UGANDA CASE STUDY
I. Introduction: A Country Snapshot
1.1 Politic History and Local Government
1.1.1 Recent Political History. Uganda became independent on October 9, 1962, as the primus inter pares of the three East African territories. The Pearl of Africa, the original objective of the railway that opened up Kenya as a by-product, comprised a set of traditional kingdoms inhabited by related peoples of the Interlacustrine Bantu, together with an ethnically divergent northern and eastern area inhabited by Nilotics and two groups classified as Nilo-Hamitic -- the Teso (second largest group in Uganda at independence) and the Sabaot. An unstable coalition of two of the three then-extant political parties took power under the leadership of Milton Obote. The coalition proceeded to come apart as Obote turned more to a socialist path to development, resulting in Obote's assuming a more and more autocratic style of government. In 197_ [?], when Obote was out of the country attending a Commonwealth Heads of State meeting in Singapore, his Army Commander, Idi Amin, staged a coup that ushered in a period of erratic, destructive personalized rule.
Ultimately, up to half a million Ugandans were killed in the excesses of the Amin period, the economy was left in complete shambles, and a further period of vacillating and unstable coalition (in which two presidents came and went in a matter of months) eventuated in the triumphant return of Milton Obote for a second attempt at governance. The second coming of Obote, however, was confronted from the beginning by the falling out of the allies from the liberation war and the return of the key military figure among them, Yoweri Museveni, to active military opposition. Thus, the five years of Obote II, as it is known, during which some efforts were made at restoring economic structure and forward movement, were punctuated by a bush war that was reportedly equally as costly in terms of lives as the Amin period. Museveni's forces ultimately triumphed, with the active support of the Baganda civilian population that surrounds Kampala, and who paid a very high price for this victory in terms of the destruction of most of the infrastructure of a once-proud kingdom, and the loss of another several hundred thousand lives.
The deaths and destruction rained on the Baganda by Obote's forces (nb. Obote's supporters argue at least half of it was occasioned by the Museveni bush fighters) significantly heightened a regional antagonism between the south and Obote's north. When Obote's mismanagement of the armed forces became too egregious, his army command rebelled and overthrew him for a second time in mid-1985. The ensuing six months saw a weak period of rule by the coup makers, both military men from the northern Acholi area (both named Okello) while an effort at negotiating an end to the "bush war" with Museveni was mounted under the auspices of Daniel arap Moi. Museveni calmly participated in the proceedings but continued to advance on Kampala, and ultimately entered the city in January, 1986, in defiance of the efforts at negotiating a cease fire and a peace.
1.1.2 New Institutions. A new institutional infrastructure for Uganda and a "ten point plan" for development were quickly adopted. From the beginning of the bush war in 1980/81, the areas conquered had been governed by local councils called "resistance councils" which took over the provision of local government services to the best of their ability. Upon gaining the seat of government, Museveni and his 27 original bush allies set about systematically putting resistance councils in place of the previous administrative infrastructure. The councils were hierarchically related to one another, starting from village level and culminating in the "National Resistance Council", the supreme legislative organ which ruled Uganda until the full return to a civilian parliament in June of 1996, based on the new Constitution of October 9, 1995.
1.1.3 Local Councils. The councils are basically related pyramidally, with the village council, LC1, comprising all adult members of the village and electing a nine person committee on the basis of direct, universal adult suffrage. This committee in combination with all the other village committees in a parish -- similar to a Kenyan sub-location, or a Malawian Group Village -- comprises the LC2, the parish council. It in turn elected a nine person executive committee, duplicating the offices of the executive at village level, to handle the administration of affairs and implementation of decisions reached by the council. The nine person executive, the LC2 Committee, in turn joins together with all other LC2 committees in a sub-county (similar to a Kenyan location, or a Malawian Area Development Council area), and becomes the sub-county council, or LC3. The LC3, like its lower level subordinates, elected a nine-person executive, the LC3 or sub-county committee, which has become a key node in the system after the streamlining done by decentralization and the new constitutional order. This committee is the key local government unit below district level, and is the obvious locus for much concrete planning and provision of capital infrastructure.
From the sub-county level the system diverges from the strict pyramidal model a bit. Each sub-county elects the standard nine person executive committee, which join together to form the county level (or LC4) council. However, this council does not similarly produce the committees that constitute the final body, the district council (LC5). Instead, the LC3s -- the sub-counties -- each send a member to the District Council, electing the representative from among themselves. The LC4 or county level each elect a woman representative to the district council. The district council thus comprises around 30 persons (up to around 50 in districts with many sub-counties) who have been chosen through elections by an "electoral college" comprising their colleagues at the next lower level in the system. The indirect nature of the electoral process has come in for substantial criticism -- the electorate for district councilors is a tiny fraction of the district's population, those who managed to get elected to the sub-county councils. There are many supporters of the system as well, however (although they predictably tend to be disproportionately those who have benefited from it themselves.)
The main changes in the system under the constitution adopted in 1995 were the singling out of the LC3 (sub-county council) and LC5 (district council) as the two main focal points for practical governance related to significant service provision. These were made legal entities. The other three levels -- village, parish and county (LC1,LC2, and LC4) are administrative units -- but are nonetheless politically important in that the indirect electoral system is still in place. However, the sub-county council chairmen are now to be elected directly, by universal adult suffrage, (this will happen finally sometime in December, after having been delayed fifteen months), as are the District Chairmen (this will happen mid-February). Finally, in a move to incorporate the Beijing recommendations on advancing women politically, one third of the seats on all local government bodies are mandated to be held by women. The local council elections in process presently will be the first test of whether and how this can be accomplished.
1.1.4 Local Government Responsibilities and Finance. Last, but most importantly, the Government of Uganda has since 1993/94 implemented a far-reaching decentralization, vesting responsibi-lities for most service provision with these district and sub-county councils and transferring to them all powers with respect to staffing. Finance was decentralized in phases, with 13 districts receiving transfer of funds for the main services from the Treasury (through Local Government) in 1993/4, 14 the following year, and the final 12 in 1995/6. A somewhat flexible set of common institutions is set out in the Local Government Act for district functioning.
Currently the districts receive around a third of Uganda's recurrent budget in the form of three types of allocation: conditional grants (tied), bloc grants (untied), and equalization grants (not actually functioning yet, but meant to reduce the disparities). They have as well local sources of income, primarily the Graduated Personal Tax (which was never abolished in Uganda). Nonetheless, between 60% and 80% of an individual district's budgeted revenue comes from central sources, and the actual collection rates of local taxes in the past two years have been disastrous, so that in fact around 90% for many councils is centrally transferred funding, which is both tied and unpredictable. In sum, this is the somewhat promising situation that the UNCDF project faces at inception. There is much general political commitment; there are many administrative pitfalls.
1.2 Demographics
Uganda's population is estimated at around 21 million presently, distributed unevenly between the more populous southern Bantu-speaking areas (the kingdoms), which comprise up to two thirds of the population, and the northern and eastern areas. Nilo-Hamitic groups inhabit the east (the Teso, Karimojong, and Sabaot), together with several Bantu peoples in the southeast; Nilotes inhabit the four northern districts comprising Langi and Acholi; and the far northwestern area, the "West Nile", is primarily the home of Sudanic peoples, with both Nilotes and Sudanics having closely related groups across the troubled northern border in the Sudan. Up to a third of the population of Uganda is ethnically Baganda, and the kingdom of Buganda occupies the south central area of the country. Buganda comprises 11 districts, the eastern region now has 12, the western region similarly 12, and the north 10, for a total of 45.
Population growth is high, topping 3% per year recently, despite the ravages of AIDS and the generally high levels of morbidity and mortality. It is impossible to put into words the impact of the AIDS epidemic on the social structure and institutions of rural Uganda (and urban, for that matter.) Funerals have become a weekly event for most people. There is no single family which has not felt the brunt of this scourge in one or another way --if not through losing members, through gaining them, i.e. the orphans of brothers and sisters, children, and collateral kin. The impact on productivity and income is severe. The absorptive capacity of the extended family has been stretched past its limits. While the phenomenon of street children is not as bad as Nairobi in numerical terms, it is growing daily, and the face of urban misery which was briefly invisible in Kampala (from about 1990 to about 1995) is now at center stage once again.
1.3 Economic Trends and Constraints
Uganda is touted as the economic miracle of the continent, especially given the depths of destruction it has confronted. Growth rates have been high and sustained for the past eight years, since the Museveni government changed direction (sometime in 1988/89) and set out on a course of structural adjustment and liberalization. The President is committed to this course and is, oddly enough, the "progressive" balance to a legislature which is more likely to tend toward market interventions to support politically sensitive groups. Asian properties have been restored to their owners, parastatals are being privatized, the IMF fiscal and monetary programmes are being pursued. The results are primarily visible in terms of the rehabilitation of commercial infrastructure in Kampala, the rehabilitation of the road infrastructure using donor funds, and the greatly increased marketing of local production, including local foodstuffs and light manufactures.
Unfortunately, some elements of the rosy picture painted in the western press are deceptive. A good deal of the increase in GDP over the past five or six years, according to expatriate advisors to the Ministry of Finance, is simply a measure of external inflows of capital in the form of donor project funding and economic support funds. The high growth rate is inevitable given the low base from which the country started to rebuild in 1988. In per capita terms, however, it is still the case that income has not reached the levels it had attained in 1970, prior to the Amin coup, and it is in fact very far from reaching this point.
Most analysts agree that rural Ugandans have not yet felt the benefits of economic rehabilitation and growth; household incomes have not increased measurably, and the most vulnerable groups -- rural landless households, female-headed households, northerners generally, esterners to a lesser extent -- are worse off now than they were five years ago because of the negative impact of structural adjustment and the curtailment of social service provision. (This is notwithstanding the onset of "universal primary education" for the first four children of a family, which has thus far been a dismal failure.)
Economically, then, the picture is not as bright for individuals as for the large-scale commercial sector in terms of the concrete results of market-oriented, structurally adjusted reforms. At the local level, revenues are derived primarily from taxes on individual income (i.e. Graduated Personal Tax) and they have been decreasing steadily over the past few years, compounded by the fact of three "election years" in a row (1994/5, 1995/6, and 1996/7) with the concomitant pressures to "forgive" local taxes. Local councils are thus in theory operating on a revenue base that is around 75% from central transfers and 25% from local revenue; in practice, both the central transfers and the local revenues have decreased, the former by a modest percentage but the latter by about half in many of the councils, leading to crises and inability to meet monthly wage bills on time in many areas. (Which in turn leads to the need to "borrow" from the teachers' salaries account to finance immediate expenditure needs, and other examples of "robbing Peter to pay Paul".)
This, then, is the economic context in which UNCDF is beginning to implement a Local Development Fund program. It is characterized by a fairly high level of political and institutional commitment and a fairly low level of resource availability in real terms, which might be quite the reverse of several of the other cases, particularly Asian cases. Sustaina-bility will be the key issue in Uganda.
II. Case Study Parameters and Caveats
A few caveats must be borne in mind in reading this case study. First, the program is just beginning. The documentation and recall of the formulation process is thus quite recent, detailed and enthusiastic, but the implementation process has yet to begin. Uganda has a way of taking much longer to implement projects and programs than anticipated, despite enthusiasm and intelligence in the formulation and design, in the author's experience. (For example, the constitutional review and drafting took two years longer than projected; all of the subsequent elections "slid" a further year or year and a half.) The project team needs to err on the side of caution in projecting time lines for achievement.
Second, the team visited only two of the four (now five) project districts, and these the two most accessible to Kampala and most incorporated into the market economy. The UNCDF project staff generously committed staff time to organizing a field visit to the new, fifth district now to be incorporated into the program, which is a very different type of district and one just at the beginning of formulation, rather than about to undertake implementation. The team felt, however, that a cost/benefit calculation of the very brief time available to it for field work argued in favor of discussions with more than one district now ready to undertake implementation, since the likely implementation issues the program would confront were stressed as the main focus in our review of terms of reference in New York. Mukono and Jinja districts being contiguous and close to Kampala made it possible to maximize field discussions even in our few days in Uganda.
Third, field interviews were limited to the district level, generally the council chairmen and a few key officers, rather than the sub-county level, where most actual capital project implementation will occur. One member of the team has, however, interviewed officials at the subcounty level in three other, non-project districts on issues directly relevant to implementation issues UNCDF is likely to confront, and their implications for a "scaling up" strategy.
III. Local Development Fund Stage and Progress
3.1 Timetable and Initial Focus. The Uganda team began planning in 1996, did baseline data work in early 1997, and then waited, taking advice from local stakeholders, for the local elections to take place, rather than begin building institutions and capacity with the current councils. The local council elections are just now being held, to be completed by mid-February, when District Chairmen will be elected on the basis of universal adult suffrage for the first time in Uganda, representing a step away from the indirect, delegatory approach that has been the government's favored political model, but which is seen by western governments as increasingly open to manipulation and monopolization by small numbers of the pro-NRM elite.
Pilot elements of the District Development Project (the name by which UNCDF's program is officially known) focus on the allocational formulae for district and subdistrict shares of the funding available, and monitoring and evaluation methods for the initiation of actual transfer of funds from district to subcounty level. This seems to have been well received and designed, and accords well with the approach and priorities of the relevant government agency, the Decentralization Secretariat of the Ministry of Local Government.
The team felt that the criteria for allocations and for transfer decisions were fairly clear, and we have made a fairly reasonable assumption that these were being well disseminated to the sub-counties who will have to demonstrate their utility. What was not as clear was that the baseline data on cost of transfer and administration/monitoring, which is not negligible in the context of Uganda's current capacity at subcounty level, is going to be captured adequately for the critical scaling up decisions that the program is meant to inform.
3.2 Project Management. The management structure comprises a district advisory committee with representatives from the four districts as well as the Ministry. This vehicle buys some commitment at the district level to decisions made centrally (since the districts are systematically involved), as well as excellent two-way communication. It is not clear that this is being internalized by the relevant government units as central to the functioning of decentralization; it may be that the relevant units have used this method previously, since a great deal of consultation at district level has been undertaken by the Decentralization Secretariat. Nevertheless, the management structure seems to the team to be one of the key areas in which lessons should be consciously learned and documented, with a view to their being proferred as institutional approaches to scaling up. This is not necessarily straightforward, since the main difficulty in scaling up is the need to deal with an overwhelming number of units all of a sudden, where previously a set of pilot units were easily managed within a flexible, experimental structure.
3.3 Potential constraints. Project management is heavy on two-way traffic involving the flow of resources and communications between the MLG/Decentralization Secretariat, the districts and the subcounties. It did not seem to the team that politically astute linkages to the Ministry of Finance and to Planning were in place, and there was some evidence that this could bring problems down the road, since the Ministry of Finance has not demonstrated overriding zeal in its commitment to the success of decentralization. Further, the success of the project as a pilot will depend on effective donor coordination, inhibiting the proliferation of donor projects in other districts which introduce other systems for allocation, resource transfer, planning and budgeting.
Some coordination is provided through the Social Sector donor group which meets monthly, but this can clearly not deal at any level of detail, and possibly not early enough in the process of donor project formulation to head off potential conflicts. The Ministry's response is a donor coordination unit in MLG, but this has so far been weak and not particularly focused on action, as opposed to data gathering. DANIDA is intending to address the problem, since they are funding this effort. It seemed to the team that the Project needed to develop close informal ties in to this unit, perhaps providing assistance in the form of a "template" of donor assistance drawn from UNCDF (and the parallel Bank-funded Peri-Urban Infrastructure Project.)
In summary, there is an unusually good opportunity for learning lessons, both for Uganda (relevant to scaling up) and for the wider development community (relevant to how to operate an experimental, pilot project for institution-building at local level through capital development funds). It was not clear to the team that enough attention is being paid to setting out lessons to be learned, capturing results, and taking maximum advantage of the amenability to this approach. The project staff avow this approach, but discussions with several of them (and with the Permanent Secretary and his very able staff) did not elicit concrete examples and the specific methodologies that are being followed. These need to be in place at this point, as implementation begins, if the lessons are to be captured in time to make them useful for reprogramming, rather than simply as a retrospective exercise. This is an analytic activity, half way between social science and management consulting. It would be intriguing to see whether the introduction of the methods of these fields into capturing lessons are feasible and useful.
3.4 Local Government Structure, Function and Issues
The local council structure and its origins were outlined above. The 1992/93 revision of the Local Government Act radically altered the relationship between district and central levels of government, transferring both responsibility and resources for most service provision to the district councils and to subdistrict councils as appropriate. The district councils were given complete authority for recruitment of staff -- hiring, firing, disciplining, promoting -- including the staff that were currently on central ministry payrolls, in 1996. A phased transfer of recurrent funding for the main services was begun in 1993/94 with 13 districts, continued with 14 in 1994/95, and concluded in 1995/96 with the remaining 12.
Recent changes have included the decision to focus on the subcounty and district as the key units for financial and legal responsibility, relegating the other three levels -- village, parish and county -- to administrative status; the decision to elect district and subcounty chairmen directly, on the basis of universal adult suffrage, through a secret ballot; and the decision that a third of all council positions must be held by women (n.b. the seats are not elected on a reserved basis; the mechanism for ensuring adherence to this principle is unclear).
Financial issues are the most critical ones facing the councils at present. Many councils have performed poorly since the advent of decentralization, with major scandals characterizing several (fortunately, not the pilot UNCDF districts, so far). Revenues have decreased rather than increasing. The implication for the District Development Project is that it will be especially important, and difficult, to build into the district and subcounty planning and budgeting processes the type of assurance of adequate recurrent funding -- operations, maintenance, supply, let alone set asides for depreciation that are logically part of capital development -- given the difficulty with day-to-day salary and other recurrent expenditures. Simply devising a model for estimating recurrent implications accurately, for several different types of capital project, and connecting it up in some reasonable and non-duplicative fashion with revenue projections, would be a major accomplishment and area for proferring useful lessons to the other districts (via the Ministry) and more widely than just Uganda.
3.5 Project Coverage
A key issue in pilot projects is how broadly to cast the net in terms of variance in existing economic and institutional capacity. Choosing too many districts, in order to permit good tests of the impact of variance along these dimensions, generally leads to an abandonment of the "pilot" principle and a de facto scaling up effort. Choosing too few may make it impossible to determine which factors really account for the variance in performance, since too many variables are in play at the same time. This may turn out to be the case with the Uganda pilot. The districts chosen are a good representative of the four de facto regions of Uganda (which, however, have no administrative manifestation); of urban/rural differences; of relative wealth and poverty; and even of degree of ethnic and religious heterogeneity. The pilot districts probably are weighted a bit highly on the side of income, but they seem to avoid an over-emphasis on the "politically advantaged".
Zeal in incorporating a nomadic northern district was probably misplaced (although the team had the impression the zeal was on the part of the government and was reluctantly entertained by the project staff). This could become a millstone, since it is a district not particularly representative of the rest of the north and logistically very difficult to service. The project staff might attempt to find another donor to take it over at some early point ( USAID is interested in "adopting" this district, Kotido, if it ever gets its act together, which is by no means certain or in sight.)
3.6 Project portfolio
It is not clear what the project portfolio will include; the project document limits the menu to the main "public goods" (schools, roads, water and health facilities). Recent experience with the investments that subcounties are making with the small funding available to them suggests primary school rehabilitation and refurbishments (staff housing, class room blocks, infra-structure); clinics; and administrative buildings. Income generating projects administered by councils are theoretically not eligible but this may be a mistake; it is not clear whether they will be entertained or will surface in the proposals, but market infrastructure and processing facilities would seem to be a feasible and sustainable type of investment in Uganda. Local councils have in many cases had good results with privatization and have been able to stabilize and greatly increase their revenues with such infrastructure, which may in the long term be a more important gain than extending the social infrastructure before they can sustain it.
3.7 Key Implementation Issues: Sustainability
The main implementation issue in terms of project portfolio is the close attention to staffing and the long-term recurrent implications of staff, supplies, and ultimately capital replace-ment costs. There may be, as in Malawi, a huge backlog of capital reconstruction and infrastructure provision which does not have salary implications -- indeed, which increases the productivity of existing staff -- housing for teachers, housing and better facilities for local clinical personnel, housing and structures for local magistrates and other administrative personnel.
On the other hand, there may be a tendency toward what Ng'ethe calls the "neighborhood effect" -- duplicating the efforts of the other local governments in the area; keeping up with the Jones's. This certainly had a powerful impact in Kenya on the harambee effort (before it became an extortionate, state-sponsored, regressive form of taxation divorced from the provision of local infrastructure), and also on the local development funds such as DANIDA's Rural Development Fund. The reduplication of facilities all of one type may have unsupportable short-term implications for staffing, and for longer term salary expenditure. This needs to be carefully monitored by the project and appropriate incentives for diversi-fication of project types introduced if it appears likely to be a problem. Further, depreciation of facilities, all of which have a concrete and relatively predictable life time, should be thought about if sustainability is to be taken seriously.
In view of the strong tendencies toward unsupportable, reduplicative social infrastructure, the team recommends that assistance in long term financial planning should be provided as an important adjunct to the efforts at building local capacity in planning. The team's experience in East Africa has been that the most serious capacity gaps are in budgeting and financial planning on a realistic basis, not in project identification, development planning, or even project costings and appraisals.
Implementation problems are likely to occur in two areas, the size and nature of local contributions, and the long-term sustainability of the infrastructure provided. With respect to local contributions, issues always arise over what counts and what doesn't, as well as whether there is good control over the tendency to budget local contribution accounts against multiple projects, or otherwise blur the commitment of local contributions to a single project.
With respect to recurrent implications, the identification of recurrent implications in the budget and the wherewithal to meet them is frequently very deceptive, because budgeted revenue is frequently wildly overestimated and under-collected. While central transfers are less vulnerable to gross fluctuations and unrealistic estimation, they have nonetheless been below what was printed in the budget document in the past two years. The Project Document indicates that costs of educational personnel are "to be met by the Central Government under UPE". Really?!!
Would that it were so. This is the sort of assumption that may signal trouble. Even now, last year's substantial increase in teaching staff is in the process of being off-loaded because the central government can't/isn't paying them. If the DDP funds 80% school projects in the four districts, with heavy increases in staff and consequent salary implications, is the central government bound to increase the salary grant significantly to these districts? Not likely; they'll tell the LC5 that these additional personnel are its own responsibility. The ability to budget realistically and on a long-term basis for the recurrent implications of capital development will be either the Achilles heel of this project or a signal accomplishment and contribution to the art of development implementation in Africa, depending on how seriously the issue is taken during the life of the project.
3.8 Size of allocations and disbursement issues.
The size of allocations will determine what the project portfolio will look like, rather than the schedules in the constitution that outline local service provision responsibilities. Large amounts of funding will precipitate larger project proposals -- eg. major road, bridge and water supplies -- while smaller ones will skew the portfolio toward school infrastructure and health facilities, as well as administrative blocs.
The project is in something of a bind in deciding it necessary to follow the statutory allocation formulae for transfer of central funding to the districts. This entails a 35% retention of funds at district level, a 65% pass-through to the subcounty level, and a 30% pass-through of that latter amount on down to the parish level. Taking the Project's first year capital fund budget and dividing it by the four districts presently at the planning stage (while recognizing that an equal allocation is not actually the procedure, the algorithm being partly population-based), this would yield an average of around $175,000 at district level, $22,750 at subcounty level and $1,625 at parish level (assuming equal district allocations, around 10 participating subcounties that meet conditions, and around 6 parishes per participating subcounty; all of these assumptions can be questioned, but the orders of magnitude become a bit clearer this way).
Is this a reasonable size for a capital development fund, and one that will induce the institution-building that the project envisions? It may be somewhat questionable. The district allocation might permit the financing of two or three reasonable projects, but it could equally be enough for only one "expensive" item. There will be extreme pressure to retain it for headquarters use of one sort or another, since this allocational formula has greatly reduced the amount retained at district level of the bloc grant from the center. In terms of the subcounties, there is a tendency towards equity in allocation decisions rather than rational, sustainable capital investment; one or two reasonable projects could be initiated with the funding available, but there may be a tendency to spread it far more widely and thinly, with consequent difficulties in monitoring and evaluating implementation and impact. With respect to the parish allocations, if they are indeed this small it is difficult to see how anything much more than the rehabilitation of one building or the grading of a fairly minor stretch of road can be implemented with this amount.
In turn, this allocation pattern is likely to lead to the tendency to comingle and re-package funding from different sources -- a universal tendency in rural Africa, and one which is actually an attempt to re-rationalize the funds that have been spread so thinly to meet equity concerns! This leads to major headaches for donors in establishing discrete accounting procedures and evaluating project effectiveness and impact. It should be observed closely by the Project staff.
With respect to disbursement, it is not clear what the "triggers" for disbursement after the first round will be, since the performance conditions have yet to be elucidated. The team strongly recommends that what is needed is a system which is not bureaucratically complex and incomprehensible to the subcounty councils, but which contains elements of flexibility that take into account the very different levels of performance among and between districts, and among and between subcounties within a district (related to economic performance and potential). This also will be a key issue in the pilot, experimental institution building effort -- how to design an accountable disbursement system which doesn't get bogged down in interminable delay because of complex, re-centralizing procedures of some monitoring system, but which also doesn't proceed to the state of misallocation witnessed in several councils since the onset of decentralization.
IV. Capacity Building
Capacity building is billed as a key element of the project. The strategy is generally to apply a portion of project funding to the institutional development needs that will produce an effective and participatory capital investment planning capability at district and subcounty level. Identification of those needs is yet to be undertaken so not much can be said about it. It was clear to the team that one immediate problem is the extreme lack of personnel resources the project faces. The staff we met with were of very high quality, but the task they face is rather daunting given their small number. This has been the case with the Decentralization Secretariat since the onset of this whole effort in Uganda. Similarly, while the Secretariat has undertaken major efforts in building capacity through short training courses at the district and subcounty level, they amount to a mere drop in the bucket compared to the needs.
The one recommendation the team feels confident in making is that the capacity building strategy cannot be designed in isolation from the construction of an effective monitoring and evaluation strategy for the main capital investment program. The latter should draw out the priorities for the former. Effective monitoring and evaluation is critical if any lessons are to be captured and made available to the Ministry, and the wider donor community, in support of a scaling up strategy. At the same time monitoring and evaluation should be neither exorbitantly costly (and thereby unsustainable after the completion of the project) nor unduly centrally-run (and thus a negation of the participatory approach to local institutional capacity building). A plausible strategy, therefore, is the design of a monitoring and evaluation strategy as the central, immediate activity -- and one involving the district teams/councils -- and the derivation of capacity enhancement needs in this process.
If this is not done it is likely that the capacity building portion of the funding will confront numerous, uncoordinated demands and may be allocated in less than optimal ways or ways for which it is impossible to demonstrate impact. Of course, it may also be the case that the Ministry may deliberately make some decisions about capacity building needs and funding in the experimental, pilot mode in order to test training procedures, modules, resources, or the like.
V. Scaling Up
Intentions for scaling up are clear; the project serves as the vehicle for testing procedures. There is a counter pressure at the moment, though, in terms of the need to identify sources of funding for areas which have been "left out". GOU tends to think somewhat in terms of finding donors for them as a main element of "scaling up". This doesn't necessarily bode ill for UNCDF, indeed it could give the project staff a useful bit of leverage vis-a-vis other donor initiatives and proposals, but it could cause unanticipated time demands on senior staff (particularly the regional adviser). Since the Ministry's new "Donor Coordination Unit" is approaching its task largely in these terms, and is likely to receive further assistance from DANIDA in sharpening its focus, it might be useful for the Project staff to maintain a close contact with this unit and to deflect demands of this sort to it, working through DANIDA where necessary.
In terms of a schedule for scaling up, it is too early to project anything concrete along these lines, but that there should be one as procedures are tested and either proved or rejected, or more likely modified, should be kept in mind. Uganda has a miserable record with respect to keeping to deadlines and schedules. The fact that it stuck to the phased decentralization over a three year period is a major accomplishment which should not be assumed to be easily repeatable. The idea of scheduling the introduction of new procedures should be periodically reaffirmed as appro-
priate in order to ensure that the project does have the intended impact, since this is one of the Government's stated objectives.
To expand on the issues involved in scaling up in the Ugandan context, the team feels strongly that actually getting the lessons learned, i.e. getting the institutional advances made adopted, costed, staffed, and organizationally serviced, will not come automatically. Accomplishing these goals will undoubtedly involve UNCDF in issues of personnel and organizational structure in MLG, and ultimately in larger issues of personnel, including possibly even a review of the negative implications of absolute district control over personnel. While donor involvement in host country personnel systems and policies is almost uniformly a mine-field littered with skeletons and completely counter-productive, it can frequently simply not be avoided. UNCDF should have a strategy in mind for dealing with this, whether it be through the provision of technical assistance in thinking through ministerial personnel structures, organization, or other modalities for providing technical assistance from MLG to the District Councils.
VI. Donor coordination
There is an unusual degree of cooperation among donors in Uganda. DANIDA did trailblazing work, the Bank and UNCDF have come in in tandem and very programmatically, and two or three newer donors are amenable to following an effort at building on what already ests (Irish, Austrians). The Donor Coordination Unit in the Ministry of Local Government is meant to tackle donor coordination on a systematic basis, but according to DANIDA is not doing so in as pro-active a manner as may be necessary. Donor coordination should clearly be a priority for the Regional Adviser, but he needs also to be transferring skills and information about it to local project staff. Given the Bank's central interest in the same portfolio of objectives as UNCDF, through its Peri-Urban Infrastructure Project, it may be possible to minimize the time demands of this activity by having a good, informal level of communications and understanding with the Bank officer responsible for PUIP and with DANIDA, who might both be better positioned to take the lead and spend the time necessary.
VII. Main Findings and Recommendations for Country Action
The team's main findings relate to the need to elaborate an organizational strategy for learning lessons. In particular, the project should:
1. Devise a methodology for capturing the costs of fiscal transfer, supervision, monitoring and evaluation in place such that the project can give useful advice on cost implications of scaling up to GOU.
2. Incorporate variables, such as the size of allocations as a factor in determining project portfolio, into the pilot projects.
3. Reconsider the decision to reject economically productive projects that can be provided on a sustainable basis (e.g. markets, milk coolers/storage, vegetable processing, coffee factories, butcheries, fish cold storage and landing piers, etc.). The long-term sustainability of social infrastructure depends on increasing local economic activity and consequent revenue potential. Of course, management of such activities by the local councils should be opposed, but Uganda has had some success with privatization of such facilities at local level while maintaining the fact of public ownership and in the process increasing and stabilizing revenues.
3. Devote more attention to the issue of sustainability of projects. The team found the two district councils visited convinced (one more so than the other) that no problem would arise about sustainability. However, one member of the team has repeatedly heard from council chairmen and Finance Officers that the funding available to operate and maintain local services is totally inadequate -- that what little local revenue is collected (a disappointing fraction of the revenue theoretically available from GPT) is wholly used on salaries for locally employed staff not covered under conditional grants. The extreme constraint on local revenue availability will have predictable effects: pressure to allocate funds to recurrent spending; to the completion of existing infrastructure (perhaps this is okay?); to small capital projects for which the recurrent implications are not clear or budgeted (or for which the same recurrent funding is budgeted against two or more such capital projects), and others.
4. Consider some modest assistance on budgeting and realistic, long-term financial planning; these, in the team's view, are the weakest link in local institutional capacity.