite of very little external assistance partly because the village chief held the greatest number of livestock, thus enhancing his personal interest in the project. Of course, an important feature of the whole experiment is that he did not abuse his power as chief (cited from Swallow and Bromley, 1994b: 5).
Besides their leadership role, the elite may provide the authority structure that is required for proper enforcement of regulatory rules. Wade has clearly this latter function in mind when he writes: 'Many who might be tempted to free ride are socially subordinate to others in the user group, and are checked from doing so by sanctions which derive from the wider order of caste and property without the [irrigation] council having to use its own authority' (Wade, 1988a: 193). The same basic position is asserted elsewhere, even in stronger terms: corporate organisations, to be effective, should be based on existing structures of authority. In practice, this means that the council will be dominated by the local elite which is a disturbing conclusion for democrats and egalitarians. But rules made by the majority of villagers would carry little legitimacy in the eyes of the powerful. (Wade, 1987: 230)
The very fact that Japanese rural societies—a model case of successful village-based collective action for CPR management—had a deeply asymmetrical social structure provides a significant confirmation of Wade's thesis. The same can be said of lineage-based societies in Imperial China. The representatives of dominant lineages had a great deal of influence in the selection of the forest manager and in the setting and enforcement of all the rules regulating the access to and the use of the local forests and wildlands. Acting as an elite who controlled a large portion of the land and the benefits it yielded, clans generally applied a combination of appeals to moral norms (Confucian appeals to order, respect for the ancestor, and family loyalty) and coercive enforcement (fines and punishment) (Menzies, 1994: 80-5). Still another illustration is provided by rural Mexico where the Indian caciques (rich Indians acting as patrons) used to mobilize labour and assume leadership for the management of common lands, including important conservation measures such as steep-slope management and erosion control (Garcia-Barrios and Garcia-Barrios, 1990). As Bardhan puts it: 'In many local communities some rudimentary forms of cooperation have been sustained and enforced over the years by traditional authority structures. While there may have been some bit of a sharing ethic, the predominant social norm was often that of an unequal patron-client system, in which the powerful who might enjoy disproportionate benefits from the institution of cooperation enforced the rules of the game and gave leadership to solidaristic efforts' (Bardhan, 1993b: 638). In the Pithuwa irrigation system in Nepal, for example, it so happens that many of the large landowners have their lands located near the tail of the system (which is also near to the east-west highway and thus low-cost transportation of produce to markets). Even though the area was not organized prior to canal construction, thanks to the initiative of some prominent farmers, the whole irrigation project became self-managed through evolution from organization on one branch at the tail of the system to the organization of the entire system (Laitos et al., 1986: 126-7, quoted from Ostrom and Gardner, 1993: 105).
It is interesting to note that, in the literature about co-operatives as in the literature dealing with CPR self-management, the importance of social homogeneity has often been stressed. Here also, however, some dissent begins to be voiced, grounded on solid field experiences. For instance, in a recent report issued by the World Bank, we are aptly reminded that 'the idea that cooperatives can function only if members have the same background can be refuted by the examples of successful ones in both developed and developing nations' (Braverman et al., 1991: 13). Thus, rural co-operatives in the Netherlands were often created by groups 'of influential, better-off farmers who took the initiative to start a service or credit co-op and to contribute the bulk of initial share capital. Smaller producers would join at a later stage, contributing smaller shares' (ibid. 6).
After having stressed the potential positive role of economic inequality for collective action, we must add the caution that there obviously exist certain limits beyond which social heterogeneity becomes harmful for this purpose. Again, this point has not escaped Wade's attention: 'where stratification breeds class antagonism, . . . the bite of reputation loss may be reduced, because the reference group is confined to the subordinate class and collective free riding by members of the subordinate group might be encouraged' (Wade, 1988a: 193). Factional antagonisms are a constant threat for corporate village organizations. In Kottapalle village, Wade notes, the problem has been overcome 'by giving these antagonisms and suspicions an institutionalised expression': the composition of the irrigation council is carefully balanced between the two existing factions, decisions are taken by consensus, not majority vote, etc. (ibid. 195).
In fact, it is not only the objective economic situation of various users and the technical conditions characterizing the CPRs (such as, for example, whether the lands of the rich are located near the head-end or the tail-end of an irrigation system) that matter. The personalities of the village elite as well as the general social atmosphere prevailing in the society are also crucial factors influencing the possibility of cooperation. This seems to be the main lesson to draw from the following account of various field experiences with hill irrigation systems in Himachal Pradesh (Chard, 1994). These systems, known as kuhls, consist of small gravity channels constructed along mountainsides and they typically deliver water to several villages. In one case, heterogeneity promoted orderly and equitable distribution of water by supplying the necessary leadership. Thus, in Kalowan kuhl, till 1960 no set timings were observed for interand intra-village distribution of water and this situation of quasi-anarchy led to frequent disputes. In 1960, however, a local landowner who is also the village revenue collector mobilized farmers of the command area to frame a set of norms for water distribution. The agreement reached among the water users concerned was recorded in a document which contains elaborate information on days and timings of water supply to each of the six villages served by Kalowan kuhl and to all beneficiaries. Everyone has full knowledge of this information.
In Gandhori kuhl in contrast, no such co-ordinated move happened to the great profit of the local elite. Indeed, high-caste farmers have frequent access to water even though a majority of them have their fields at the tail-end of the system. The opposite is true of low-caste members who do not get water frequently despite the fact that the fields of many of them are located nearer to the channel. Revealingly, about 86 per cent of low-caste farmers stated that they would prefer government control of the kuhl while only 30 per cent of high-caste members expressed this preference. The lower-caste farmers understandably felt that under government control they would get their due share of kuhl water, whereas high-caste farmers prefer local control to keep their hold of the irrigation system and thereby get a larger share of the available water (ibid.). Of course, if the village elite have their irrigable lands more favourably located than in Gandhori kuhl it is still easier for them to enhance their interests at the expense of poorer CPR users. As pointed out by Pranab Bardhan, where the rich farmers are able to get enough water for their land without having to organize corporately and without having to incur large additional expenditures themselves since they own the land immediately below the canal outlets, they are highly likely even to block the formation of a co-operative water control committee at the expense of the small cultivators lower down. This is because such a committee might curtail their own irrigation freedom (Bardhan, 19936: 637).
Of course, the more hierarchical the social structure, the more important the personal qualities of the leadership. Thus, in the monastic forests of Imperial China to which we have already referred, we learn that 'under a weak or unscrupulous abbot, wood might be illegally felled and sold'. In contrast, there are examples of 'strong, devoted monks, who not only protected the land under their care, but led their community in restoring degraded lends' (Menzies, 1994: 69). Some time during the twelfth century for example, the abbot of the Fu Yan Chan (Zen) monastery at Heng Shan (Hunan province) stated: 'There are remote places on this peak where none of the Cunninghamia (a tree species) is in good condition. This means that we cannot be sure of the flow of the water and the supply of firewood to our retreat. We must not be remiss in our mission and foolishly fail to make provision for them' (cited from Menzies, 1994: 69). It must be added that, under the leadership of such enlightened abbots, not only the monk community but also the entire population of villagers depending upon the monastery (usually as tenants) were driven to better management of local forests.
The importance of a more or less harmonious social atmosphere (even in the presence of significant wealth differentials and inequalities in socio-political status) for efficient intra-village economic transactions may actually drive the upper strata to behave like good patrons and to accept a situation in which a relatively egalitarian access to local CPRs is guaranteed to everyone in the village, especially so if they are essential for the survival of the lower strata. In such a case, co-operation in the management of these CPRs under the leadership of the upper strata arises not because the richest people in the village have a greater interest in the CPRs as in Wade's example above, but because their natural leadership position enables them to ensure the success of CPR management schemes. In this context, it is useful to bear in mind some of the results achieved in Chapter 5. Indeed, assuming that the upper-class people have an AG payoff structure, or a CG payoff structure, or even a PD payoff structure, while the lower-class people have an AG payoff structure—given relatively harmonious social relations, they are willing to co-operate if the upper-class people show a good example, yet will behave non-cooperatively otherwise—we know that co operation will get established if the upper-class people assume a leadership role.
That egalitarian access to village CPRs can exist in differentiated societies is actually testified by several important examples. Thus, in situations where irrigation allows new land to be brought under cultivation, communities which have a highly asymmetrical social structure are sometimes observed to ensure fair distribution of the new lands to all members, possibly as a result of external pressures (exercised by donor or state agencies). For instance, in the strongly differentiated communities inhabiting the Senegal River valley (on the Senegalese bank), methods of alloting irrigation plots do not discriminate according to the social identity of the members. The assignment of plots through a lottery system tends to be favoured because, when some plots are better located than others, it is the only system that is considered legitimate by the villagers. In several village irrigation schemes, it has been found that, on the one hand, members of the minorities or lowest castes (including previous slaves) were not barred from access to irrigated land and, on the other hand, members of the highest castes did not arrogate to themselves more than one parcel or larger parcels than those assigned to the other participants in the scheme (Boutillier, 1982: 303-4; Diemer and van der Laan, 1987: 133). However, previous slaves are not entitled to hold significant positions in the irrigation group's committee since they may only be appointed as messengers or public criers (Diemer and van der Laan, 1987: 136, 141; see also Bloch, 1993; Shanmugaratnam et al., 1992: 27, 36, for similar observations about both Senegal and Mauritania; see also Tang, 1992, reporting evidence based on a large sample of irrigation systems).
Situations where social heterogeneity is combined with a relatively egalitarian treatment of all villagers with respect to use of local CPRs have been clearly pointed to by McKean in the case of Japan. In Japanese villages, indeed, inequality in private landholdings and political power went hand in hand with equal access to CPRs. Random distributions, assignment to parcels or products of the commons by lottery, and rotation systems were thus frequently used to avoid unfair allocation of use rights regarding these resources considered as vital to all people, rich and poor.
There thus exist a variety of situations that can arise when heterogeneity takes on the form of unequal endowments. As we have just noted, unequal endowments in private landholdings and social and political status can go together with a relatively egalitarian access to local CPRs. In such circumstances, the village elite can possibly provide the leadership required for collective action in the CPR domain. This is all the more likely if agrarian relations are of the patron-client type. When unequal endowments in private wealth are also reflected in unequal shares in the local CPRs, several possibilities present themselves that are more or less conducive to collective action. To begin with, the elite can enjoy a strategic position in their access to the CPRs and they may have no need for any kind of corporate organization. This corresponds to the case referred to by Bardhan in which landlords have all their irrigable lands located immediately below the canal outlets. This is the worst case since no collective action will take place and the poorer resource users will be at the losing end (see Chapter 5 for the corresponding game-theoretical model).
Alternatively, the strategic position of the rich may not enable them to dispense with the collaboration of the other users: for example, their lands are favourably located (near the headend of the water control infrastructure), yet the maintenance of the infrastructure on which they depend requires more labour efforts than they can supply themselves. Under such conditions, they may want to play a leadership role to ensure that the necessary public goods are provided. Their dependence on the labour of the poorer sections of the village community may confer on the latter a certain bargaining strength which can be used to ensure a co-operative solution in the allocation of the resource flows (similarly, the distribution of the water will not be monopolized by the landlords). The situation is still more favourable to co-operation if the elite have a disproportionately high interest in the CPRs but they do not hold any kind of strategic position: this is the case mentioned by Wade in which the lands of the rich are scattered all along the irrigation system. Here, again, one can expect them to assume leadership and, moreover, to be particularly attentive that water is equitably distributed throughout the whole system.
This typology is not exhaustive. It should nevertheless be sufficient to convey the point that social, political, and economic heterogeneity is not necessarily dooming cooperation in CPRs to failure. In important circumstances, heterogeneity may even promote co-operation compared with a situation in which there is more homogeneity in wealth and other kinds of endowments. Moreover, implicit in the above discussion are a number of criteria that must be taken into consideration when one wants to assess the impact of heterogeneity on collective action potential in specific field situations. In particular, the following questions have to be raised to delineate the problem: