Project Evaluation Summaries - 1998
Prepared by the Policy, Planning and Evaluation Unit (PPEU)

---

Djibouti

---

Project Numbers:	DJI/86/CO4, DJI/89/005
Project Title:	Community housing in Balbala
Type of Evaluation:	Final
Government Executing Agency:	Djibouti District, District Technical Services (STD)
UN Cooperating Agency:	UNCHS- Habitat
Sector:	Human settlements
Sub-Sector:	Housing and equipment
UNCDF Budget:	$US 3,590,623
UNDP Budget:	692,600
Gov't. Budget:	1,002,200
Total Project Budget:	$US 5,285,443
Total expenditures at evaluation:	$US 6,975,049
< evaluation:	$US 4,179,061
Duration:
Date Project Approved:	April 1991
Date Project Began:	June 1991
Date of Evaluation:	December 1998

II. Background

Balbala, one of the three main areas of the city of Djibouti, is located on the left bank of the Wadi Ambali, on basalt cliffs. The quarter covers more than 320 hectares. In 1981 a master plan identified the quarter as an area for future urban development because of the land development taking place there as well as the existence of unplanned housing in the area. Balbala has undergone rapid expansion since 1982, and during the 1990s it has accounted for the largest share of the city's growth. Accordingly, the master plan accords priority to providing sanitation and other services to poor neighbourhoods as well as land development for the creation of plots in the Balbala quarter. In the early 1990s the population of the entire city was estimated at 240,000 inhabitants, with a population growth rate of 3.8%. The Djibouti authorities had built housing units for this growing population, but they were too costly for those with low or irregular incomes.

In Balbala residents frequently built houses on unregistered land that was considered part of the national territory, using unreliable materials such as wood and corrugated zinc roofing; they would then move in with temporary occupancy permits.

III. The Project

The approach selected encompasses housing construction, institution-building and a cost-recovery system that will make it possible to continue the project on a sustainable basis. The level of services proposed is compatible with the income of the target population.

Several formulas were adopted. At the technical level, priority was given to plots of land measuring 105 square metres, a graduated level of services and modest facilities and equipment. In financial terms this meant that, to meet the objective of gradually recovering the costs from the most disadvantaged population groups, costs attributable to the beneficiaries had to be limited to the cost of on-site installation of infrastructure, with an adjustment for plot size and location. During execution, the beneficiaries selected equipment of a higher quality than anticipated, which increased the costs attributable to them.

At the time the project was designed, a preliminary urban development programme was nearing completion. In essence, this programme involved restoring old quarters by installing service infrastructures, with housing construction left to individuals. The programme introduced a system for recovering costs and establishing a working capital fund that was adapted for the present project.

Development and immediate project objectives are to:

• Provide economical housing on developed plots for 1,390 low-income families in the Balbala quarter.

• Strengthen the institutional organization of the Djibouti district; this is necessary to ensure that a community housing programme will continue to operate.

• Provide technical and financial support to artisans and small construction companies for the production of standardized building materials.

• Improve the living conditions of the low-income population by providing adequate housing.

• Promote the establishment and continued operation of district-level institutions.

• Help create jobs.

• Promote the standardization of affordable equipment and housing for low-income households.

IV. Conclusions of the Evaluation Mission

The mission noted that the physical objectives of the project had been achieved. From the technical standpoint, the infrastructures and buildings were of high quality and adequately met the needs of the target group. Overall, the product was well received by the beneficiaries.

As for the sustainability of the project, a second phase can be envisaged only if the level of development is adapted to the income level of the target population, as originally planned.

The project had to be extended several times owing to the many external problems encountered. The magnitude of these problems and their impact on project execution could probably have been reduced by closer monitoring of subcontracted services, particularly those subcontracted to the Société Immobilière de Djibouti (SID) and to SCET/Tunisie, in terms of both the technical aspects as well as the organization of work and the timely application of certain coercive measures.

The transfer of skills was satisfactory. National team leaders and staff are now able to carry out the tasks assigned to them. However, their enthusiasm, which was augmented by supplemental cash payments from UNDP, may wane as the project draws to a close.

The termination of the service contract with SID and the follow-up to the studies and supervision of work by the project office made it possible to correct the situation. With the departure of the international experts and without the main technical staff of the District Technical Services, the new phase will require the services of an outside organization from either the public or the private sector.

The 10-year repayment period for infrastructure loans granted to beneficiaries of community housing plots made ownership possible for certain groups within the population who had previously been excluded from similar programmes.

Under the project some 50 workers were trained in construction; their skills appear to be in great demand among prospective builders both in Balbala and in other quarters of the city.

The project has also had a positive impact on women, who account for 30% of all direct beneficiaries. Women played an important role in construction, concluding contracts with labourers and overseeing the progress of work.

A. Evaluation of the results obtained

The construction of housing prototypes, which had not been anticipated in the project document, made it possible to evaluate different materials and to choose cement blocks, which proved more suitable than wood or brick.

Service levels in the district improved considerably in comparison with the emergency measures that consisted of developing the land and marking off plots. However, improvements to infrastructure during construction resulted in significantly higher prices for the beneficiaries.

The institutional framework of the project sought to ensure coordination by the district of the activities of participating institutions. The underlying principle followed was that strengthening the role of institutions would ensure that they continued their activities once the project was completed and that costs would be recovered. Strengthening the role of the district made it possible to identify the institutional foundations of a permanent community housing programme.

The institutional aspect of the project was well designed at the outset, but its framework was soon modified because of regional conflicts and a deterioration in the country's economic situation.

A computerized system for processing data relating to cost recovery was set up with support from the UNCDF and the UNDP, which are responsible for the salaries of the consultants recruited for this project. Owing to a lack of financing, it is not certain that this system can be continued after 31 December 1998. However, cost-recovery activities should be able to continue until 2009.

The criteria used in selecting beneficiary households, as set out in the project document, were well defined on the basis of 1991 infrastructure costs. Estimated costs for 1996 were revised upward. Given the system of calculated adjustments, price increases had the most serious effect on plots other than those designated for community housing.

B. Evaluation of the project concept

The system of loans in the form of materials and the issuing of a collective building permit proved to be entirely satisfactory for the beneficiaries, since they were able to begin building as soon as the plots were allocated and complete construction without delay, which was not the case with other plots, where little building was taking place. It also had an important catalytic effect: the construction of additional rooms and sanitary facilities was observed, which meant that the beneficiaries were making additional financial investments. It should be added that the UNCDF made it possible for materials to be purchased tax-free, further reducing prices. However, this financial arrangement did not lend itself to the notion of a working capital fund. Future purchases of materials with the amounts recovered will not be tax-exempt (unless the government decides otherwise). Accordingly, either the amount of materials must be reduced in order to stay within the limits of the loan package and the repayment capacity of beneficiary households, or the amount of the loans must be increased by about 19%, a step that may put them beyond the reach of low-income households. Nevertheless, this pilot programme showed that it was possible to set up cost-recovery systems for community housing projects.

The issuing of a collective building permit presents many advantages while freeing households from long administrative procedures. Nevertheless, the system should be improved if it is to achieve its primary goal, which is control of the land and utilization of the work prior to construction.

With the resources available for the launching of a possible second phase of the project and the funds recovered, it would be possible to continue the phase at least until 2009, provided that the current team remains in place.

The Land Allocation Commission has done its job well, but the list of eligible owners has become outdated owing to the delay between the time applications are filed and the time parcels are allocated.
 
 V. Recommendations

In the event that there is a second phase of the project, the evaluation mission recommends the following modifications:

• The Djibouti District should be the owner of the site.

• The Ministry of Urban Planning and Housing or a technical consultancy firm should be responsible for drawing up the plan and preparing the technical studies. In this case, the two institutions should coordinate their respective tasks under the project.

• In order to improve established financial procedures, an audit should be conducted. Contrary to the terms of the project document, no audit was conducted while the project was being implemented. To ensure the sustainability of the project, a clause should be added to the terms for the establishment of the revolving fund stipulating the purpose of the account and the conditions of its use. It is recommended that a committee be appointed to ensure the transparency of management.

• The District of Djibouti and the Djibouti Electric Authority should seek a solution to the problem of water supply. The project built eight public standpipes, but at the time the evaluation was conducted, only one of them was operational. Drinking water supply is a major problem, and improving the population's health depends on it.

• The price of land, set by the state, is too high. Prices should be reviewed and the budget line miscellaneous expenses, which the evaluation mission considers unwarranted, should be eliminated.

VI. Policy Implications and Lessons Learned

The decision to build prototype housing was made by the project office in response to the mission's conclusions regarding the use of wood and owing to the population's poor image of wooden housing, which had to do with its flimsiness. With these prototypes it was possible to side step a number of subsequent problems and, in particular, to confirm the cost of building a basic unit of housing and keep it within the limits of available resources.

The excavation plans referred to a volume of approximately 12,000 cubic metres for 677 plots. Given the relatively slow completion rate, attributable to the type of soil and the dry climate prevalent in Djibouti, it would be desirable, should the project have a second phase, to provide for smaller, and thus less expensive, foundations.

In terms of supplies, the use of group orders resulted in more advantageous prices and shipping costs. However, this arrangement involves administrative and warehousing costs that must be underwritten by the district or by the beneficiaries themselves.

In addition to bringing improvements to some 1,164 families, this project provided a development model whose infrastructure, collective equipment and housing construction standards can serve as a benchmark for future operations.

VII. Evaluation Team

The evaluation team was composed of:

• Ms. Alicia Casalis, Consultant and Head of the Mission, UNCDF

• Mr. Jean-Marc Rossignol, Consultant, UNCHS

• Mr. Ahmed Ali Hemed, Urban Engineer and National Consultant