V. OVERALL FINDINGS
The major findings, conclusions and recommendations of the review follow.
1. UNCDF's Comparative Advantage. The pilot project approach and the ability to be experimental and to design procedures for use in scaling up strategies is a characteristic that distinguishes the local development funds program from other donors efforts. Local government institution building efforts should be systematically pursued, with the capital investment funds as the central and concrete inducement. Building effective local governance institutions should represent the main objective, taking priority over other general goals ( poverty alleviation, participatory project identification, and straightforward infrastructure provision). Maintaining the pilot project approach and resisting the pressure to become the lead actor in a scaling up to nation-wide implementation is important. Strategies for doing this should be formulated from the lessons learned to date.
2. Political Commitment to Decentralization. The two cases reviewed demonstrated widely different levels of political commitment that are likely to have implications for UNCDF. In terms of the "conditions" vs. "context" argument raging in UNCDF, (which as political scientists we cast roughly in the language of Plato vs. Aristotle, i.e. "ideal types" vs. "rank epiricism", the language of our youthful college careers) we come out somewhat equivocally on the side of context -- that is, somewhat Aristotelian.
We think many situations present scope for some degree of forward movement in building and institutionalizing effective local governance structures. Administratively-led institutional development is not worthless, although it is insufficient. An irreversible and concrete, demonstrated commitment to political devolution cannot be a sine qua non for undertaking a local development funds program, or UNCDF's portfolio will be very small indeed. At the same time, the refusal to even contemplate political devolution in concrete terms should perhaps be grounds for rejecting a program proposal. We think that rather than setting conditions, what UNCDF needs is good baseline work on the political economy of the countries it is considering working in -- a competent assessment of the context, rather than either a setting of minimum "conditions" in mechanical fashion, or the alternative commitment to working in any and all contexts. Some contexts are simply going to be unproductive; some "conditions" are going to be agreed to without any capacity or real commitment to deliver. A careful assessment of the political economy of central-local relations should form a part of UNCDF's pre-project appraisal.
3. Resource Constraints. Extreme resource constraints, particularly personnel of adequate skill and experience, are likely to constrain ability of projects to manage complex designs. Our review suggested that the Malawi project more clearly faces these constraints. The essential background work on fiscal procedures for an eventual, delayed decentralization were prioritized and set out some time ago, but the relevant governmental units are unable to devote the appropriate staff resources to these tasks. In the Ugandan case, the complexity of the project design suggests that a lot is riding on the ability of the Project Management Unit to design and implement a rather sophisticated monitoring and evaluation scheme. We were not persuaded that sufficient resources have been or can be brought to bear before the implementation phase begins.
4. Learing Lessons. We felt that both cases manifested an inadequately developed infrastructure for learning lessons. In the Ugandan case, this is something that needs to be an immediate priority; it is part and parcel of the design of the monitoring and evaluation system discussed above. But it is more than this, since lessons are always learned from unanticipated results, loopholes, errors, misunderstandings, and the creative efforts of field staff to deal with all of these. It is clear from responses to questions to other field projects (Mozambique and Palestine, as well as Malawi) that the main method for capturing lessons is through key stakeholder workshops and periodic staff meetings.
The team suggests that much more formalization of the lessons the projects intend to learn, the evidence that will be considered in drawing conclusions, and the time frame for the experiments would be appropriate to a more systematic pursuit of the experimental, pilot program effort we identified as UNCDF's "niche" in the small-scale capital investment universe. Paying greater attention to systematic monitoring and evaluation will assist greatly. Monitoring appeared to be conceived primarily as the periodic reporting of expenditure and of project physical completions, rather than the broader issues of local government institutional functioning, constraints, costs, unanticipated or new implementation procedures, and the like. These need to be captured and routinely transmitted to decentralization units and other key stakeholders if they are to inform scaling up decisions.
5. Participatory project appraisal/methods. We found a generally positive approach to participatory project appraisal and formulation methods. We also identified in one of the cases an insufficiency of resources/time in training staff to use such methods, while in the other there were indications that the initial participatory methods had been very intensive and time-consuming and that later phases of formulation found it necessary to curtail these. We felt there was insufficient thought given to the issue of the strategic points at which participation should be incorporated, the degree to which participation can raise unreasonable expectations for funding (especially in the context of a significant amount of competition among providers of small scale infrastructural grants, such as is the case presently in Malawi).
We are concerned about the degree to which community participation can produce conflict rather than consensus in the context of extreme resource scarcity. We also wonder about the degree to which participatory decision-making, in terms of project selection and prioritization, has some potential for conflict with local institution building on a representational, elective basis. In the ideal case, participatory appraisal should be about the transfer of participatory skills to elected representatives; we have insufficient evidence on which to base a determination of whether participatory activities to date have been undertaken with this objective in mind, or on a more fuzzy "community empowerment" model. Finally, the nature, composition and agendas of those "participating" need to be carefully explored to ensure that we are not making unwarranted assumptions about these.
6. Sustainability: Project and Program. Insufficient attention is paid to long-term sustainability and its local financial mobilization and budgetary implications. Providing capital infrastructure, especially in the somewhat narrow band that UNCDF' projects encompass (e.g., excluding most income generating infrastructure, although we have seen that this is not a hard and fast rule), while trying to work within existing local resource mobilization and financial planning and management capacities, could be very frustrating and disillusioning. We found project staff (and their host country counterparts) taking for granted the assumption of major elements of the operations and maintenance costs of typical types of infrastructure funded by the LDF programs.
We do not share their confidence in the long term. We felt the costs of the implied supervisory activities for both of the cases we reviewed were likely to be higher than anticipated, and that this was a central issue in determining whether the procedures being used by the projects -- whether they were successful in the pilot districts or not -- could be replicated. Several issues need further attention if sustainability is a central objective, as it should be. The main such issue is the long-term salary burden of staffing for any substantial increments to capital investment in social infrastructure such as education and health.
Country/Case-Specific Findings: Uganda
1. The "lesson learning", pilot nature of the effort is admirably supported by the government's objectives but appeared to be running ahead of project's capacity to systematize, sythesize, and package these lessons. The remaining period prior to implementation should be devoted not just to assessing the degree to which districts and subcounties meet the conditions set for funds disbursement, but to developing an ability to capture the experimental outcomes.
2. Commitment in the political sphere, even together with the firm and perceptive commitment of the key host agency bureaucracy, potentially can be overridden by unconvinced centralizing administrative organs -- the Ministry of Finance, Planning, or sectoral ministries with reservations about local capacity. The Danes warn of significant resistance in some of these quarters to consolidation of devolution. No doubt the continuing embarrassment of financial scandals at district level will assist them in consolidating opposition -- as has the very political "use" of decentralization by the NRM government to buy off or obstruct political organization at levels above the district, i.e. to inhibit support for federalism. UNDCF may need to contribute to defensive strategies to maintain the substance and the momentum of decentralization.
3. Implementation problems are likely to be design issues rather than training issues. For example, where control over resources and the need for accountability is concerned, it is not accounting skills that is the problem, according to field informants, but value for money audits, quality control over contractors, and other mechanisms for inhibiting the propensity of individuals to attempt to capture public benefits unfairly. Targeting expenditure for capacity building should take account of the need for assistance in designing quality control and audit procedures -- which have long term "overheads" implications -- rather than simply skills transfer. While training may be the solution proposed for most problems confronting effective local government performance in Uganda, the Project should take pains to determine whether it is really skills or systems design that is at the heart of the problem, and focus capacity building assistance accordingly.
Malawi
1. There appears to be plenty of scope for experiment and indications that a good amount of creative experimentation is taking place or is at least being proposed. There is much less clear evidence of the "lessons learned" strategy requirements outlined above -- institutional infrastructure and focus, systematic tests of hypotheses and indications of what constitutes adequate evidence to confirm or disconfirm a procedure, good synthetic abilities, and a real strategy of what to do with the lessons, such as who to target as audience, how to disseminate and package them.
2. Longer term political bureaucratic commitment is being held hostage to both political opposition -- or rather, a complicated geo-political fallout from the initial multi-party power configuration -- and severe capacity constraints. The current government would reportedly lose substantial local control to the opposition if local council elections were held now. Therefore devolution is still on hold in practical terms, despite a UNDP-led effort to bring the major stakeholders around a table to agreement. Further, some lead institutions (specifically, the Director of the Budget, the Accountant General and others) are not able to bring the necessary resources to bear to implement the decisions they themselves have taken that would drive the process forward in at least a technical, administrative mode.
There is substantial room for maneuver in the technical design of decentralization, but no sign of firm political commitment to the significant legislative changes that it would imply, nor to long-term devolution of adequate finances; nor is there enough in-house capacity, it is reported, to complete the technical design of the system in-house. The likely trajectory and time-table for devolution is appears quite precarious.
3. On the positive side, immediate institution-building goals are to some extent being realized -- DDOs are proving popular as a focal point for development efforts at district level, even with the Bank-funded social action funds that are disbursed directly to communities without going through the district. Linkages between districts, sub-district "area" level (similar to Uganda's sub-counties), and villages are being gradually but precariously built by the planning system. Bureaucratic delay may cause this to stall; nonetheless it is being used as the model for scaling up.
4. Planning criteria and the planning system seemed to us completely overdesigned from the point of view of the paper and pen requirements. They incorporate too many donor fads in the criteria for project selection (e.g. gender-based selection criteria, environmental concerns, poverty, etc.) and completely fail to address the inevitable and strong intrusion of political criteria. Field officers made clear that most of the criteria are ignored; equity of project distribution, local community contribution and ability of the district team to implement and supervise a project remain the chief prioritization indicators -- not different from the case with other funding. The 16 prioritization criteria are chiefly useful for defensive reasons when politicians attempt an unfair distribution of funds.
5. The team was somewhat taken aback by the the degree to which, in the field, the Local Development Funds projects are character-izable as "brick-led development" -- projects identified and negotiated for on the basis of the pile of bricks that is made available by the local community prior to project proposal or appraisal. Participatory planning is beginning to take on the character of an extortionate, narrowly-focused, capital goods provision; the ultimate equity of the community contributions is not completely transparent, and could in fact be regressive, if, for example, it is primarily already over-burdened women who are the main contributors of labour for the brick-making. This may be inevitable; it bears some thought.