Elites’ roles and community relations

Although both case study locations were subject to colonial, Frelimo and NGO-led interventions, they differ significantly in terms of their historical development trajectories. The first case study, where GA-Moz worked, was the Locality of Chicomo, which was located in Manhiça District in Maputo Province, and consisted of five bairros.Footnote ² In the early days of colonial rule, Manhiça's residents mainly lived in dispersed homesteads alongside the floodplain of the Incomati River under a system of hereditary leadership. Although quite close to the capital city and accessible from the national road, there was little major external investment in the area, although the creation of the nearby Maragra sugar plantation resulted in the installation of a regulo by the Portuguese authorities in the late 1930s. The settlement as it is recognised today was formed as a communal village following the Incomati River floods in 1978, when the government permanently resettled some 800 households from the low-lying floodplain to areas of higher ground (Republic Popular de Moçambique 1983; van den Berg Reference Van den Berg1987). At the same time, due to its status as one of the first communal villages, Frelimo's district leaders took a direct interest in the settlement, and moved in one of its respected party leaders to take control of local affairs. At the time of the case study, this individual, who was known as the ‘president’, was supported by the party secretaries, each of whom was responsible for leading one bairro. Each secretary was, in turn, supported by a small group of ‘problem-solvers’, who represented the remnants of the old system of traditional leadership in the area. Even though these two systems of authority worked together in small groups, day-to-day tasks were still split along official ‘executive’ and informal ‘traditional’ lines, with secretaries acting as local government representatives, and traditional leaders being called upon to resolve domestic conflicts, and perform rituals and ceremonies for the local population.

The president held considerable influence within this system, and a notable social and material gap divided him from the political non-elite. His appointment, officially by election, occurred across the locality once every four years, although it was considered highly unlikely that local residents would vote against a representative associated with the Frelimo party. Much of the president's authority stemmed from his image as the settlement's local ‘protector’ or ‘provider’, a representation created during the community resettlement process when he played a pivotal role in diverting government resources to the local population. This role was extended during the civil war when Manhiça came under sustained threat from opposition Renamo forces with the result that the district was cut off from national Frelimo control for extended periods of time. The president's role was also consistent with that of the national Frelimo Party's generally ‘paternalistic’ attitude towards its rural subjects in which the party took on the job of ‘modernising’ poor people (Hanlon & Mosse Reference Hanlon and Mosse2010). The president's status thus transcended that of day-to-day politics within locality-life. Like Chicomo's president, bairro secretaries were also selected by vote once every four years, although there was a greater likelihood of these authority figures being deposed, suggesting that the secretary's power relative to non-elites was smaller. Respondents during interviews were more willing to criticise the performance of village secretaries, mainly because they were lower in rank and therefore ‘closer’ to the people. These dynamics were observable on a number of occasions during village meetings when non-elites would directly question local leaders on service provision, including water pumps and cattle.

In contrast to Chicomo Locality, the second study site – Jofane Town, located in Chokwé District, Gaza Province, where CDG worked – experienced high levels of agricultural and infrastructural investment during the colonial and socialist periods. Following independence, the national Frelimo authorities transformed the former Portuguese plantation and settlement scheme in the region, the Colonato do Limpopo, into a huge agro-industrial complex, the Complexo Agro-Industrial do Limpopo (CAIL), the largest in the country at the time (Bowen Reference Bowen1989). Chokwé District was also the birthplace of the Frelimo cooperative movement, at the height of which there were 12 cooperatives locally active, with a combined membership of over 3,000 farmers (Ministério da Agricultura 1982). Following Frelimo's Fourth Party Congress in 1983, and in the context of ongoing South African destabilisation of the country, the government began a Priority Districts Programme, channelling scarce resource to regions, including Chokwé, where economic, military and climatic conditions presented the best chance of positive results (Bowen Reference Bowen1989). These economic and technological investments targeted at the agricultural sector resulted in a relatively well-developed government administration as local secretaries were often called upon to organise labour and direct resources (Hermele Reference Hermele1992). The leadership system also benefited from relatively low levels of disruption due to civil war as Frelimo's Maputo-based government invested heavily in the defence of the district due to its strategic importance as the ‘granary of Mozambique’ (Wardman Reference Wardman1985).

This legacy of external investment was still evident at the time of the case study, as governance arrangements within Jofane's bairros were found to be dominated by the secretary system. Unlike in Chicomo, arrangements also included ‘vice-secretaries’ to assist with the day-to-running of community affairs. Secretaries in Jofane were relatively well-educated, as shown by their higher literacy rates than their Chicomo-based counterparts (90% compared with 55%) (CDG 2005). In addition, traditional leaders in Jofane were more marginalised within the bairro structure compared with Chicomo, as they were consistently ranked as having lower status by community-based respondents. This difference was explained by some key informants as due to the relatively high levels of forced resettlement and labour that occurred in the region under Portuguese rule, and the unpopular roles that former regulos took within this system of oppression. As in Chicomo, selection of bairro secretaries occurred once every four years by general election. In general, secretaries were well-respected by the local populations that they governed, and community meetings mainly revolved around the transfer of information from district government ‘downwards’, as well as the collection of ‘complaints’ from the local population to pass on ‘upwards’ to district-level authority. These meetings appeared to be mainly technical, planning exercises, and there was little direct questioning of secretary authority.

In sum, the two case studies illustrate variations in elite structures, and show how these differences are linked to the particular histories of socioeconomic development found in their respective localities. Because of these differences, the basis of local elite legitimacy also varied between localities. In the first study, the president's authority derived from his association with the Frelimo Party, and from his status as a figure of ‘stability’ during the community's turbulent history. In the second study, local secretaries derived legitimacy from their long history of involvement in the provision of external inputs for economic growth. Both of these examples of legitimacy are recently constructed and differ from the ‘timeless’ logic of ‘traditional’ relations that dominate the patrimonial literature (Pitcher et al. Reference Pitcher, Moran and Johnston2009). This is not to suggest, however, that kin-based authorities were completely excluded from local affairs. The case studies also show that dominant government structures existed alongside and worked with ‘traditional’ authorities. This likely reflects the many compromises that secretaries were forced to make as the civil war progressed and the Maputo-based Frelimo Party's grip on the countryside weakened. Although the two governance systems are widely considered in Mozambique to be distinct and parallel, and therefore cannot be viewed as ‘hybrid’, these findings add weight to recent scholarship which suggests that ‘selected’ and ‘elected’ forms of governance can exist alongside one another and work together (Logan Reference Logan2009).

Project capture by elites

This section considers whether and how community-based projects were captured by local elites following their introduction to the case study areas by the intervening NGOs, GA-Moz and CDG. ‘Capture’ in this instance is considered to occur when established local elites dominate project decision-making, whether those elites take key positions within organisations themselves or affect their activities from outside. Both NGOs tried to influence this process by attempting to control the positions that local leaders took vis-à-vis newly introduced organisations. Following the civil war, the perceived ‘neutrality’ of international NGOs by the donor community provided foreign agencies with the necessary legitimacy to enter conflict-affected areas (Costy 1996). This historical role was reflected by GA-Moz's approach to community interventions as commissions were designed along a technical rationale intended to ensure their independence from local community affairs. As one Maputo-based GA-Moz food security manager explained: ‘The commissions are separated from other groups. They are linked to the outside world by their results … For example, water is useful for the whole village, and the people in commissions are accountable to the community in this way.’

Commissions were therefore designed on a ‘counter-elite’ logic (Wong Reference Wong2010) in which legitimacy stemmed from the technical performance of trained commission personnel rather than the political power of local leaders. This, it was conjectured, would spread decision-making and financial management across a greater number of people, thus allowing for monitoring of each others’ work and greater transparency (cf. Thorp et al. Reference Thorp, Stewart and Heyer2005). The strategy is consistent with the idea that the interests of poor people can be protected when the representation of elites is sufficiently diversified for a division of opinions to develop among them (Gow & Vansant Reference Gow and Vansant1983). It was also hoped that commissions would provide marginalised community groups, such as women, with the opportunity to improve their social status by becoming more involved in community affairs.³ GA-Moz's fieldworkers, for example, encouraged the nomination of certain non- and low-ranking elites to run commissions by creating pools of candidates from which community members could select. These positions were then consolidated through the provision of management training and reinforced annually by externally organised ‘participatory assessment exercise’ meetings.

In contrast to GA-Moz, CDG's preference was for community authorities to run local, externally introduced organisations. This approach is referred to by Wong (Reference Wong2010) as a ‘co-opt’ elite strategy, in which elite structures are deliberately ‘absorbed’ into externally facilitated organisations. This is because CDG believed that cooperatives would function more effectively with community elites at the helm due to their greater management and problem-solving capabilities. It is consistent with the idea that if the more powerful have similar interests to the less powerful, then the former might greatly enhance the probability of successful organisation through investment of their resources in group management (Ostrom Reference Ostrom1999). The co-opt approach was also ideologically motivated, as CDG viewed cooperative leadership positions as the ‘proper place’ for secretary leaders. As one CDG food security manager explained: ‘Secretaries have always looked after this area and run local affairs. It is what the people expect. This place was the birthplace of … the Party. [Frelimo] has been able to change a lot of things on the ground as a result.’ In this quotation, the NGO manager is linking effective local leadership with the governing national party, effectively associating secretary leadership with ‘benevolence’. This ‘co-opt’ approach, however, can also be viewed as a form of control as CDG aimed to draw local leaders into an externally initiated governance structure designed to manage a cooperative agricultural system.

In sum, elite domination of project decision-making was actively discouraged by GA-Moz but generally favoured by CDG. The positions taken by the case study NGOs, as well as the leadership outcomes for the organisations that they introduced, are summarised in Table 1.

Table 1

Leadership outcomes for newly introduced organisations

Table 1 shows that all newly introduced organisations were ‘captured’ by community leaders in ways that were not intended by the NGOs. In the case of water commissions in case study 1, respondents reported that the original leadership structures were still intact within the majority of organisations. However, commission members struggled to maintain water pumps and had increasingly sought out bairro secretaries to help resolve water management disputes between themselves and community members that resulted. It was widely believed by respondents that this was the correct role for a secretary as water was something that was available for everyone and this was analogous to the secretary's function of being ‘one of the people’. It was also suggested that the presence of problem-solvers within the secretary's group would increase the ability of community-based authorities to settle water-related disputes.

In the case of animal commissions in case study 1, the system of decentralised, non-elite management intended by the NGO had been replaced by a centralised animal management group run by the locality's president. According to GA-Moz's fieldworkers, the president's interest in the commission stemmed from the high economic value of the good being managed, and the prestige associated with cattle-ownership in southern Mozambique in general (cf. van den Berg Reference Van den Berg1987). The NGO attempted therefore to establish a link between elite interest in commissions and personal gain. The president, however, rebuffed this claim, stating that it was his ‘duty’ to take charge of the commission, a belief consistent with his historical role as the main ‘provider’ to the local population. This position was reinforced by the majority of community-based respondents who stated that the president had assumed this role due to his ‘expertise’ in animal husbandry. The overall message here was that, given the high value of the good under management, it was only fitting that an individual who commanded high levels of respect should be assigned the task. In the first case study, therefore, an association was created between the nature of the good being managed and local leaders’ status in community affairs.

In the second case study, a bairro secretary and vice-secretary were initially elected by the local population to run the cooperative as CDG intended. However, contrary to expectations, the community-based leaders declined to take up the positions. Publicly, they stated that they were ‘too busy’ running bairro affairs to take on any extra responsibility but, in later interviews, expressed doubts about the viability of cooperative forms of agriculture, citing past limitations and mistakes made by Frelimo in the region in the 1980s (cf. Coelho Reference Coelho1998). The secretary explained: ‘In the past under Frelimo we had large state farms and these provided the people with all that they required. I was often involved in informing the people about these activities, and organising them [for work]. But the cooperatives in this region were not so successful … I learned the hard way that they do not deliver and now I don't involve myself in them.’ The careful consideration shown in this quotation by the secretary about which types of external investment to support reflects the important role they can play in building the legitimacy and reputation of a local leader. In this case, the secretary backed an alternative group of cooperative leaders who represented ‘middle farmers’. The parents of these individuals had been part of the registo dos agricultores Africanos (Register of African Farmers) established by the Portuguese authorities in the region in the 1950s in an attempt to create a class of peasants who would produce for the market (van den Berg Reference Van den Berg1987). Middle farmers were not officially recognised by the socialist Frelimo administration but captured many government-mandated collective forms of production in the late 1970s and 1980s nonetheless as a route to capital accumulation (Harris Reference Harris1980; Dinerman Reference Dinerman2001). This was found in Jofane where, according to CDG interviewees, ‘capture’ by middle farmers occurred as they sought out personal gain. In this region, therefore, CDG essentially found itself organisationally allied to middle-income farmers. However, most non-elite cooperative members disagreed with this analysis, stating that former privadores were well-placed to run the cooperative due to their experiences of ‘modern’, irrigated agriculture.

In sum, elite capture of NGO projects was commonplace, although the variety of explanations offered for its occurrence were far more diverse than those suggested by the CDD-elite capture literature. NGO personnel often described a ‘push’ from elites to gain political and material benefits from newly established organisations, but many community members offered an alternative view, explaining how elites were ‘pulled’ into such organisations by newly elected leaders. This suggests that elite involvement in projects can represent a logical response to the complex demands placed on organisations during CDD, in addition to a bid to gain materially and politically from externally initiated programmes.

Project benefits capture by elites

The previous section demonstrated that project capture occurred widely in the two NGO-initiated CDD projects. This section examines whether NGO-provided goods continued to flow to ordinary communities under these regimes, indicating that benevolent capture was taking place, or were seized by elite groups, implying malevolent capture. Table 2 summarises the elite capture outcomes in both case studies, and main checks and balances on elite activity that were found. As set out in the second section, checks and balances are classified as either ‘top-down’, ‘horizontal’ or ‘bottom-up’.

Table 2

Elite capture outcomes, and checks and balances on elite behaviour

Table 2 shows the presence of both benevolent and malevolent capture in the case studies, as well as a variety of checks and balances in operation. In the case of the animal commission in case study 1, no evidence of malevolent capture of cattle or commission funds by the president was found in either NGO documentation or key informant interviews. GA-Moz reported, however, that cattle were being distributed by the president to Chicomo's better-off households and that this was evidence that the local representatives of the Frelimo party had captured the distribution programme. Indeed, the national Frelimo Party has a history of rewarding people who were loyal to it during the war (West & Myers Reference West and Myers1996). This finding reflects fears that community-based economic and political decentralisation is being used by Frelimo's national political elites as an inexpensive way of extending the state's reach into rural areas, thus helping to undermine Renamo political influence which is traditionally based in the countryside (Alexander Reference Alexander1997; Kyed & Buur Reference Kyed and Buur2006).

Because of these concerns, GA-Moz's fieldworkers attempted to retain partial control of the cattle allocation system via the introduction of ‘transparency and accountability’ rules aimed at ‘ensuring distribution…to a wide range of families, including more vulnerable female-headed households’ (GlobalAid 2006). This was important because, as one GA-Moz fieldworker explained, ‘cattle ownership helps to raise the status of women in the community relative to men’. However, the president was able to resist these attempts, arguing that better-off families should be prioritised as they were more capable of taking care of animals. Interviews with local Frelimo representatives revealed that the party was likely to be having a positive effect on keeping the president's activities in check. This is because ‘number of cattle distributed’ was a key indicator of local development success in the District Development Plan (Ministério da Administração Estatal 2005). As a local party representative, the president was more likely to be responding to these internal forms of checks and balances than external ‘accountability measures’ imposed by the NGO.

Although top-down accountability measures exerted by GA-Moz appeared to be relatively ineffective, there was more success in the case of CDG's cooperative in the second case study where no evidence of elite capture of project benefits could be found. This is partly because cooperative forms of organisation were relatively familiar to cooperatists, and this allowed group norms of behaviour to be relatively quickly established by the intervening NGO. However, evidence from across SSA shows that even well-organised cooperatives have been prone to elite ‘takeover’ (Bingen et al. Reference Bingen, Serrano and Howard2003), and the presence of a statute by itself cannot explain the absence of malevolent elite capture. In this case, uncertainty over cooperative leadership following the secretary's resignation created the necessary space for a competitive political system to emerge. This consisted of cooperative leaders, which, as described in the previous section, consisted of ‘middle’ farmers, and a small group of former community secretaries that had banded together in response to the NGO intervention. This split reflects old divides that were observed within the 1980s within cooperatives between party and state officials on the one hand and cooperative leaders on the other hand (Coelho Reference Coelho1998). The group of former secretaries regularly attended Union meetings and challenged the organisation's leaders on a range of issues varying from the reporting of cooperative accounts to use of the tractor. This split was possible as there was greater overall capacity amongst local leaders in Jofane compared with Chicomo due to the history of investment in the region.

These arrangements were reinforced by ordinary cooperatists who were able to maintain a degree of control over elite activities through the withdrawal of cooperative behaviour. This was possible because, according to Dinerman (Reference Dinerman2001), the main reason ordinary people have historically engaged in cooperatives in Mozambique is to spread income risk. This suggests that farmers position themselves relative to cooperatives depending on what they can gain from their participation, not according to the dominion of their leaders. For this reason, mobilising people for important work on collectively managed land such as irrigation or application of pesticide can prove difficult. This was observed by Wardman (Reference Wardman1985) and Bowen (Reference Bowen1989), and also reflected present-day attitudes amongst cooperatists in the second case study who viewed their participation in group-based work as conditional upon an ongoing flow of monetary benefits towards them. Withdrawal of labour from the cooperative was therefore an effective way for people to exert control over the Union leadership. It reflects the argument advanced by Pitcher et al. (Reference Pitcher, Moran and Johnston2009) who stated that the act of ‘buying’ support from clients by patrons is more difficult than often presumed.

Although elite capture of projects was observed as relatively benign in the cases of the animal commission and the cooperative, respondent attitudes towards water commissions in case study 1 were found to be more mixed. According to GA-Moz, commission positions were strictly ‘voluntary’, and use of water funds by commission members for personal gain was referred to by agency personnel as ‘eating’. This reflects Berman's (Reference Berman1998) reference to Africa's politics as focused on ‘devouring’, or getting one's share of resources, or Bayart's (Reference Bayart1993) ‘politics of the belly’. Some commission personnel acknowledged that funds had been removed from their organisations’ coffers in this manner, but justified this as payment for their work. Community members, on the other hand, held mixed views about the use of commission funds in this way, with some interviewees condemning leaders’ activities, and others offering their leaders support in spite of on-going elite misappropriation. This latter example illustrates how ‘benevolent’ and ‘malevolent’ can mean different things to different people, and draws attention to the need to understand who is defining such terms in studies of elite capture.

Some commentators have observed in southern Africa that accountability to the local community by user committees is often lacking or weak (Matose 2009). This appeared to be the case in the first case study where NGO-introduced rules around recording and reporting funds were largely ignored by commission members. One NGO member explained how local elites were able to manipulate such arrangements: ‘The people who work in the water commissions hide the registration of the money to take advantage. When you ask for the accounts, they say: ‘we lost them’, or ‘the money is with this or that person’. If you have everything registered then the loss of money will be easier to find. When the money is passed from one person to another, a quantity is always taken out.’ This does not mean, however, that ordinary community members were powerless in the face of such problems. Over half of respondents interviewed acknowledged that commission members were struggling to maintain water services under difficult circumstances and that leaders deserved some form of financial compensation for their efforts. Others, however, reported that they had ceased paying their water rates in protest, with the result that the organisation was chronically underfunded. This form of ‘passive’ resistance to elite domination was similar to that which was occurring in the case of the cooperative. In addition, a smaller number of respondents stated that there was a need to ‘remove’ their secretary as a consequence of these failings. Although this threat had yet to be carried out, key informants reported that some community members had begun a campaign to remove their bairro leader on the basis of ongoing water management problems.

In these cases, the ability of NGOs to impose externally devised checks and balances onto communities was mixed. In the first case study, they were found to be relatively ineffective as user-based commissions represented relatively unfamiliar ways of working to the communities involved. The tendency, therefore, was for these NGO-initiated groups to be drawn into local governance arrangements relatively quickly where externally introduced ‘transparency’ procedures could be easily manipulated. In this way, commissions appeared to replicate social relations that were already present in the community, acting as an extension of the local-level politics rather than as arenas in which new arrangements of actors could be realised (cf. Coelho & Favareto Reference Coelho and Favareto2008). In the second case study, the union statute drawn up by CDG proved more effective in regulating elite activities, although cooperatists had an historical template on which to work, and this led to the replication of old elite–non elite relations that had played out in the Chokwé region in the past. However, these externally introduced accountability arrangements are only a part of the picture. As Table 2 shows, there were a range of alternative checks and balances in place, some of which proved relatively effective in regulating elite activities, but which were rarely taken into account or encouraged by the intervening NGOs.