District Assemblies in a fix: the perils of the politics of capacity in the political and administrative reforms in Malawi

Abstract

This article appraises the prospects of District Assemblies institutionalising local governance structures and processes that are responsive, democratic and capable of improving the livelihoods of the poor. District Assemblies have been established against the backdrop of a radically reconstituted policy, governance and administrative framework following the transition from one-party rule to multiparty democracy in Malawi in May 1994. The underlying argument of this paper is that the potential success of the District Assemblies is effectively hampered by widespread self-seeking tendencies and orientations among various stakeholders in grassroots development, strategically veiled as lack of capacity on the part of the decentralised planning framework. Unless these seemingly idiosyncratic tendencies are subordinated to the underlying noble cause of the decentralisation policy and institutional reforms, the trinity of good governance, development and poverty reduction in the evolving structures and processes of local government will remain a virtually unattainable ideal.

1. Introduction

Since the turn of the 1990s, democratic decentralisation has become more or less an article of faith as the centrepiece of sweeping policy reforms currently under way in virtually all developing countries. These fairly standardised reforms are principally geared towards achieving the trinity of sustainable good governance, development and poverty reduction. There is immense confidence in the ability of democratic decentralisation to deliver, mainly because previous failed attempts at decentralisation were ‘largely initiatives in public administration without any serious democratic component’ (Blair, 2000: 21). Decentralisation is in fact billed as one of the emerging trends and fashions that have returned poverty and poverty reduction to the foreground in the discipline of development studies. The underlying rationale for democratic decentralisation as a vehicle for achieving sustainable good governance, development and poverty reduction is that it brings the government closer to the grassroots both in spatial and institutional senses, hence making it more representative, accountable and responsive, which in turn improves the appropriateness and effectiveness of development projects (Crook & Sverrisson, 1999; Crook, 2002; UNCDF, 2002). Thus the fundamental developmentalist logic of decentralisation is that local institutions can better discern local needs and aspirations and are therefore more likely to respond to them adequately.

The reforms leading to local governance are consequently seen as providing a structural arrangement through which local people and communities can participate in the fight against poverty at close range, since the ultimate objective is to institutionalise grassroots participation in the policy processes. Participation of the grassroots is believed to make development plans and interventions more relevant, give people more self-esteem, and help to legitimise the development planning process and the state machinery as a whole, since communities are given the opportunity to influence the resource allocation processes which have a direct bearing on their livelihoods (Boone, 2000; Kauzya, 2000; Heller, 2001). This ultimately creates an atmosphere in which various segments of society work collaboratively to alter their pattern of livelihood at least according to shared hopes, fears and aspirations. In other words, decentralisation heralds a permissive and enabling atmosphere for communities to effectively realise their full potential for dignified and fulfilling lives, as it is not only an institutionalised but also a legally underpinned form of participatory development (Mikkelson, 1995; Varshney, 1999; Chinsinga, 2003).

The advent of District Assemblies (hereafter referred to as DAs) in Malawi is very much linked to the phenomenal political, economic and social transformation that the country has undergone in the recent past, following the collapse in May 1994 of Dr Banda's Malawi Congress Party's (MCP's) three-decade authoritarian one-party rule. This not only marked a disjuncture from the official rhetoric of institutionalised denial of crippling poverty but also stimulated unequivocally critical policy debates pertaining to poverty for the first time since independence in July 1964 (Anderson, 1995; Chinsinga, 2002) and eventually culminated in the United Democratic Front (UDF) government adopting poverty reduction as its operative development philosophy (GoM, 1995). This government has thus clearly distinguished itself from its predecessor by placing poverty reduction at the centre of its overall economic and social agenda (Harrigan, 2001; Chinsinga, 2004). In fact, the collapse of the authoritarian state machinery was sealed with the adoption of the liberal democratic constitution in 1995 which, inter alia, guarantees a full range of civil, cultural, social, political and economic rights, democracy, good governance and the rule of law. This constitution further recognises a viable system of local government as an integral building block of a functional and potentially vibrant democracy. Appropriate instruments, notably the new local government Act of 1998 and the 1998 decentralisation policy, were promulgated in a bid to concretise the fundamental ideals embodied in the liberal democratic constitution. The ultimate objective of these legal instruments is to build up local capacity and grassroots institutions and adequately institutionalised arenas of participation.

Given that Malawi's socio-economic profile is characterised by marked polarisation between the rich and the poor, there is little doubt that the reforms propagated by the new government have raised high hopes and expectations, especially among a populace reeling from three decades of one-party dictatorship, severe poverty and underdevelopment. The key question, then, is how have the decentralisation and institutional policy reforms spanning nearly a decade created an enabling atmosphere for sustainable development, governance and poverty reduction? This is certainly a broad question. To begin to answer it, this paper takes a critical look at the prospects of the newly inaugurated DAs for institutionalising governance and administrative structures that are responsive, democratic and capable of transforming the livelihoods of the poor on a significant and sustainable scale.

DAs constitute the focal point of the decentralised planning framework whose primary mandate is to coordinate the activities of various stakeholders in grassroots development at district level. Procedurally, therefore, any stakeholder in grassroots development is required to operate within the ambit of the decentralised planning framework, clearly stipulating its objectives, plans and operating procedures. This is considered particularly critical because DAs as planning authorities ought at least to know how much is being spent in monetary terms, and where the monetary resources are being distributed in a DA so that they have an idea of how people in the various areas under their jurisdictions are being serviced. This would thus also help them to direct stakeholders to development priorities as embodied in their respective development plans. The tendency of most stakeholders, especially non-governmental organisations (NGOs) and donors, however, is to circumvent the decentralised planning framework altogether, claiming that it does not have the requisite capacity to deliver. But this claim has turned out to be a strategically veiled pretext for insulating their institutional perspectives, ideals and agendas, to the detriment of the progressive development and institutionalisation of the DAs. The NGOs' and donors' main motive for shunning the decentralised planning framework is that they view each other, and the DAs, not as partners but as competitors in grassroots development. There is thus a competitive spirit, a sense of power struggle, whether real or imagined, among stakeholders in grassroots development (Chinsinga & Dzimadzi, 2003). Stakeholders usually strongly desire to be the only organisation within a particular field and to be able to demonstrate visible evidence of the impact of their activities within the shortest possible time – if not at a stroke. Thus, while there seems to be reasonable consensus about the desirability of a decentralised planning framework as a vehicle for sustainable governance, development and poverty reduction, the key stakeholders are, however, principally driven by the overwhelming desire to protect, advance and proclaim their own perspectives, ideals and agendas, including prejudices, at whatever cost.

The main argument of this paper is that the decentralised planning framework should be understood within the context of the principle of subsidiarity. This entails defining rules for cooperation among stakeholders at various levels of society (Ikhiri, 2001; Scholz, 2001). The issue here is that the activities of various stakeholders should not, in so far as local governance and development are concerned, be competitive, but rather complementary. Seen this way, the decentralised planning framework would largely underlie the broad parameters geared towards creating unity of direction among stakeholders in grassroots development without necessarily undercutting their own institutional perspectives, ideals and agendas. There could, for instance, be a whole range of satellite organisations working on flexible terms with communities but tied in one way or another to the decentralised planning framework so as to ensure bureaucratic accountability and continuity. This would, however, be achievable only if there was coordination and partnership among the stakeholders and a shared commitment to making the decentralised planning framework work. Unless the deeply entrenched idiosyncratic tendencies among the stakeholders are effectively tamed and genuinely subordinated to the underlying noble cause of the decentralisation policy and institutional reforms, the trinity of good governance, development and poverty reduction will remain a virtually unattainable ideal.

2. District Assemblies in contemporary and historical perspective

DAs are broadly defined as decentralised political and administrative authorities with elected representatives of the people primarily entrusted with the tasks of democratising state power and advancing participatory democracy and collective decision making at the grassroots. Their overall mandate is to consolidate and promote local government development as a source of democratic and autonomous decision making at the district and local levels (Chiweza, 1998; Mbeye, 1998). DAs have been constituted under the auspices of decentralisation policy reforms whose objectives include the following:

• To create a democratic environment and institutions at the local level which will facilitate the participation of the grassroots in decision making and therefore consolidate the democratisation process;

• To eliminate duplication of institutions responsible for governance and rural development at district and local levels, with the aim of making the public service more efficient and cost effective;

• To promote accountability and good governance at local level; and

• To mobilise the masses for socio-economic development at the local level in order to help the government reduce poverty.

DAs have two organisational structures: political and management. The political structure is headed by a chairperson chosen from among councillors during their first meeting and the management structure is headed by a District Commissioner. The management structure of DAs is popularly referred to as the Secretariat and the political structure is known as the Assembly. The Secretariat is thus an organ of execution, while the Assembly is an organ of debate. DAs are thus integrated decentralised authorities combining the strength of deconcentrated line ministries and departments with the autonomy and revenue powers and functions of devolved democratic local government. All government ministries and departments represented at the district level constitute the Secretariat at least as projected in the policy statutes. Membership of the Assembly is made up of: 1) one elected member from each ward within the local government area; 2) Traditional Authorities (TAs) and Sub-Traditional Authorities (STAs) from the local government area as non-voting members; 3) Members of Parliament from the constituencies that fall within the local government area as non-voting members; and 4) five persons as non-voting members to be appointed by the elected members to cater for the interests of such special interest groups as the Assembly may determine from time to time.

The official adoption of the constitution in 1995 providing for decentralisation was not Malawi's maiden attempt at decentralisation, but rather the culmination of a series of efforts which began with the colonial administrators as early as the turn of the last century. However, the reform efforts that eventually resulted in the inauguration of DAs can best be seen as a continuous stream of transformative action that can be traced back to the second half of the 1980s. Suffice it to note here that Malawi became independent in 1964 with a vibrant local government system approaching almost the ideal but this was to change dramatically following the 1966 cabinet crisis (Baker, 1975; Kaunda, 1992). The statutory powers of the local councils were substantially circumscribed and progressively neutralised by the establishment of District Development Committees (DDCs) to anchor rural development initiatives in 1967. Local councils were, however, not abolished and this therefore meant that from 1967 onwards Malawi maintained parallel and unavoidably rival institutional mechanisms for the delivery of grassroots development.

The need for reforms to the DDCs, which ultimately laid the foundation for DAs, was first echoed in the 1987–96 statement of development policies (GoM, 1987). This statement outlined strategies to be pursued in various sectors of the economy in order to achieve satisfactory economic, social, political and cultural progress. The major criticisms of DDCs were that: 1) they were hardly serious forums; 2) they met irregularly; and 3) they merely produced a series of wish lists for development projects. These could only be implemented if they were in line with the priorities of the central ministries. However, the most important factor underpinning the need for reforms was that for the first time since independence in 1964 the government had at least officially acknowledged that Malawi's development track record, especially at the micro level, was grossly erratic, and inconsistent with the official rhetoric, which virtually denied the prevalence of poverty. The official position of Dr Banda's government was that as long as the people had clothes on their back, enough food to eat and lived in houses that did not leak, the question of poverty did not arise. The change in the official tone did not, however, come as any surprise, because the living standards of the country had plummeted following the cumulative adverse effects of the structural adjustment programmes (SAPs) adopted in 1981 in response to Malawi's chronic structural and fiscal imbalances as diagnosed by the International Monetary Fund (IMF) and the World Bank (Chilowa, 1999; Harrigan, 2001).

The most important catalysing factor for the reforms effected to the DDC institutional planning matrix was, however, the publication of the joint Government of Malawi/United Nations (GoM/UN) Situation Analysis, which, for the first time ever, systematically compiled the official statistics of the incidence and magnitude of poverty in the country. This, inter alia, strongly gave credibility to the contention that the DDC institutional framework was virtually incapable of facilitating rural development in an effective and sustainable fashion (GoM/UN, 1993; Kawonga & Lungu, 1994; UNCDF, 1994). More importantly, the joint GoM/UN Situation Analysis marked a critical turning point in the official rhetoric of Dr Banda's government, which either viewed poverty from a grossly minimalist perspective or denied its existence altogether. The findings of Situation Analysis pegged the prevalence of rural and urban poverty at 60 per cent and 65 per cent respectively.

It is against this background that the fifth GoM/United Nations Development Programme (UNDP) Country Programme specifically targeted the institutional policy reforms of the rural development management regime. Its objective was to achieve a complete overhaul of the management regime in order to make it responsive to local needs and priorities, as the people at the grassroots would be directly involved in the identification, design, implementation and management of projects and programmes. This would not only result in the equitable distribution of resources but would also direct resources to areas where they were needed most. Poorer regions would catch up with those that were relatively prosperous, and there would be equal opportunity for all (GoM, 1996; Oyugi, 1996).

The drive for renewed momentum in rural development culminated in the district focus policy initiative modelled on Kenya's experience in 1993. The substantial appeal of the Kenyan experience was aptly captured by Barkan and Chege (1989: 432), who characterised it as ‘offering much for those who wish to learn more about the prospects for, and limits to, decentralisation in Africa’. The aim of this initiative was to empower districts to plan and implement district-specific projects that would otherwise be under the jurisdiction of the central government; to make districts into focal points for planning and implementing district-specific projects, transforming them into major forces and instruments for the design of rural development (GoM, 1996; Chinsinga, 2003). This policy initiative was launched in six districts, designated as Local Impact Areas (LIAs), in order to ‘pilot participative, transparent, accountable and cohesive structures for planning, implementing and monitoring activities at district level’ (UNCDF, 1994: 8). These districts were Mulanje, Thyolo and Mangochi in the south, Mchinji and Dedza in the centre, and Nkhata Bay in the north.

Several major changes were introduced within the framework of the DDC institutional planning matrix. A District Development Planning System (DDPS) and District Development Fund (DDF) were established to allow districts to prepare and implement their own projects (Kawonga & Lungu, 1994; Anderson, 1995). The DDF entailed commitment by the central government to provide block grants to district authorities for small-scale development projects such as improvements to feeder roads and water supplies, or the construction of community centres and health clinics, or the creation of income-generating projects identified by district residents and their elected representatives. The office of the District Development Officer (DDO) was also created to strengthen the DDCs' planning and budgeting capacity – this office served as a secretariat and repository of information for development planning and implementation in each district. The rationale for the district focus initiative was that districts needed to become financially autonomous in order to assume a leadership role in local governance and development. Whereas before the reform MPs would devote considerable time to lobbying their pet projects at the appropriate ministry headquarters in Lilongwe, they were now forced to conduct a greater proportion of such activities back in their districts by fulfilling their statutory role as members of DDCs. This would improve the public's access to the locus of decision making, since district institutions would play a greater role in the policy-making process.

Perhaps the most significant change resulting from the adoption of the district focus initiative was the inauguration of the District Executive Committee (DEC), which comprises heads of government ministries and departments, NGOs and even donors represented in a district. The Committee's mandate was to serve as a technical or advisory body to DDCs (Anderson, 1995; Oyugi, 1996). Before this change, all heads of government ministries and departments and NGOs, in addition to political and traditional leaders, were members of the DDCs. The justification for the formation of the DEC was that the ordinary members were not in a position to effectively contribute to the deliberations of the DDCs since technocrats dominated, and in most cases completely overshadowed them in development proceedings that were supposedly to be legitimately theirs. The establishment of the DEC, therefore, meant de-linking technocrats from the mainstream DDCs.

The efforts to promulgate a decentralisation policy framework benefited from reviews of the previous failed initiatives, particularly the district focus policy initiative following a cabinet directive in 1994 (Oyugi, 1996; GoM, 2001). The reviews focused on DDCs prior to and after the district focus policy initiative. Four major weaknesses of the DDCs, and lessons learnt from them, were specifically highlighted. It was pointed out that: 1) DDCs had neither a statutory basis nor executive powers; 2) they controlled no funds and had no authority to enforce decisions except through coercion, since departmental members remained responsible to their own ministries; 3) the behaviour of civil servants at committee meetings (by withholding essential information) alienated most people, who had begun to feel that they were merely being used by the civil servants to legitimise decisions already made by themselves; and 4) they were subject to excessive political direction, to the detriment of their development objectives (Chiweza, 1998; Mbeye, 1998).

Even though the district focus policy initiative had in many ways revitalised and reinvigorated the DDCs, it was argued that its major weakness was that it did not entail any measure of devolution. District-level activities remained squarely under the total superintendence of the central government machinery. Furthermore, institutions established under the auspices of the district focus policy initiative were not anchored by any legal or legislative mandate, and emphasised development much more than governance and representation. This means that it was only after the advent of democratic rule in May 1994 that local government restructuring became democratised in terms of both process and objectives. These reviews were further complemented by study tours to countries such as Uganda, Ghana and Germany, which already have in place fairly well-functioning systems of local governments, and later followed by broad-based consultations with public and private stakeholders, political and traditional leaders and civil society whose outcome was the 1996 draft national decentralisation policy (GoM/UNDP, 1996). The national decentralisation policy was effectively adopted in October 1998. This paved the way for the enactment of the Local Government Act in December 1998. The Act came into effect on 18 March 1999. The distinctive feature of the 1998 decentralisation policy is that it provides for devolution. Building on the constraints of, and lessons learnt from, the previous decentralisation initiatives, the ultimate objective of the policy is to institutionalise real decision-making powers and authority in local jurisdictions with clear geographical boundaries, legal status and personnel of their own, to such an extent that a great bulk of their activities will be substantially outside the direct control of the central government. The local government will therefore no longer be a public body exercising delegated powers but rather a deliberative assembly with legislative and executive powers recognised by and embedded in the constitution.

Suffice it to note here that the DAs, which were formally constituted following the November 2000 local elections, retained almost wholesale the innovations introduced to the district planning framework under the auspices of the district focus policy initiative. The major change, however, arising from the constitution of the DAs was the merger of local councils and DDCs, which until then had essentially existed as distinct and parallel planning structures for grassroots development for almost three decades. The merger was imperative because it resolved an outstanding institutional anomaly that enormously constrained the potential efficacy of each planning and administrative apparatus. The councils were legally constituted and yet perpetually resource constrained; the DDCs were richly endowed with resources, yet without legal or corporate status (Kaunda, 1992; GoM/UNDP, 1998; Chinsinga, 2002). The activities of the DDCs were as a result well insulated from legislative oversight, or indeed any other kind of scrutiny – making them highly susceptible to abuse and patronage.

Following the merger of the councils and DDCs, an interim administrative structure made up of the offices of the District Commissioner, the Director of Administration and the Director of Planning and Development was created in each DA to oversee the implementation of the reforms in earnest. It is worth restating here that the underlying objective of decentralisation policy reforms in Malawi is to institutionalise participatory democracy and development anchored by a well coordinated and decentralised planning framework. The rationale for such a framework is that synergies between stakeholders create a favourable atmosphere for creativity and higher levels of productivity, as previously disconnected structures are brought together at the local level. This planning framework, which was essentially a bequest from the district focus policy initiative, was established: 1) to provide a comprehensive resource profile of the districts to enable orderly planning, implementation and management of their own development; 2) to provide a coordinated system of development planning at the subnational level; and 3) to ensure a systematic selection of projects for funding (GoM/UNDP, 1998; Gama et al., 2003).

To come up with a functioning integrated decentralised planning framework, each DA is expected to institute a data bank, which in turn serves as a springboard for developing Socio-Economic Profiles (SEPs) and District Development Plans (DDPs). The data banks are established on the basis of district-wide needs assessment using the participatory structures of the decentralised planning framework, which extends down to the grassroots. Coupled with macro-level district data, usually extracted from official statistics, the primary data gathered from communities is analysed and the product of this analysis is the SEP. The SEP sets out details of the status quo of development in the district, highlighting, among other things, its prospects and potential in various socio-economic spheres. The SEP consequently forms the basis for articulating a DDP, which ‘provides a comprehensive but inherently flexible, pragmatic and adaptive policy direction and guidelines for the development of the district’ (GoM/UNDP, 1998: 25).

With a functionally robust and integrated decentralised planning framework, DAs would therefore not only coordinate district-level development initiatives with considerable facility but also be in a position to advise development partners which sectors and what areas in the district deserve priority consideration. The planning framework would further provide unity of direction and ensure that duplication in development initiatives is minimised as far as possible. However, the possibility of creating a functionally robust and integrated decentralised planning framework is beset by some problems which, as further explained below, underlie the politics of capacity. The problem of poor capacity on the part of the DAs, this author argues, is flagged as a strategically veiled pretext to cover up for the deeply entrenched self-seeking tendencies and orientations driving most stakeholders in grassroots development in their relentless quest to shield their institutional perspectives, ideals and agendas and project them onto the communities at whatever cost.

3. The interplay of the politics of capacity in perspective

What the author refers to by the phrase ‘the politics of capacity’ is the tendency among stakeholders in grassroots development to circumvent the decentralised planning framework, as explained above. According to Chinsinga and Dzimadzi (2001) and Dzimadzi and Chinsinga (2003), most stakeholders do this because they have widely differing goals, strategies and mandates which they are hardly willing to reconcile for the sake of this framework. Under pressure to achieve demonstrable results, they undercut the framework in order to project their own institutional needs, perspectives and ideals onto rural communities, and they justify operating outside the framework on the pretext of expediency, efficiency and effectiveness, arguing that the framework is unduly constrained by, and often trapped in, frivolous bureaucratic procedures, besides having dubious capacity to deliver. This entrenched practice on the part of most stakeholders has in various ways been responsible for the rather limited success so far of efforts to progressively institutionalise the decentralised planning framework. The stakeholders' obsession with their own institutional needs, perspectives and ideals, coupled with their operating strategies, has more or less disempowered the DAs and restricted their capacity to do good and to develop a positive role and legitimacy of their own. The rest of this section presents some examples which illustrate the way the enduring politics of capacity among the stakeholders in the political and administrative reforms has undermined the activities of DAs.

While the NGO fraternity in Malawi and the world over is driven by the rhetoric of participation and empowerment of the grassroots in development activities, their working relationships are surprisingly adversarial and competitive. NGOs are thus not only strongly opposed to the decentralised planning framework but also intent on outdoing and outsmarting each other in their respective spheres of endeavour. In this struggle, financial inducements have become a major arsenal among NGOs in their attempts to project their institutional perspectives, ideals and agendas onto rural communities (Chinsinga & Dzimadzi, 2001; Dzimadzi & Chinsinga, 2003). Thus while the objective of the decentralised planning framework is to progressively create empowering circumstances by cultivating a spirit of self-reliance and agency among communities so that they realise that they are not merely passive victims but masters of their destiny, most NGOs are increasingly mobilising communities with the help of financial inducements.

NGOs make use of financial inducements to target either communities or government extension workers. They are compelled to motivate extension workers financially because in most cases they use them not only as an entry point but also as a link with the communities that they work with. This practice has created a thriving rural development market among communities and extension workers with the highest-bidding NGOs, especially on the part of the extension workers, getting their priority attention. This use of financial inducements has consequently diminished the appeal of participation in local government structures. Since the majority of rural constituents' livelihoods are precarious they readily make themselves available to participatory development initiatives mobilised by NGOs. The NGOs have thus overshadowed the local government structures because of their prompt feedback and a stream of benefits vested in communities, particularly those holding positions in project management committees. Given a choice between participating in NGO-led projects and participating in the development activities mobilised through the local government structures (the DAs), people would opt for the former because of the benefits associated with them.

This situation is encapsulated in Chinsinga and Dzimadzi's (2001) account of their encounter with the people of Njema village in Mulanje district. The highlight of this encounter was the villagers' comparative analysis of the NGO and local government structures, especially in terms of resources, feedback and impact. Using OXFAM as an example of an NGO which is very active in the area, they pointed out that NGOs are far superior to the local government structures since they are seen to be prompt and expedient in responding to people's needs. They argued that OXFAM not only responds speedily to their problems but also rewards at least in some way all those people involved in the development projects. It is therefore not surprising that they say: ‘We are very willing to work on these development projects [NGOs] since from our point of view it is not time completely wasted. For those in committees, the benefits are even quite substantial. They are exposed to several training workshops, which usually attract decent allowances’ (Chinsinga & Dzimadzi, 2001: 80). It must nevertheless be noted that training workshops offered by NGOs to committee members turn out to be regressive in most cases when it comes to fostering mobilisation and empowerment at grassroots level. Esman and Uphoff (1984: 249), for example, argue that:

The most prominent members are invariably selected and then given training and control over resources for the community, without any detailed and extended communication with the other members about objectives, rights, or duties. Creating groups through these leaders, in effect, establishes a power relationship that is open to abuse. The more training and resources they are given, the more distance is created between leaders and members.

The NGOs' track record was contrasted with the modus operandi of the local government structures. Taking the example of a school project, the people of Njema pointed out that it took almost three years for it to get off the ground. Within this three-year period, no effort was made to update them on the status of the project. Meanwhile, the bricks that had been moulded in anticipation of the project were damaged and the sand that had been collected was washed away. In a similar vein, the activities of the National Initiative for Civic Education (NICE) – a nation-wide NGO – have had a negative effect on the activities of the DAs. Likewise, NICE offers inducements in kind to communities participating in their civic education programmes. It has, according to Englund (2003), developed what it calls a ‘goat culture’ tradition. This is a team-building practice using local resources – preparing and eating food together while discussing issues. Thus, to mark the end of usually day-long civic education initiatives, NICE sources a goat for the occasion. The underlying rationale, however, for the goat culture tradition is to enhance patronage of its events.

Given these practices, the grassroots are apparently demoralised and reluctant to engage in local government development initiatives, not only because the implementation of the projects is uncertain but also because they have virtually no immediate material benefits. Because the majority of rural Malawians struggle for daily subsistence, they obviously prefer to make themselves available to development endeavours which yield immediate benefits. These tendencies have equally taken root among the local government workers at DA level. Dzimadzi and Chinsinga (2003: 75), for instance, noted:

Indeed, in all sites, participation in panel discussions for the impact assessment was threatened by lack of allowances. In Ntcheu, for instance, DEC members flatly refused to participate in discussions having established that the DA had no funds to pay them for sitting allowance and neither could we.

This thriving rural development market therefore makes it exceedingly difficult to mobilise let alone implement grassroots development projects through the decentralised planning framework. This has progressively deprived it of valuable opportunities that could have steadily supported its development as well as its legitimacy in the eyes of the stakeholders and, perhaps even more importantly, the people at grassroots level.

The same is equally true for the majority of donor agencies who, while strongly supporting the decentralisation initiative, often circumvent the evolving decentralised planning framework. The very same excuse of excessive bureaucratic delays is likewise flagged. Most donors argue that the decentralised planning framework does not yet have the capacity to deliver the desired strategic impact. They allude, inter alia, to: 1) excessive delays in decision making; 2) weak accountability and transparency mechanisms; and 3) high labour turnover. It is, for example, under the pretext of these considerations that the African Development Bank (ADB) has created a virtual impasse in the activities of Mulanje and Mwanza DAs. These DAs had more or less completed the preparation of their DDPs when ADB came along with an aid package to help them institutionalise the decentralised planning framework. But to benefit from this aid package the DAs had to come up with new DDPs altogether. The argument was that the DDPs they had formulated were not robust enough since, according to the ADB's diagnosis, these DAs did not have the capacity to come up with viable DDPs (Chinsinga & Dzimadzi, 2001). As such, these DAs ended up having a pair of DDPs – one tailor-made to the specifications of the ADB under the superintendence of their technical advice. The consequence of this prescription as a prerequisite for the ADB's funding package was that both initiatives stalled because ‘stakeholders and communities felt cheated because they had just completed a similar process which they were being asked to repeat when they were yet to experience the benefits or impact of the earlier process’ (Chinsinga & Dzimadzi, 2001: 70–1). These demands therefore effectively disempowered the DAs and the momentum that they had gained prior to the ADB's approach to finalise their DDPs was dissipated as a result of the need to restart the process.

Donors further justify operating outside the decentralised planning framework on the basis that it affords them an opportunity to carry out their activities without undue bureaucratic and even political constraints. Their concern is that the lists of projects based on field and desk appraisals are rarely adhered to. Instead, it is power politics that ultimately determine the way projects that have qualified for funding are actually allocated and distributed across DAs (Dzimadzi & Chinsinga, 2003). While these arguments are to a certain extent plausible, they nevertheless mask the underlying credible reasons for ignoring the decentralised planning framework. The fact is that most stakeholders, including donors, are intent on protecting, advancing and satisfying their own interests. In fact, according to Dzimadzi and Chinsinga (2003), donors would like their specific labels attached to development programmes and poverty reduction initiatives that have directly benefited from their resources. The pool nature of the DDF funding mechanisms makes them reluctant to channel their resources through the decentralised planning framework because they fear they may not get credit at all for development interventions that have massively benefited from their resources.

The issue of stakeholders shunning the decentralised planning framework is particularly critical for the Malawi Social Action Fund (MASAF) since massive resources are being channelled through it primarily to cushion the cumulative adverse effects of the SAPs and foster grassroots development. This fund was launched in 1995 as one of the policy instruments to anchor the Poverty Alleviation Programme (PAP) billed as a ‘community led and partnership driven poverty reduction instrument focusing on developing local community assets’ (Dulani, 2003: 3). The MASAF has three different programmes: the Community Sub Project (CSP), the Public Works Programme (PWP) and the Sponsored Sub Projects (SSPs). CSPs are essentially demand driven, PWPs are the responsibility of the DAs, and SSPs target vulnerable segments of society through NGOs or Community Based Organisations (CBOs). Quite strikingly, Parker and Serrano (2000) make a strong case for social funds to be integrated into the decentralised planning frameworks. Their argument is that if social funds operate outside the integrated decentralised planning framework they will be less efficiently allocated. Since development projects are considered purely on a case-by-case basis, it is almost impossible to fit them into a broader picture. They further argue that the massive resources channelled through social funds could help institutionalise and legitimise decentralised planning frameworks, especially at grassroots level.

However, the MASAF, just like the NGOs and donor agencies, does not work directly through the decentralised planning framework and like them it resorts to citing the questionable capacity of the decentralised planning framework to justify the way its operational procedures circumvent it. It argues that the decentralised planning framework would ‘replace collegial with hierarchical decision making’ (Van Donge et al., 2001: 23). To underscore the problem of the DAs' lack of capacity, the MASAF increased the number of zone managers for the CSP from nine to twelve and employed four zone managers for the PWP, which had initially been designated as the exclusive domain for DAs, and three regional managers for the SSPs. The MASAF also shifted the responsibility for the PWP funds from the DAs to its zone managers. The MASAF claims its actions are merely stopgap measures subject to review once the decentralised planning framework acquires the requisite stature and becomes fully functional.

Social funds which are not integrated into the decentralised planning framework are invariably a cause for concern, because this makes it almost impossible to achieve the goal of empowerment billed as the mainstay of the decentralisation policy and institutional reforms. Unintegrated social funds are procedurally basically one-shot games which offer virtually no chance for the beneficiary communities to build up skills, especially to learn how to bargain strategically with external agents (Evans, 1995; Tendler, 1997). The grassroots can attain these skills only if they are provided with an enabling and supportive bureaucratic programme environment underpinned by tolerance, credibility, predictability, and rights (Whitehead & Gray-Molina, 1999; Joshi & Moore, 2002). Social funds must thus be seen not merely as islands of reform and innovation but as an integral part of the wider public sector reform processes. There is thus need to routinise and formalise the process through which participatory inputs are translated into outputs. The argument is that participation can never be comprehensive or continuous, since grassroots capacity for participation is uneven and cannot be sustained throughout the planning and implementation process, especially if attempts to systematise the requisite institutional framework are virtually non-existent.

The decentralised planning framework potentially presents regular and predictable opportunities for citizen–state interaction, but if participation is to be more than a temporary component of projects or interventions, if it is to be generalised across space and time, then it must be properly and adequately institutionalised. The decentralised planning framework in fact underpins a representative and downwardly accountable form of local governance which, if properly nurtured, can at least effectively institutionalise sustainable forms of local participation. The appeal of the decentralised planning framework hinges on the fact that only state-supported institutions can guarantee participation on a sustainable basis, not merely because they are adequately institutionalised but also because they are legally underpinned. Popular participation can be significantly consequential only where the structures exist for meaningful and not just symbolic participation.

3.1 The problem of capacity: a genuine cause or a sheer pretext?

The author argues that citing lack of capacity on the part of the DAs as an excuse for sidestepping the decentralised planning framework is to a large extent ill construed. It is indisputable that the decentralised planning framework is weak. This should, however, not be flagged as a pretext for debilitating it even further. The majority of the stakeholders overlook the fact that ‘it is conceptually normal to start from the point of weak local governance capacity and work towards strong local governance capacity’ (Kauzya, 2000: 7). This, indeed, makes sense if one views development as a progressive and qualitative movement from inability to ability, from incapacity to capacity. The stakeholders are therefore hardly justified in holding back until the decentralised planning framework attains the requisite capacity to serve as an effective vehicle for grassroots development.

The grim reality is that if stakeholders continue to stand aloof, the decentralised planning framework will never acquire the desired stature. The stakeholders' argument is a chicken-and-egg one, ‘since more often than not it is only the pressure of [the implementation] of decentralisation which motivates actions necessary to improve capacity’ (Ribot, 2002: 42). Without such pressure, there is little opportunity for the decentralised planning framework to gain the experience needed to build capacity, and no opportunity for it to demonstrate that capacity has been gained – for surely capacity cannot be developed in the abstract. The stakeholders must therefore see the evolution of the decentralised planning framework as a process of capacity building of which they are an indispensable integral part.

The lack of institutional capacity should not be an insurmountable barrier to the use of the decentralised planning framework; rather, using the framework consistently would contribute to its steady development. The danger is that holding back for as long as standard level capacity is absent amounts to bringing the institutional transformation to a virtual standstill. The fact is that capacity cannot just be legislated into existence; it has to be steadily built up. In fact, the most optimistic way to look at it is that the expanding scope of responsibilities implied by the decentralised planning framework may ‘provide the incentive for public officials to invest in capacity building or seek creative ways to tap into existing sources of capacity’ (Litvack et al., 1998: 27) or in Ribot's words (2002: 42), ‘motivate the existing staff at the local level to recognise their own potential and demonstrate their real abilities’. The point here is that the DAs must at least be allowed to operate, make mistakes, learn from their mistakes and, more critically, establish sound and interactive relationships with the grassroots as well as with other stakeholders in grassroots development.

3.2 The proliferation of NGOs and grassroots development: a blessing or curse?

The number of NGOs operating outside of not only the decentralised planning framework but also the national budget has apparently substantially swelled in the last the four years following the suspension of the Poverty Reduction and Growth Facility (PRGF) macro-economic stabilisation initiative in November 2001 (Fozzard & Simwaka, 2002). This was triggered by overwhelming fiscal slippages which, inter alia, included: 1) diversion of donor resources to non-priority areas; 2) unbudgeted-for expenditures, notably external travel; 3) the disbursement of resources to the poor without a viable bureaucratic mechanism for accountability; and 4) a dramatic increase in official corruption and patronage.

The decision of donors to withhold budget support has led inexorably to the proliferation of international NGOs across the country, especially at district level. This is mainly the case because most donors, for example the Department for International Development (DFID), the European Union (EU) and the United States Agency for International Development (USAID), have expanded their establishments following the suspension of the budget support in an attempt to closely monitor the disbursement of their resources to the poor. The progressive expansion of their establishments notwithstanding, they are also sponsoring many of the NGOs, which, as a matter of practice, tend to operate outside the evolving decentralised planning framework. These NGOs boast of superior resource flows and hence have a magnetic appeal to the grassroots and yet their activities are largely transitory since they do not have any defined bureaucratic framework.

This explosion of NGOs operating outside the decentralised planning framework is a cause of concern mainly because the exact roles of NGOs in grassroots developments remain steeped in controversy. While Edwards and Hulmes (1998) argue that NGOs can be effective mechanisms for grassroots development if they adopt the three-pronged operative strategy of helping the poor to secure their livelihoods, attempting social mobilisation with the aim of empowering the poor, and trying to influence the overall political process by playing an advocacy role on behalf of the poor, the reality, however, is that most of them are far more concerned with service delivery, ‘at the neglect of the broader task of social mobilisation’ (Osmani, 2001: 138). The implications of this skewed emphasis are that NGOs provide pure benefits and not rights either in the moral or legal sense (Moore & Putzel, 1999; Joshi & Moore, 2002). In this regard, it is argued that it would be unusual to expect NGOs to ‘encourage or even tolerate the autonomous and potentially antagonistic mobilisation of their own client groups’ (Joshi & Moore, 2002: 49). There is also increasing evidence that the majority of the NGOs tend to be apolitical. They prefer to maintain a distinction between politics and development, which, in all fairness, is practically infeasible (cf. Englund, 2003). NGO officials usually contend that ‘NGOs and associations are concerned to create development through activities with the population whereas politicians are involved in politics which, according to NGO leaders interviewed, has nothing to do with development’ (Engberg-Pedersen, 2002: 168).

An issue of significant concern in the present article, however, relates particularly to the limited scope of NGO activities to improve the livelihoods of the rural constituents. These activities are small scale and dispersed, frequently experimental and flexible in practice, temporary and unstable and, more significantly, not enforceable in the same way as those run directly by government agencies (Osmani, 2001; Joshi & Moore, 2002). A large number of NGOs are susceptible to capture by elites at the expense of the grassroots they proclaim to serve. In Malawi, for instance, nearly all NGOs, especially within the realm of governance and service delivery, are essentially a network of returning political exiles that look at them primarily as a source of livelihood and not necessarily as a vehicle for promoting democratic participation and delivery of social services (Van Donge & Chinsinga, 1996). These NGOs are thus essentially characterised more by the logic of rent than the logic of productivity.

In view of these limitations of NGOs as actors in grassroots development, the author argues that it is imperative that they work within the decentralised planning so as to achieve legitimacy and facilitate bureaucratic continuity in order to secure long-term sustainability in the effort to transform rural livelihoods. If, indeed, NGOs could work within this framework, they could play a critical role in transmitting information about constitutional rights, potential allies of the poor, and political opportunities within easy reach of the poor, so as to harness their political capabilities and help them toward agency and self-reliance (Johnson, 2001). The observations made by Deolalikar et al. (2002) make the case for some kind of a workable relationship with the decentralised planning framework even stronger. Taking to the example of Bangladesh, which holds the Asian record for the number of NGOs relative to the population, they contend that ‘all NGOs taken together do not reach more than 20 per cent of the rural poor’ (Deolalikar et al., 2002: 33). Furthermore, the main defect of NGOs in this regard is that they are not legitimate objects of popular mobilisation in the way governments are (Moore & Putzel, 1999; Joshi & Moore, 2002). This dilemma is perhaps aptly captured by Moore & Putzel (1999: 5):

It is legitimate in most political systems to criticise governments for failing to provide for the poor, or for performing badly. These concerns can be framed in terms of an accepted discourse of rights. It is not possible to do the same thing in relation to NGOs. They are not authorities against which rights can be asserted. There is even need to ask who NGOs are accountable to, if indeed they are accountable.

Another cause of concern arising from the proliferation of NGOs, especially in the wake of the suspension of the balance of payments support, is that the dominant players in this process are foreign nationalities. This is a cause for concern because it substantially undermines the prospects for developing capacity for strategic policy analysis at the national and local levels. The failure of the majority of indigenous stakeholders to engage in informative policy analysis and advocacy, as is presently the case, often translates into lack of control over the development process. This is particularly critical for a country like Malawi whose public sector reform efforts are geared to creating viable DAs, which should assume a leading role in planning and implementing development initiatives. The proliferation of NGOs bolstered by enormous donor financial resources therefore risks marginalising and debilitating the capacity of the evolving local government structures. The planning framework, as has been envisaged in the decentralisation policy and institutional reforms, ideally portends an integrative avenue for at least systematically regulating the process of development at district level. The tendency to shun the decentralised planning framework has therefore led to more or less a state of institutional anarchy because a large number of donors interact with a variety of governmental and NGO institutions in all kinds of ways to alleviate poverty. The inevitable consequence is an enormous waste of resources because of unnecessary duplication of activities and lack of unity of direction. This makes the sustainability of these uncoordinated poverty reduction ventures doubtful. Of course the stakeholders ought to enjoy unfettered autonomy of operation, but it has to be within the framework of standards regulated by the central government, especially in matters of strategic importance in the overall scheme of national development.

4. Concluding reflections

The widely proclaimed problem of lack of capacity as flagged by most of the stakeholders notwithstanding, the potential of the decentralised planning framework should not be hastily discounted, especially given that development, especially empowerment, is a gradual and systematic process. It has been pointed out that this problem is used by stakeholders as a pretext for circumventing the framework in the pursuit of their own interests and because they wish to be the only actor within a particular sphere of grassroots development and to achieve immediate visible results. The downside of this pressure and impatience is that it invariably degenerates into stiff competition which forcefully promotes the practice of circumventing the framework, thus marginalising efforts to institutionalise the representative and accountability aspects of decentralisation policy and institutional reforms.

The major concern, however, is that most stakeholders are less willing to play a part in efforts to foster the DAs' progressive development of capacity. As mentioned, capacity cannot be developed in the abstract: DAs need incentives if they are to develop it. Moreover, implementing reforms of this nature takes time and stakeholders ought to be patient. There is obviously some learning to be done and adjustments need to be made, but this cannot happen if they do not support the process, as is currently the case in Malawi. The bottom line is that decentralisation reforms cannot be judged unless they are actually implemented and given time to take root. The DAs must thus at least be allowed to operate, make mistakes, learn from their mistakes and, more critically, establish sound and interactive relationships with various stakeholders, including the grassroots.

It might as well be argued that the way in which NGOs and donors are currently operating portends a novel institutional blueprint for decentralisation (cf. Van Donge et al., 2001); that a functionally robust structure of district administration may be a product more of organic growth and adaptation to changing environments than of legislative design. The intrinsic appeal of this argument, however, markedly diminishes if one takes into account the fact that NGO and donor projects have well-defined life spans and that their whole range of interventions subsequently becomes the DAs’ responsibility alone – and yet the DAs are shunned on the pretext that they do not have the requisite capacity to deliver. In addition, NGOs and donors operate on very small scales and their activities are not enforceable in the same way as those run directly by government agencies. The main appeal of the decentralised planning framework therefore lies in its ability to reach almost every possible corner of the country, unlike most of the non-state actors, whose spheres of influence are often localised within the confines of urban and semi-urban centres. Furthermore, the decentralised planning framework presents a legally viable and predictable framework for interaction between rural communities and external stakeholders within the sphere of grassroots development. In stark contrast, most NGOs and donors are reluctant to invest in institutional support, especially for the building of skills at community level which forms the basis for effectively mobilising and empowering the grassroots (Platteau & Abraham, 2002).

The reluctance of NGOs and donors to invest in the institutional infrastructure of grassroots development is a critical constraint because the potential success in mobilising and empowering rural communities is dependent on synergetic interactions between external stakeholders and the rural constituents. In the words of Evans (1995) as quoted by Johnson (2001: 14), the success of any policy, ‘requires a concrete set of social ties that binds the state to society and provides institutionalised channels for continued negotiation and renegotiation of goals and policies’. This provides an opportunity for external agents, which Platteau and Abraham (2002) characterise as institutional organisers, to get citizen efforts organised and sustain their involvement. The decentralised planning regime, properly constituted, therefore offers a framework within which such lasting and mutually cooperative interaction between the grassroots and stakeholders can be achieved, since its overriding goals are to build up awareness and knowledge among rural constituents about the potential of their livelihoods and how best to bargain with external agencies. The decentralised planning framework would therefore greatly assist in propagating an organisational milieu that would help the poor to mobilise effectively and organise themselves in the long term.

Decentralisation policy and institutional reforms are ultimately meant to reconstruct local polities and local political processes in which the grassroots are effectively empowered to demand services, transparency and accountability, among other things. Hence it is imperative to ensure that this is accompanied by the capacity of local governments to respond and deliver (Ribot, 2002; Romeo, 2002). If, therefore, local governments are bypassed and remain institutionally underdeveloped, as is increasingly the case in Malawi, there will be virtually no incentive for the grassroots to engage in the local political processes and to demand transparent and accountable management of local resources. People would simply say why participate in a process that cannot deliver? It is against this backdrop that the author argues that for the DAs to institutionalise governance structures and processes that are responsive, democratic and capable of improving the livelihoods of the poor on a sustainable basis, there is need for some kind of a workable relationship between the decentralised planning framework and other stakeholders in grassroots development. This would, however, not necessarily mean that the activities of the stakeholders would be rigidly dictated by the DAs. There could, for instance, be a whole range of satellite organisations working on flexible terms with communities but tied in one way or another to the decentralised planning framework so as to ensure bureaucratic accountability and continuity. But for this to happen the stakeholders must be willing to subordinate their seemingly idiosyncratic tendencies and inclinations to the underlying noble cause of the decentralisation policy and institutional reforms, otherwise the trinity of good governance, development and poverty reduction will remain a virtually unattainable ideal.

A vote of thanks is due to Christopher Dzimadzi and Chrispin Chikwama and the two anonymous referees for their insightful comments on the earlier draft version of this paper. The usual disclaimer, however, applies.

Additional information

Notes on contributors

Blessings Chinsinga

Currently a PhD candidate at the Centre for Development Research (ZEF), Department of Political and Cultural Change, University of Bonn, Germany.