UNDERSTANDING GOVERNANCE

Central aspects of governance policy suffer from a contradiction. Governance reform expresses a bold project of social engineering, based on a holistic vision in which all aspects of social life (within the state, the market, cultural relations and the family) are rendered akin to the free-market model (Harrison Reference Harrison2004b, Reference Harrison2005). But this project is constituted in a highly centralised and statist fashion, with little consideration as to how reforms might be realised within the complex and diverse societies that collectively make up most post-colonial African nations. This article is interested in the fortunes of the governance agenda in Africa, in terms of its realisation in policies that potentially have an impact throughout a country, encompassing a diversity of localities within a national space.

Governance reform requires us to look not solely at abstractions regarding the meaning of governance, but also at its practice. One of the main programmes to realise governance reform has been the Public Sector Reform (PSR). In Uganda and Tanzania, PSR has attracted considerable amounts of external credit, grants and technical assistance (Harrison Reference Harrison2004a). These external resources have gone a long way to producing programmes that have as their overall aim a complete reconstruction of the state, based on a range of assumed complementarities between the state and society. And yet, the policy documents for the programmes in both countries give no clue as to how they might be ‘rolled out’ or ‘embedded’ within society. This is highly significant, bearing in mind the problematic relations that post-colonial states have maintained with their citizenries.Footnote ¹

Furthermore, when interviewing government and donor personnel, it became clear that governance reform was in essence a conversation produced by a restricted group of people who had attained a significant degree of ‘enclosure’ – a ‘pocket’ of ownership (cf. Evans & Ngalewa Reference Evans and Ngalewa2003): perhaps a governance realm underpinned by resource inputs and new intellectual fashions. This ‘realm’ does not sit well with the broader normative thrust of the governance agenda, which has become increasingly ‘inclusive’ and expansive (Craig & Porter Reference Craig and Porter2003), relying on terms such as ownership, participation and accountability. The implication of governance is that a sine qua non of its effective realisation is a broad civic constituency of support (Williams Reference Williams1999). If elites matter, it is only inasmuch as they either engage with, mobilise or create broader liberal bases of support for reform. Within its own terms, it is an oxymoron to speak of top-down governance reform.

It is this context that makes the recent donor commitment to local government reform (LGR) interesting. Unlike the other main axes of governance reform – financial management, civil service reform, parliamentary reform/capacity building, data management, etc. – LGR brings the governance agenda closer to more substantive issues of implementation in the locality. It requires reform-makers and implementers to imagine how the state will act and interact with societies at the ‘local’ level. In this context, ‘local’ simply refers to interactions outside the governance realm, and as such is a political construct that betrays the centralised nature of much reform practice.

And so it is within the emerging literature on LGR and governance that we begin to see more explicit formulations of the state–citizen interface and local state politics, beginning to construct clearer expectations of governance reform as a widespread and socially embedded phenomenon. The key premise behind LGR is that by localising governance reforms, citizens will identify with their purposes and consequently contribute to the progressive transformation of the local state. In a phrase: ‘sub-national governments are said to be closer to the people, have good access to local information and understand the local context well’ (Smoke Reference Smoke2003: 9).

Let us break this down a little further with reference to Tanzania – our case study and a country that has taken on and to some extent integrated governance and local government reform. The governance/LGR matrix encapsulates four processes.Footnote ²

1. 1. By making local government more open and transparent, local authorities and their decisions will become more legitimate (PORALG 2000).

2. 2. As a result of the above, local groups will be more likely to support local government (PORALG 2000).

3. 3. By devolving power to the local level, local societies will feel more disposed to engage positively with the state (Ngwilizi Reference Ngwilizi2002: 6, 15).

4. 4. As a result of the above, local government will be subjected to checks and balances emanating from local interest groups and organisations, making it more accountable (Mushi & Melyoki n.d.: 1; Ngwilizi Reference Ngwilizi2002: 21; World Bank 2004b: 2).

It is largely through LGR that we can tease out the kinds of expectations that advocates of governance reform have with regard to the latter's prospects outside the governance realm of donors, workshops and air-conditioners. In other words, it enables us to focus centrally on the ways in which governance reform relates to social relations beyond the realm of donor and high state officials.

A GOVERNANCE REALM IN TANZANIA

Tanzania presents a country case study in which a range of governance reforms have been pursued with strong donor backing and a relatively high degree of central government (GoT) support (Harrison & Mulley Reference Harrison, Mulley and Whitfield2008). This is especially the case from 1995 onwards, when donors and key parts of the government elite recognised that a disposition had been attained within which donors and the GoT had constructed a common reform agenda (Helleiner et al. Reference Helleiner, Killick, Lipumba, Ndulu and Svendsen1995; Helleiner Reference Helleiner1999). From the mid-1990s, the GoT has implemented new systems of transparency (financial management, data management, expenditure tracking), public sector management (Public Sector Reform Programme), participation by the private sector (privatisation and contracting out) and new codes and procedures related to good governance (Ethics Secretariat, Codes of Ethics, Prevention of Corruption Bureau). Additionally, as part of the Poverty Reduction Strategy Paper (PRSP) process, Tanzania has undertaken two rounds of consultation with Tanzanian ‘stakeholders’ regarding general policy development. All of these reforms have contributed to the prosecution of the governance agenda at the central level, implemented against a backdrop of (relative) political stabilityFootnote ³ with prominent high-level support, making Tanzania a ‘good reformer’ in most international perceptions.Footnote ⁴

Tanzania has indeed made a remarkable political journey over the last decade. It is now on its second PRSP (created in 2005 and known by its kiSwahili acronym MKUKUTA). As a result, Tanzania is intensively researched, and the downtown areas of Dar es Salaam are busy with consultants, researchers and donors. It is easy to be seduced by the profusion of programmes, infusion of donor money and the hubris of meetings, workshops and social networks. Interviews in 2000 and 2001 revealed that the perceived success of Tanzania's governance reforms was intimately intertwined with the construction of a quite closed and integrated governance realm, in which a certain kind of discourse – underpinned by substantial flows of resources – allowed for a high degree of ‘self-referentiality’ within what Gould (Reference Gould, Mosse and Lewis2005: 63; Gould & Ojanen Reference Gould and Ojanen2003) aptly describes as a ‘transnational aid domain’. By this, Gould means that programmes each employed similar terms as objectives and justifications, and that specific policies were articulated through similar procedures, especially logical frameworks and SWOTs.Footnote ⁵

Within this governance realm, it is often unclear to what extent specific reforms impact on their target populations, or the Tanzanian people in general; and importantly for this article, how reforms affect governance outside the central state. This point needs some conceptual elaboration.

Perhaps one of the most fruitful approaches emerges from a fairly new collection of writings within development ethnography and critical anthropology. A good example of this is Maia Green's (Reference Green2003) study of project management in Tanzania, in which the workshop constitutes the cradle for a ‘manageable’ development practice, which relies not on profound considerations of success or failure, but rather on the facility of ongoing practices and especially their conformity to global standardised models. Interestingly, she refers specifically to the LGRP: ‘Financed largely by multilateral and bilateral donors and designed by international audit consultancy, [it] essentially comprises a series of workshops through which local authorities are expected not merely to reform themselves, but to attain appropriate levels of competency for programme implementation’ (ibid.: 134). More generally, in the work on David Mosse (Reference Mosse2005; Mosse & Lewis Reference Mosse and Lewis2005) and others, ethnographies of professional development communities reveal how strongly issues of management and process influence the construction of acceptable understandings of development practice. Uma Kothari (Reference Kothari, Laurie and Bondi2005) describes how the profession of development practitioners has constructed a way of seeing and processing knowledge, and generally ‘ordering’ development, in ways which leave little space for ‘outsiders’ to influence development reform. This approach highlights how orthodox issues of policy success, learning and implementation constitute a governance realm in which procedures and languages are maintained within an integrated and introspective (transnational) development community.Footnote ⁶ These writers suggest that – as with any development policy or programme – governance reform has, in a sense, become its own raison d'être. Its impacts outside its own social and cognitive horizons may not be entirely disregarded, but they are not pivotal. In fact, perceived project failure might smoothly trigger commonly accepted procedures of review, which integrate very easily into the existing discourse and practice.

This approach is attractive because it engages with a remarkable social development that has taken place in states like Tanzania, in which government elites who are highly dependent on external sources of revenue develop close relationships with a range of donors and creditors, in the process ostensibly internalising a very powerful global ideology of development policy.Footnote ⁷ By highlighting the self-referential nature of this construction, our attention is drawn to the issue of what governance means for Tanzanian citizens in their specific localities in their encounters with the state, and how the state is changing at the local level as governance reform is ‘rolled out’ through decentralisation (cf. Peck & Tickell Reference Peck and Tickell2002).

It is important to understand the dynamics that shape bureaucracies that interact directly with citizens in specific locales, often under severe resource constraints.Footnote ⁸ These are precisely the conditions within which LGR is being carried out in Tanzania. This line of enquiry also fits well with the kind of critical anthropology just considered, because it gives us a heuristic distinction between abstract or centralised reform agendas, and the negotiations within public institutions that take place at the local level in order for them to ‘make sense’ in specific places. Consider the following passage (Mosse Reference Mosse2005: 79):

This excerpt highlights the issue well: how do policy initiatives produced within a centralised ‘governance realm’ (design fictions) then become mediated and in some sense implemented by officials and politicians at the local or ‘street level’?

LOCAL GOVERNMENT REFORM IN LUSHOTO

Lushoto is located in Tanga region, a day's travel from Dar es Salaam. Although not an especially remote District compared with those in the centre and south of the country, it raises issues that are common to the districts more generally: concerns about (perceived) remoteness, a desire by some to tap in to resources that are retained at the Centre, a desire by the young to migrate out to the regional capital or to Dar, and a fluid sense of Lushoto's intrinsic identity within a broader nation.Footnote ⁹ Any generic policy innovation which aims at ‘the local’ must necessarily negotiate this set of relationships between the central state and the districts, or as I have formulated it, between the realm of governance and the street level.

The importance of this was directly apparent at the start of fieldwork. Speaking to people in kiSwahili, largely with a research assistant,Footnote ¹⁰ one of the preliminary tasks to negotiate was how to define ‘good governance’ in a vernacular suitable for Lushoto.Footnote ¹¹ A few of the people we interviewed used the term ‘governance’ as part of their kiSwahili, and some preferred to speak in English, but the majority of interviews involved a period of conversation in which we explained what we were interested in before coming to a workable understanding of governance reforms, which were commonly denoted as maboresho (cf. Mushi & Melyoki n.d.: 19). This kiSwahili word loosely means ‘improvement’ (stemming from bora, meaning good). The phrase emerged during interviews as the nearest synonym that we could establish for the profoundly Anglo-American term ‘good government’. Maboresho might connote a general improvement or progressive reform in local government; it attached itself to a variety of instances of reform, for example the Local Government Reform Programme, Universal Primary Education, the Tanzania Social Action Fund or various capacity building programmes. The term represented a generally held sensibility amongst the District eliteFootnote ¹² that the local political milieu had changed as a result of a series of recent and interconnected central initiatives, and that this change had generally led to a period of intense reform activity and a perceived improvement in the District's politics.

But the generally positive views that the District elite held of the period from 2000 did not straightforwardly lead to a sketch of successful reform implementation. Nor did it imply that the kinds of state–citizen interaction projected within the text of reform documentation were being realised. Rather, maboresho – rather like the articulation of the word itself – encapsulated a localised negotiation of external initiatives.

The rest of this section outlines a characterisation of maboresho in Lushoto. This form of politics is crucially underpinned by an infusion of resources from external sources: the Ministry of Finance, line ministries and international donors. These resources take the form of financial transfers, investment in infrastructure and a range of training packages which often involve workshops in the regional capital city, or in Dar es Salaam.

DEVELOPMENT POLITICS

One of the cornerstone phrases of Tanzania's post-colonial politics is maendeleo (development) which, along with ujamaa and kujitegemea, constituted Tanzania's brand of nationalist developmentalism.Footnote ¹⁸ The meaning of maendeleo has been differently interpreted (Mercer Reference Mercer2002; Nyerere Reference Nyerere1974), but one key political practice that it has encapsulated is the political advocacy of local representatives who aim to capture resources from higher levels of government for their constituency. The ability to ‘bring development’ to one's home area provided a way of shoring up legitimacy both during the single party period and into the present day.

In present-day Lushoto, there are 32 constituencies, each with a councillor who represents his or her ward in the District Council (DC), which is located in Lushoto town. Councillors are elected every four years through universal franchise which runs along with the national elections. Since 1994, candidates can be supported by any party.Footnote ¹⁹ Additionally, as a result of national legislation, there are also eight ‘special seats’ in the DC reserved for women. In Lushoto, all elected councillor seats were retained by men, and all special seats were held by women.

Throughout Tanzania's post-colonial history, councillors have needed to maintain legitimacy by being ‘development advocates’ on behalf of their constituency, even if the control and location of any development resource within the ward might be a moot point. Since political and economic liberalisation, competition to be elected as councillor has intensified (Kelsall Reference Kelsall2000; Kiondo Reference Kiondo and Gibbon1995). Furthermore, one of the first successful measures taken as part of Tanzania's LGR was to reduce the power of the regional commissioner to allocate resources, and to devolve resource allocation down to the District level and to the District Council in particular. As a result, the District Council has become a key focus for those in pursuit of developmental legitimacy. So, although not an entirely new political habitus, developmental politics in Lushoto has been both modified and reinvigorated by maboresho. In the words of one councillor, there has been a ‘development wave since 2000; all councillors want a larger portion of the development budget going to their ward’ (Tito int.). Let us see how this has happened.

As already noted, decentralisation in Tanzania has been based on the Districts, which are expected to play a pivotal role in terms both of resource management and of participation. This is not only to say that councillors represent the wananchi Footnote ²⁰ in their ward, but also that the DC should work as a more active influence on the District administration. One can see this formalised in the changes made to the routines of full DC meetings. The minutes from full Council meetings in Lushoto demonstrate how the DCs have developed detailed agendas, have attempted to compel various departments of the District Administration to make reports to Council,Footnote ²¹ have created sets of focused minutes and have integrated reports in the full Council from various Standing Committees (which are legislated for nationally). The reinvigorated Council meetings contrast with a general view of meetings before maboresho in which an entire day – and perhaps even part of the night – would be spent discussing issues endlessly and without structure or agreement. Formally, now, the District Administration and the Executive (embodied in the district executive director or DED) should report to the full Council, which acts as the paramount decision-making body.

The revival of the DC has produced a routinised opportunity for councillors to lobby and advocate for their own wards. All of the councillors interviewed spent some time explaining that one of their key tasks as local politicians was to argue their case in full Council meetings for support to schools, health care, road improvement, irrigation and the improvement/construction of wells.

A second centrally legislated process has reinforced Lushoto's developmentalism, which relates closely to the new planning procedures mentioned in the previous section. The District Development Plan (DDP), which constitutes the key document to request resources from the GoT, has a process conditionality built in to it: it must be the product of ‘participation’ throughout the District. This is carried out through a hierarchy of meetings, from the Village Executive Committees, to the Ward Executive Committees (which are chaired by councillors as a result of the Local Government Act 1999), and then aggregated, through a process of prioritisation, by the Planning Department of the District Administration. The DED then presents the District Development Plan to a Full Council meeting. In effect, what is happening here is that Councillars and local representatives of the state (village executive officers and ward executive officers) draw up lists of developmental demands (almost always health and education), which they hope will make it through the opaque reworking carried out at the District level by the Planning Department. If these demands are not carried through into the DDP, councillors have the opportunity to pressure the District Administration during the reporting back to Full Council. Failure to do this effectively has a direct impact on the political cachet of the councillor: the VEOs and WEOs ‘just shoplist their problems, and expect us to deliver …’ (local government officer, in Braathen Reference Braathen2003: 6).

The importance of this Ward–District connection for councillors' attempts to represent themselves as successful development advocates was revealed in interviews with ‘special seats’ councillors, because their claims to represent a constituency are relatively weak. One of the main themes that emerged when women were speaking of their marginality as special seat councillors, instead of fully elected councillors, derived from the disconnected nature of their mandate. Special seats councillors did not represent a ward; rather, they were selected from a list according to what percentage a party had won in the District elections. Because CCM has always won all Council seats in Lushoto, the local CCM Secretariat selects the women to take up special seats after an intra-party election. The councillors then represent ‘women’ in general which, as an abstraction and as a category still subjected to the marginalisations produced by patriarchy, does not allow them to work nearly as well as development advocates. One group of women councillors complained that whenever a special seats councillor argues for developmental support for women in a certain location, other (male) councillors will rhetorically ask: ‘And who do you represent? What district do you represent?’ This suggests both that special seats have less claim to developmental resources, as it is unclear what geographical representation would underpin this claim, and also that there is an understanding that ‘women's needs’ are best represented by each district's own councillor. The disempowerment of women councillors is reinforced by the regulation that only ‘ward’ councillors can chair Standing Committees. As a result, women councillors have only chaired these committees as caretakers when the existing chair is ill or out of the district. As one woman councillor put it: ‘you need a ward to have a voice’ (Chamdoma int.).

VERANDA POLITICS

The developmental politics of Lushoto spills over into another area of District politics. This can be introduced through an observation. During the six weeks of fieldwork in Lushoto, we tried repeatedly to arrange an interview with the DED. This involved an almost-daily period sitting in the waiting room outside his office. It was readily apparent that the DED's office was busier than any other in the DA complex. The DED received a constant flow of Councillors who, according to the receptionist, were largely seeking an audience with the DED in order to request some form of District support for their ward. Officially, the DED has no discretion to act in this fashion, but the traffic of councillors suggests that, de facto, he did, and that this was a key way in which the DED assured himself of a substantial political presence in the District.Footnote ²⁶ His position as chief executive, chair of the Council Management Team, District accounting officer (which means that all Central funds are officially submitted through him), and chair of the District Tender Board, located him right at the centre of the governance/capacity building reforms. Because the District government is highly dependent on Central subventions (Steffensen et al. Reference Steffensen, Tidemand, Naitore, Ssewankambo and Mwaipopo2004: xi), the DED position is structurally highly significant, even if, formally, he is charged merely with ensuring the implementation of Council decisions.

What is striking here is the contrast with the open developmental politics of the NGOs and the Full Council meetings. The DED's office was a literal and metaphorical illustration of a more opaque and personalised local politics – a ‘behind closed doors’ set of negotiations and discretionary decisions. Although this form of District politics is the most difficult to research, it constitutes an important sphere of political activity, and raises questions about the ways in which governance reforms realise themselves at the local level.

There are no statistics and very little written information about corruption in Lushoto.Footnote ²⁷ Some Council minutes note a few specific cases of the embezzlement of public funds in order to report the disciplinary procedures that have been taken. In the District Commission, one commissioner remarked that small-scale bribes by those visiting District offices are commonplace (Stephen int.). The Prevention of Corruption Bureau had a District office and occasionally requested information from administrative offices. The DA had a ‘corruption officer’, but he was away from the District when we visited, and his office had no paperwork or computer; his office was open and unoccupied on the occasions when we passed by, which did not suggest that a great deal of significant activity was happening therein. Tenders and a code of ethicsFootnote ²⁸ were advertised on the District noticeboard, and councillors have (as a result of national legislation) been banned from participating in any tenders. Overall, it appeared that anti-corruption measures were being followed procedurally, and that small-scale bribes were part of people's negotiations with local state institutions.

A distinction needs to be made here between ‘petty’ and ‘grand’ corruption. Three members of the Tanzania Women's Association told us that high-level corruption is part of Lushoto's mainstream politics. Tenders remain opaque in how they are awarded and costed, and some considered that the DED gave evasive answers to queries from the Council regarding tenders (Shemndolwa int.). It was an open secret that the election season in Lushoto produced a vigorous circuit of electoral largesse, in which electoral patronage would be generated and favours from the local government assured. This activity took place both within the CCM selection process and during the elections proper. Rumours abounded that it was during this period that the DED had assured himself of a new four-by-four. Interestingly, after my return from Tanzania (unsuccessful in my attempts to interview the DED), the minister for regional administration and local government publicly named the DED of Lushoto as responsible for the rampant misuse of funds, overexpenditure, opaque expenditure reports and illegal expenditure (The Citizen 3.3.2006).

However much of the radio trottoir (Ellis Reference Ellis1989) regarding corruption is sensationalised, we can rely on another observation to illustrate one aspect of the secretive and clientelist aspects of Lushoto politics. Lushoto's hardwoods are being illegally felled and evacuated from the District. I was shown two places where forest had been cleared, and the tyre-tracks of large lorries that led away from these sites. The sites were located in conservation areas. Simply in terms of its scale, this kind of illegal operation is too prominent to be invisible to the District Administration: logging is noisy, takes time and requires very large vehicles. A number of people in Lushoto, speaking in general conversation and without being asked for the information, referred to a specific local businessman as the entrepreneur carrying out the logging. Occasionally, people would hear the motors of large trucks passing through the district during the early hours of the morning. This was a secretive and illegal form of primitive accumulation that could only be carried out with the complicity of individuals who could ensure the indifference of the police or other law-enforcing officials.

Those who spoke about the logging all suggested political complicity, and it is difficult to imagine how it could occur without high-level support. It is simply not possible to bring large trucks into Lushoto DistrictFootnote ²⁹ without being noticed, far less to drive these through forest tracks, chainsaw trees down, load up the trucks and drive out of the District. I heard of accounts from reliable sources – including a development worker who was born and lived in Lushoto for most of his life – of villagers trying to arrest those felling the trees, only to be faced with police harassment themselves and the quick release of the culprits. We did not pursue how this entrepreneur had established influence within the local state structures, because of the sensitivity of the topic.Footnote ³⁰

The sporadic but frequent mention of corruption during interviews led us to conclude that ‘veranda’ politics was thriving in Lushoto (see Kelsall Reference Kelsall2002). It is not possible to define how Lushoto's veranda politics interacts with the other two forms of political practice. Whereas the governance and development politics are reasonably transparent and have some clear points of interaction, veranda politics is both more profoundly opposed to both of the other political forms, but must rely secretively on political support at the highest levels.

Governance reforms are ostensibly very concerned with corruption, but only through a specific problematisation of corrupt practice. Corruption is seen as a result of inadequate training, poorly designed incentive/discipline structures and a lack of transparency in decision-making. As a result, Tanzania has implemented a series of anti-corruption reforms, some of which address themselves to local government: improved salaries, new forms of invigilation, training and incentives, and declarations of wealth. These reforms have their merits, but it is worth reflecting on what they might mean in a situation in which local government is not homogenous but parcellised. Anti-corruption measures have rested on the assumption that corruption constitutes a deviant practice within a single (no matter how intricate) bureaucratic complex.Footnote ³¹ The sketch given above suggests that Lushoto's District government has an internal flexibility that allows it to negotiate these three repertoires of political practice. Interviewees were all aware of the ‘liveliness’ and complexity of district politics, embodied in the District Administration, the District Council and the District Executive Director.

CONCLUSION: STREET LEVEL GOVERNANCE OR LOCAL REALM?

It is striking how much of the reform activity taking place in Lushoto is governmental in its focus. That is, reforms have been oriented towards the modifications of the internal practices of the District, with a view to achieving the requirements of the GoT. Reforms in resource management have been internally oriented, rather than concerned with modifying the ways in which Tanzanian citizens might engage with the way the District government uses its resources; Councillors use their constituencies as a resource to be mobilised in order to capture resources from the District, but there is little evidence that this has led to a change in the way people in the villages and wards relate to the District Council; participatory research has been carried out, but the reports emanating from this reveal that people's participation has been entirely procedural – a more effective way to generate statistics rather than to empower or ‘conscientise’. This concern was expressed in detail by Lushoto's politically savvy district commissioner (Goroi int.).

Thus, what has happened so far as a result of local government reform is not clearly a rethinking of the ways in which Tanzanians relate to the state, but a drive to strengthen the political and administrative procedures of the District government. The semi-transparent clientelism of Lushoto's development politics, and the more secretive networks of twilight politics, can coexist with efforts at local administrative strengthening. Whether a revived role for the next administrative tier – the ward – will emerge as the LGRP progresses remains to be seen. An important facet of the answer will derive from how we understand the prospect for more profound change at the local level.

Veranda politics constitutes a political practice which is entirely ignored in the governance reform literature. Whereas it is recognised that councillors need ‘education’, district administrators need training and civil society will become more disposed to exercise pressures on local government, the ability of some local ‘big men’ to give or receive bribes and subvert the legal infrastructures of the District is not in any way amenable to the kinds of processes that are encapsulated by governance reform.

Thus, the reality of local-level governance reform is defined by a two-fold set of interactions. In the first place, a tension – but not an irredeemable tension – between Lushoto's capacity-building/resource management focused governance reforms, and its reinvigorated and electorally competitive ward-focused developmental politics. Because these two forms of politics are fairly transparent and rely on some shared discursive and normative aspects (notably over universal primary education, which everyone saw as a great success), this interaction does not augur that all governance reform will founder at the local level. Rather, its prospects depend on how these two realms interrelate in the future. The second set of interactions is between governance reform and veranda politics. Herein lies a more profound opposition. Veranda politics is extremely difficult to quantify and even describe in a comprehensive fashion; it may involve a small number of people in a bounded set of accumulation activities, which may come to an end with the depletion of certain opportunities for enrichment. Be that as it may, the governance reform agenda has no intellectual tools either to understand twilight activities or to engage with them.

In this respect, it is worth returning to the earlier characterisation of the governance realm. Here, a key motivation for agents within this realm was to maintain a set of practices and discourses that served to maintain an integrated form of politics; issues of policy success/failure were necessarily articulated within this motivation. In essence, the practice of governance reform is more important than any ostensible criterion of objective success. We may borrow from this approach to understand Lushoto's governance. Rather than search for a single coherent reform agenda that might negotiate Lushoto's internally differentiated local governance, we might consider how externally conceived reform agendas affect the facility of local government – that is, the latter's ability to maintain a workable set of inter-relations between its three distinct repertoires of political practice.

These comments are provisional, in the sense that local government reform has a long way to go. But, they do suggest that the success of the reforms depends mainly on the ways in which the institutions of the state at the District level reconstitute themselves to manage a national programme. The extent to which citizens, or local people, engage with this process seems to be limited to their ability to participate in ward development meetings. There are good reasons to assume that this form of participation is very limited (Shivji & Peter Reference Shivji and Maina Peter2000).