Introduction

In the aspiration to upscale their activities, development agencies have a tendency to develop generic models for interventions in the global South. We define models as a prescribed set of structures and principles, which together form a blueprint to intervene in a given situation. Underlying these models is a particular, stabilized interpretation of what is identified as the problematic situation that the model is to address (Rap, 2006). The stabilized interpretation of the problematic situation allows the model to be broadly applicable, independently from the different socioeconomic realities in which it is implemented.

Generally development models share two features. First, they embody an ideological dimension both in the definition of the problematic situation and in the design structures and principles that constitute the model itself. Thus, these are often ‘enriched’ by ‘desired characteristics’ such as participation and empowerment that have become ‘seemingly universal values’ in the development discourse (Cornwall, 2007, p. 472). Second, models tend to be presented as embodying a dimension of success and qualify as ‘success stories’ (Molle, 2008, p. 138). Often anecdotes and ‘best-practices’ are utilized to highlight this ‘success’. Apart from producing ‘success’ of models through the selection of favorable pilot sites and through sympathetic evaluations by promoters, ‘cultural, esthetic and ideological understandings and practices’ are employed to confirm and bolster the ‘success’ of the model (Rap, 2006, p. 1313).

In the process of bolstering success, some actors from the national and international development establishment ‘work hardest of all to maintain coherent representations of their actions as instances of authorized policy, because it is always in their interest to do so’ (Mosse, 2004, p. 639). Successful models constitute a means for national elites, NGOs and other agencies from the development establishment to ensure funding from bilateral and international donors, and to legitimate their policies and development strategies (Mosse, 2004; Molle, 2008; Osserwaarde et al, 2008; Rusca and Schwartz, 2012). Similarly, through embracing and disseminating a particular model, consultants and ‘experts’ may develop lucrative careers and gain support from multilateral and bilateral donors, who see models as a way of efficiently upscaling their development activities and the associated transfer of funds (Rap, 2006).

The politics of developing and promoting such models have been subject to considerable research in past years (Roe, 1991; Mosse, 2004; Rap, 2006; Molle, 2008). In this article we focus on the relationship between the generic model and the everyday practices of contestation and consolidation of its implementation.Footnote 1 We explore what happens to these blueprint models and their associated design characteristics when they are applied in an everyday life context, with specific social configurations, spatial characteristics and ecological constitution. We argue that through everyday practices, design principles associated with a generic model are mitigated and adapted. By incorporating socially embedded institutions that reflect local power structures, models carry the danger of reproducing and even increasing existing inequalities. As a result, initial claims, such as participation or inclusion of the poorest, often fail to materialize (Fritzen, 2007; De Wit and Berner, 2009; Rusca and Schwartz, 2014). We develop these arguments by first presenting a literature-based discussion on the relation between generic development models and context-specific legal, administrative or social-cultural orders. We then discuss the case of the Water Users Association (WUA) model, established in the context of a water supply project implemented by the international NGO WaterAid in peri-urban Lilongwe, Malawi. The project, which is presented by WaterAid as a successful model to be replicated within and outside Malawi, was initiated in 2007 and is currently being upscaled to other low-income areas in Lilongwe.

Fieldwork for developing the case study was undertaken from November 2008 to February 2009 and from November 2013 to January 2014. Forty-eight semi-structured interviews were held with representatives of WaterAid (both at the headquarters in London and at the office of the Country Programme in Malawi), as well as WUAs representatives and key informants involved in or knowledgeable about the WUAs. Focus group discussions were held with WUAs, Community-based Organizations and the local NGO partners. Opinions of local consumers were gathered through semi-structured questionnaires with 30 households in the area benefiting from the project.

Generic Models and the Everyday Practices

The Context Specificity and Generic Models

In contrast with the idea of the broad applicability of models, various authors have highlighted how the complexities and uniqueness of context influence the outcome of the implementation of development models (Riggs, 1962; Cleaver, 2002, 2012). The influence of context on generic models has been explored from different disciplinary perspectives and draws from diverse empirical evidence. As far back as 1962, Riggs argued that bureaucracies in developing countries incorporate elements of both the Western bureaucratic model and the traditional societies, leading to ‘a set of new administrative structures, different from both the traditional and the modern, and a product of the mixture’ (Riggs, 1962, p. 22). At the core of this argument is the idea that groups in society are simultaneously exposed to and draw from different legal, administrative or social-cultural orders, developing new sets of mixed arrangements and rules. On a similar note, legal pluralism refers to the ‘coexistence and potential interdependence of local laws and disputing processes with the law and the administrative apparatus’ (von Benda-Beckmann and von Benda-Beckmann, 2006, p. 2).

Critical institutionalism draws similar conclusions on the relation between and interdependence of different institutional arrangements, by rejecting the dichotomy between ‘traditional’ and ‘modern’ institutions (Cleaver, 2002). Prevailing institutions are a mix of traditional and modern arrangements, resulting from a process of bricolage, through which norms, values and arrangements are consciously or unconsciously molded to serve a given purpose. In the process, the boundaries of ‘traditional’ and ‘modern’ become indistinct ‘and tradition becomes reinvented’ (Cleaver, 2002, p. 24). In a paper significantly entitled ‘How institutions elude design’, Cleaver and Franks (2005) argue that simple evolutionism, the idea of replacing traditional with modern institutions, does not reflect the dynamism and complexities of institutional formation and development. The process of bricolage is authoritative and, as ‘some bricoleurs are likely to posses more authoritative resources than others’ (Cleaver, 2002, p. 19), social inequalities may be preserved and reproduced. Cleaver (2002, p. 28) concludes that ‘the introduction of new bureaucratic institutions or organizational arrangements are not necessarily robust and enduring, nor do they automatically ensure beneficial collective action and optimum resource use’. Consequently, institutions that are not developed through an internal process of bricolage, but imposed externally, may be perceived as illegitimate.

Generic Models in the Water Services Sector

Generic models have played a fundamental role in shaping urban water supply strategies and development projects in the global South. The ‘modern infrastructural ideal’, promoting delivery of standardized water services by a single monopolistic provider to the entire urban population (Graham and Marvin, 2001), has guided development initiatives in the urban water sector over the past 50 years. With the onset of neoliberalism in the 1980s, the nature of the organization providing services and the institutions under which such an organization operates have been subject to considerable debate (Brown et al, 2000; Bakker, 2003; McDonald and Ruiters, 2005) and preference shifted towards private organizations or public-private partnerships. The ideal of standardized services provided through a single network, spatially bounded and uniformly managed, remained, however, predominant.

In recent years international donors and lending agencies appear to be questioning the ‘modern infrastructural ideal’ (McGranahan et al, 2006). ‘New’ service delivery models, including co-production (Harvey, 2007; Ahlers et al, 2014; Marston, 2014) and formalization of small-scale independent providers (Njiru, 2004; Samson, 2006; ADB, 2008; Schaub-Jones, 2008; World Bank, 2009; Ahlers et al, 2013; Cheng, 2014), are increasingly being propagated by donors and lending agencies as acceptable service modalities that may co-exist or operate in partnership with the formal utility (Batley, 2006; Ahlers et al, 2014). The shift to acknowledging multiple service modalities has not undermined the faith in ‘blueprint models’, which international donors and development agencies remain devoted to for their interventions (WSP, 2009, 2010; WaterAid, not dated).

The WUA Model

The WUA model emerged in the 1970s, as an alternative to state-run irrigation systems, increasingly perceived as underperforming (Rap, 2006; Meinzen-Dick, 2007). Decentralization of operation and maintenance tasks to WUAs has been promoted as a way to reduce overhead costs and increase efficiency (Smet, 2003; Suhardiman and Giordano, 2014). A similar assumption is made on revenue collection: not only are local farmers better placed to collect revenue, but this organizational form is also likely to increase willingness to pay within the association (Smet, 2003; Manor, 2004).

Promoted by development and lending agencies as an ‘institutional panacea’ (Meinzen-Dick, 2007, p. 15203) to optimize performance of irrigation systems, the WUA model is designed around desired characteristics of collective management, empowerment and participation, and goals of efficiency and full cost recovery (Watson et al, 1997; Smet, 2003; Rap and Wester, 2013; Suhardiman and Giordano, 2014). A study by the World Bank’s researchers Watson et al (1997, pp. 19–20) calls for the ‘need’ to establish WUAs to achieve ‘more efficient water delivery services’ while at the same time ensuring that projects are ‘better adapted to local needs and constraints’. According to the same study, members of the association jointly ‘develop common experiences and social norms over time’ (Watson et al, 1997, p. 107) that form the basis for decision making within the association. In this respect, WUAs are seen as a

demand-oriented approach because their organizational structure can be so designed so that they enable users to express their preferences, negotiate pricing and other contributions, and provide a platform for users to exercise their ‘voice’ with outside organizations. (Watson et al, 1997, p. 101)

The model is therefore based on the assumption that user participation will not only lead to more democratic and equitable water allocation, but also empower farmers, allowing for increased autonomy and direct control over the resource. WUAs are seen as a way to give ‘people at the grass roots greater influence over decisions that affect them’ (Manor, 2004, p. 194; see also Uphoff, 1992). Development and lending agencies promoting this model assume that users have shared interests, and their ability to influence decisions makes farmers more willing to contribute to the financial sustainability of this service provision modality. Last, members of WUAs are believed to have local knowledge that will improve sustainability of service provision.

More recently, possibly because of the ‘optimisms’ with which WUAs have been presented in the irrigation sector (Meinzen-Dick et al, 2002; Rap, 2006), this model has become more popular in urban water supply projects in low-income areas or informal settlements. In this context, WUAs have been strongly linked to point source technologies such as water kiosks (GTZ, 2009; WaterAid, not dated). Water kiosks are communal facilities situated in fixed locations where consumers can purchase water. The water supplied at the kiosk may originate from an independent water source or be supplied by the formal water utility (Kariuki and Schwartz, 2005). In the latter case water kiosks represent a low-cost technological alternative to extend services, where the water utility is unwilling or unable to provide services through in-house connections. The formal water utility supplies bulk water to the kiosk, from where it is sold to individual consumers. The WUA is one of the possible organizational modalities to operate these kiosks. WaterAid defines the WUA model as a management system constituted by a ‘cooperative water society where the communities establish a legal business entity and register it with the Government to operate all water facilities in a designated area’ (WaterAid, not dated, p. 6). In these areas WUAs sell water at the kiosks, collect revenue and pay the water utility for the bulk water.

While much research has been undertaken on irrigation WUAs (Rap, 2006; Meinzen-Dick, 2007; Rap and Wester, 2013; Suhardiman and Giordano, 2014), very little is known about the equivalent organizational structure in the urban water supply context. Water Users Association in irrigation and water supply services therefore share the same guiding principles of participation, equity, autonomy, and goals of efficiency and full cost recovery, attributable to the neoliberal ‘development orthodoxy’ (Suhardiman and Giordano, 2014, p. 91). Furthermore, both types of association are structured around users committees that are to streamline decision making and are usually established as formal organizations such as trusts, companies or some form of voluntary society (Smet, 2003; WaterAid, not dated).

WUA for Urban Water Supply in Lilongwe, Malawi

Water Supply Services in Lilongwe

Lilongwe, the capital of Malawi since 1975, has the highest population in the country, with estimates ranging between 972 000 (Urban Structure Plan of Lilongwe, LCC, 2013) and 800 907 (Lilongwe Water Board, 2013). Of the total population, approximately 70–75 per cent lives in 26 low-income areas (LIAs), in poor housing conditions and with little access to urban services (NSO, 2008; UN-HABITAT, 2011). The government has, since colonial times, relied on local chiefs for community development, and cultural, religious and administrative matters (Cammack et al, 2009). After independence and the onset of one-party rule in 1964, the newly established central government enacted the Chiefs Act, which conferred the President with the authority to appoint Chiefs and to decide upon their salaries (Eggen, 2011). Traditional systems and authorities therefore continue to play a major role and Malawians are nowadays at the same time ‘citizens of the state and subjects under a state-enforced chieftaincy system’ (Eggen, 2011, p. 313).

Until the beginning of the 1970s, water service provision was limited to the oldest parts of the city, which in Figure 1 are identified as Areas 1, 2 and 3. In 1972, shortly before the relocation of the capital in Lilongwe, services were extended to Area 20 (Capital Hill and Parliament Building), Area 12 (residential area for senior government officials), Area 15 (residential area for middle-level government officials), Area 27 (industrial area) and Area 35 (Army offices and residential area). Between 1975 and 1978, the network also reached Area 13 (hotels and office buildings), and Area 16 (banks), Area 19 (part of the City Center) and the residential Area 25.Footnote 2 According to the latest census (NSO, 2008), the Lilongwe Water Board, the sole authority with the mandate of providing services within the city limits (Waterworks Act, 1995), serves 31 per cent of the total urban population through in-house connections and yard taps and 41 per cent through the kiosks (NSO, 2008).Footnote 3 The kiosks are run by different management systems. Some of the kiosks are operated directly by utility employees who are responsible for selling water at the kiosk and collecting revenue. While this service modality ensures high revenue collection, it has high operation costs, as the salary of the kiosk attendant is borne by the water utility: the service delivery to low-income areas is already subsidised […] and you are having a lot of people employed, with very high staff costs. At the same time we were given the task to minimize numbers.Footnote 4 Other kiosks are run by private entrepreneurs, CBOs or faith-based organizations. The Lilongwe Water Board subsidizes the bulk water supplied to the kiosk, and the costs of the private operator or CBO operating the kiosk are added to the end fee charged to the consumer. Numerous cases are known where local elites, composed of both traditional leaders and politicians, embezzled the revenue collected at the kiosks, with the result that many communities ended up being in debt with the water utility. According to the Water Board, there was a lot of political interference, politicians used to go there, play their games (ibid.). WaterAid’s analysis suggests that abuses by politicians and traditional leaders have led to communities’ disillusionment, lack of a sense of ownership and in some cases acts of vandalism against the water facilities.Footnote 5 Furthermore, private operators charged elevated water fees to residents, who paid at least twice the price charged for water supplied through in-house connections in higher-income neighborhoods. The Lilongwe Water Board summarizes the situation as critical: despite the subsidy on bulk water, there were issues of ownership, issues of vandalism, issues of profiteering, with privates trying to maximise the profit.Footnote 6 Service provision to low-income areas is described by the General Director of the Lilongwe Water Board as a very big challenge. First in the sense that you don’t have access through which you can pass pipes. When it comes to the construction of the facility [i.e. kiosks] you don’t have room. […] Revenue collection has been a challenge because there has been a culture of non-payment.Footnote 7 Because of this ‘culture of non-payment’, the General Director Lilongwe Water Board was considering rationing or even cutting services at the kiosks: we reached a point in which we had accumulated over 31 million Malawi Kwacha [i.e. approximately 218 000 USD] in arrears and in our terms, it’s quite a bit of money. We reached the point to say enough is enough (ibid.). In areas where consumption rates were below 200 m3/month and, as such, were considered financially unviable, the Water Board planned to disconnect the kiosks. Implementing this measure would have meant closing 40 per cent of the operational kiosks. WaterAid Malawi’s intervention was targeted at reversing the rationing and kiosk disconnection policy.

Figure 1
figure 1

Map of Lilongwe.

Source: Based on National Statistics Office of Malawi (2008), population and housing census 2008.

Full size image

Designing the Model: Goals, Principles and Institutions

In 2009, WaterAid’s mission was to ‘overcome poverty by enabling the world’s poorest people to gain access to safe water, sanitation and hygiene education’.Footnote 8 The INGO’s main interest in the project was the provision of affordable and safe water to the poor, while ensuring empowerment and participation of local communities in decision making on water service provision to their areas.Footnote 9 Given Lilongwe Water Board’s monopoly on service provision within city boundaries and the dependency of low-income areas on the bulk water supplied by the latter, WaterAid aimed to achieve this goal by ensuring continuity of service provision by the water utility. The Lilongwe Water Board, on the other hand, aimed to comply with its obligation to provide potable water to low-income areas within city boundaries, while operating as a full commercial entity and thus achieving full cost recovery (Waterworks Act, 1995). For the utility, key priorities were recovering outstanding debts, ensuring high revenue collection, and minimizing its presence and role in the provision of services to low-income areas, while at the same time fulfilling its statutory obligation of supplying water to all residents. For the Lilongwe Water Board the only way out was to source out.Footnote 10 To achieve these goals, the INGO and the water utility designed and established a management system capable of ensuring ‘mutual satisfaction by the communities and water service providers’ (WA-LWB MoU, November 2004). This translated into the overarching objective of ‘providing sustainable, affordable and safe water to the poor while embracing full cost recovery’ (WaterAid, not dated, p. 8).

In practice, two entities were established for this purpose: a service provider, the WUA, responsible for operating the kiosks, and a Kiosk Management Unit (KMU), under the Finance Department of the Lilongwe Water Board. The KMU is entirely dedicated to supporting management of kiosks and acting as a regulator and mediator on the basis of consumer complaints. According to WaterAid, this management system was selected and approved by the local communities. Despite the fact that previously implemented management options of direct management by the LWB, management by private operators and management by point water committees were considered non-viable by the formal service provider or ‘had previously failed miserably’Footnote 11, communities were given the opportunity to choose between these options and that of management by WUAs. All six targeted communities opted for the establishment of WUAs. Interviews with consumers, however, raise questions on the inclusiveness and effectiveness of this participatory process. Consumers interviewed appeared to be unaware of the existence of the WUAs, let alone the selection of the management system and the election of WUA members.

The resulting model is strongly based on ‘modern’ bureaucratic institutions, which are to guide the functioning of the WUAs. First, emphasis is placed on the concept of autonomous service providers operating under private sector institutions and neoliberal principles (Baetti et al, 2006; Suhardiman and Giordano, 2014): not only are the WUAs to operate under principles of full cost recovery, they are also entitled to make profit, which is to be reinvested within the community. Second, WUAs are formalized through the registration as cooperative water societies, thus enhancing the capacity of the LWB to exert its authority over these service providers: dealing with a legal entity makes it much easier when it comes to enforcement.Footnote 12 Furthermore, the structure and functioning of the WUA mimics that of a formal water utility, with a hierarchic structure, a written constitution, and various policies and procedures to guide employees. All WUAs are composed of a Board of Trustees, an Executive Committee and a Secretariat. The role of the Board of Trustees is to give the strategic direction to the Executive Committee and monitor and approve its decisions (Constitution, Article 4.7). The Executive Committee supervises the activities of the Secretariat and facilitates the relationship between the Board and the Secretariat (Article 5.7). The Committee members meet once every month (see Table 1: Structure and responsibilities within the WUA). WUAs operate according to the same constitution and policies, some of which deal with the daily functioning of the WUAs, others with additional and extraordinary activities. For policies regarding procurement of goods and services, cash management, and auditing, the WUAs have aligned themselves to the procedures adhered to by the formal Water Board. In addition, the Associations have also developed detailed policies for employees, board members and executive committee members regarding annual leave, maternal and paternal leave, access to loans, and funeral allowances.

Table 1 Structure and Responsibilities within the WUA
Full size table

Although the Secretariat is the organ that operates the WUA on a daily basis, its role and responsibilities are not mentioned in the constitution. The Administrator heads this office and is responsible for managing the kiosks, ensuring that people are supplied with save [sic] and clean water all time, and making sure that kiosks are not braking down. In practice, the tasks of the Administrator are to monitor fellow employees who are under [his/her] position. We have kiosk attendants, kiosk inspectors, plumbers and a guard. […] I collect the sales made by the kiosk attendants in 44 kiosks and put it in the bank account weekly. […] I do the payments for the employees, to the Lilongwe Water Board, calculating the price we are buying from them, and pay for the purchase of goods.Footnote 13 Last, the relationship between the KMU and the WUAs is regulated by a service contract: the Lilongwe Water Board is responsible for providing water and for responding to any calls of the WUA in a timely manner (Area 50 Service contract, Article 7.2). The WUAs are responsible for providing services at the kiosk, collecting revenues from the consumers and paying the bill, including actual consumption and a portion of the arrears, to the Water Board (Area 50 Service contract, Article 7.3; see Figure 2: The WUA Model).

Figure 2
figure 2

The WUA model.

Full size image

Implementing the Model: Context Specificity and Everyday Practices of Service Provision

While the design principles and structure of the WUA reflect ‘modern’ institutions, the outcome of the everyday implementation includes socially embedded institutions, which adapt the model to the local context. For the management system to be implemented and effectively work, the system was to gain support of local chiefs. First, ensuring locations for the kiosks required consensus from landowners, the chiefs. Similarly, because of their mandate to keep financial and statistical records of the local population (Cammack et al, 2009; NSO, 2008), Traditional Authorities (TAs) were best placed to report on vandalism. Incorporating elites, and in particular the chiefs meant a smooth and more effective implementation of the new management system for project partners. Last, regardless (or because) of their authority in the community, chiefs are nominated by the President of Malawi and mandated to participate in any developmental activity within these communities. Their involvement is obligatory.

Rather than aiming at a broader and more ‘democratic’ approval of the water supply system, therefore, the model implemented by WaterAid ensured first and foremost the support and approval of local elites, by granting them a number of privileges ranging from key positions within the Associations, to authority in decision making, to financial benefits and increased status within the community. In particular, rather than addressing the challenge of misappropriation of funds by local elites, the practice of embezzlement was institutionalized by incorporating the perpetuators in the WUAs: now they are no longer playing their games in the low-income areas. We have them in the Board, we have them in the Executive Committees.Footnote 14 While the Board is per policy composed of the most prominent members of the community (traditional chiefs, religious leaders and local members of parliament), officially it is the community members who elect the Executive Committee representatives. Community members aspiring to this position are to apply before elections and, if shortlisted, can run for office. In practice, however, it is chief's discretion to choose out of his constituency someone who will be able to speak on their behalf.Footnote 15 Furthermore, responses from the household survey suggest that the majority of the people living in the project area were unaware of both the elections of and application procedures for the Executive Committee. As a result, both the ‘strategic’ and the ‘supervisory’ organ of the WUAs were co-opted by local elites.

Second, the authority of elites is ensured through a strategic distribution of responsibilities and benefits within the Association’s layers. WUAs are designed in such a way that key responsibilities for service provision, such as operational tasks and financial management, are with the Secretariat, while decision-making power is with the Board and the Executive Committees. The members of the Board have self-regulating powers, are responsible for designing policies and regulations, establish their own honorarium (Constitution, Articles 4 and 8), and decide upon the employment of kiosk attendants. On the other hand, the Secretariat, responsible for all activities of the WUAs, has no decision-making power and is composed of only one member and the field staff (that is, kiosk attendants and a plumber). Although the structure of the WUAs appears rather top-heavy, members of the WUA themselves do not seem to question or challenge this structure. On the contrary, both Administrators and Committee Members agree that it is necessary to have elites guiding the organization: we need them, they know betterFootnote 16.

Third, the decision-making power ensures entitlement to other benefits, such as financial incentives, which are directly decided upon by Board members. Formally, members of the Executive Committee and the Board of Trustees are not employed by the WUA but act as advisors. As such, they receive a monthly honorarium rather than a salary. Members of the Secretariat, however, are employees fully dedicated to service provision and thus entitled to a salary. Although an official distinction is made between salaries and honoraria, when comparing the financial benefits of employees and advisors it appears that Board and Executive Committees gain the most from this distribution (see Table 2: Honoraria and salaries). The Board meets every 3 months, but members receive a monthly honorarium. This honorarium is approximately equivalent to between one-third and one-fourth of the monthly salary of the Administrator, who works full time and is ultimately responsible for the kiosk attendants and their performance, as well as for all financial aspects, including the payment of bulk water to the Lilongwe Water Board.

Table 2 Honoraria and Salaries (Malawi Kwatcha - MKW)
Full size table

In addition, WUAs’ policies and regulations may potentially increase the status of local elites and, thereby reinforce inequalities. Policies explicitly reaffirm the distinction between elites and ‘ordinary’ community members (Policies and procedure, Mgona, October 2008, p. 29). This distinction is then used to define the amounts that the WUA allocates to funerals, where ‘ordinary community members’ receive 10 times less than leaders such as village heads, politicians and church leaders (see Table 3: Amounts donated for different community events). The same distinctions are made within the Water Users Associations: Board and Executive Committee are entitled to privileges that are equal to and exceed those granted to the employees. When it comes to funeral allowances, kiosk attendants are denied any financial support, while Board members and Executive Committee members are entitled to allowances equal to two to 6 months salaries of the kiosk attendants (see Table 4 WUAs Policies and procedures for employees and other active members). Such benefits, we argue, lead to an additional indirect benefit: the power to distribute profits within the community by selecting events and projects to be developed increases the status of local elites within the community. The profit-making mechanism becomes another way of ensuring privileges to the ‘important’ community members and further increases the tension between the design principles embedded in the model and the practices that undermine them. As a member of a local NGO implementing the project highlighted, you are getting money from poor people, but you are conducting meetings at hotels, lake shore areas […] all while the access to water is less than 100 per cent.Footnote 17

Table 3 Amounts donated for different community events
Full size table
Table 4 WUAs policies and procedures for employees and other active members
Full size table

Mechanisms to re-balance these asymmetries are included in the policies, but rarely implemented. It is established, for instance, that each WUA is to organize an annual general meeting to present and discuss financial reports and to decide on WUAs’ surplus spending. In practice, however, annual general meetings do not take place regularlyFootnote 18 and, in some cases, the selection of attendees only includes influential community members (clergy, public companies employees, chiefs). ‘Ordinary’ community members seem to be displaying surprisingly little contestation: some seem to be unaware of the system, while others consider WUA bodies to be reserved for influential community members and employees.Footnote 19

The Impact of Institutional Bricolage on Service Provision

The elaboration regarding the composition, enumeration and secondary benefits of the different organs of the WUA becomes relevant when it comes to assessing the service provided by the WUAs. First, WUAs are an expensive service provider. At the beginning of this research (2008–2009), the Water Board was selling water to the WUAs at MK2 58 per m3. The WUAs, in turn, were reselling this water at more than double this price at the kiosk (MK 125/m3). In 2014, the ratio of the price of the bulk water (MK 91/m3) and the final price to consumers (MK 300/m3) had grown from 1:2.15 to 1:3.29. The high price charged at the kiosks to consumers is the result of a number of factors. As the WUA has to follow the principles of cost-recovery and profit-making, all costs associated with the incorporation of the local elite in the WUA and the repayment of arrears automatically translate into higher prices at the kiosks. The top-heavy design of the WUA means high personnel costs. Although the WUA is only responsible for a small component of service delivery (revenue collection), the overhead resulting from the many Board members and Executive Committee members is considerable.

Second, the WUA not only has to cover its service costs, but also has to pay the arrears that have resulted from the embezzlement of kiosk funds prior to the establishment of the WUAs. As such, the payments of the previous debt are included in the price charged to consumers at the kiosk. According to WaterAid, there is ‘commendable improvement in billing and repayment of arrears’. From the peak value of MKW 30.5 million in January 2006, arrears dropped to MKW 14.5 million by January 2007 (WaterAid, not dated, p. 7).

Last, despite the increase in price, the WUAs were hardly able to ensure improved service provision to their consumers: asset maintenance and continuity of the service remain of poor quality. In case of breakage, a response from the Lilongwe Water Board may take up to 3 months. The complex bureaucratic procedures adopted by the utility for the procurement of goods further slow down the simplest repair. Some kiosks have been abandoned because LBW considered the costs of repairing them too high and therefore not viable. In some areas shortages last –2–3 daysFootnote 20 while other areas are only supplied 2 days per week.Footnote 21

Conclusions

Development projects are not implemented in a vacuum. Models are designed as generic constructions, but have to deal with the complexities and diversity of the local social and political context. To be ‘understood’ and accepted locally, generic models are transformed to reflect socio-political realities and, as such, carry the danger of reproducing existing disparities and local power asymmetries. The implementation of the WUA model is the result of a negotiation, an authoritative process (Cleaver, 2002) in which some groups and interests prevail to the detriment of others. In the everyday implementation of the model goals, guiding principles and institutions that prevail and guide service delivery are the ones that align with the interests of the most powerful players, the Lilongwe Water Board and the elites within the targeted community. For the utility, the priorities were recovering outstanding debts, ensuring high revenue collection and minimizing its direct presence in low-income areas, while at the same time ensuring access to potable water. In this respect, the WUA model was highly effective: not only were the WUAs able to meet these targets, but they were also able to ensure payment of outstanding debts and make profits. Ironically, WUAs, which operate in lower-income areas, succeeded in achieving one of the objectives that formal utilities often fail to achieve, that is, causing the former Managing Director of Uganda’s National Water and Sewerage Corporation to state that full cost recovery is a myth (Muhairwe, 2006).

For the local elites the management system became an opportunity to reinforce their privileged status within the community and gain access to resources and decision-making power. As for WaterAid, the everyday implementation of the model ensured the efficient achievement of the overall project objective, having water flowing again in the targeted communities. On the other hand, this objective was achieved to the detriment of norms, goals and guiding principles, reflected in WaterAid’s mission statement and agreement with the Lilongwe Water Board. Full cost recovery principles prevailed over the objective of sustainable and affordable water supply to the poor. Similarly, hierarchic and oligarchic structures and institutions prevailed over principles of empowerment of the community at large and democratization of decision-making processes. Local elites, previously reported by the Lilongwe Water Board and WaterAid as embezzling funds from the kiosks, are now formally recognized and legitimated by the same organizations as leading authorities in the WUAs. The WUA model internalized the ‘costs’ of embezzlement, which were transferred from the water utility to the consumers.

In this local adaptation of the model, therefore, design characteristics associated with empowerment and inclusion of the poor are transformed into ‘disempowering forms of participation’ (Hickey, 2002, p. 843) or ‘conservative’ forms of participation that only allow for the inclusion of those community members who are already in a position of power. Rather than giving ‘ordinary’ community members a platform to influence decisions that affect them, as prescribed by the generic model, the WUA becomes a mechanism for existing elites to fortify and extend their elite positions. The interaction between the context and the model thus led to increased inequality within the community, while at the same time ensuring smoother implementation of the model.

This research raises further questions on the role of ‘ordinary’ community members in challenging or facilitating the process of elite capture. Elite capture might also be agreed upon or even be encouraged by community members, who might opt for someone they consider more clever, vocal, educated or capable to represent them. On a similar note, this research shows that the everyday implementation of the model leads to uneven outcomes and distribution of benefits among community members. Further research is required on how benefits are distributed among community members and, more specifically, whether and under what circumstances benefits trickle down to ‘ordinary’ community members. Last, focusing on inequalities within community members in low-income areas shall not distract attention from the larger issues of inequalities within the city and among different areas in the city. The objective of ensuring equitable water supply within the city has so far eluded all actors concerned: the public water utility, community based organizations, local leaders and NGOs.