NUMBER 148 -EEDPPrecis Operations Evaluation Department May 1997 A Success and Challenge: AGETIP in Senegal Senegal's recent program to combat stemming from high unemploy- The projects included a prog- unemployment is widely regarded as ment in urban areas. The govern- ram of urban public works, con- a development success and a model ment realized that a solution to struction, rehabilitation, and of efficient and effective public works high unemployment needed to be maintenance; a package of services management. The public works and em- found rapidly, but it also knew that related to work organization, man- ployment agency (AGETIP) has become government agencies and public agement, and training; and a com- well known in the region for managing enterprises could not deliver con- ponent for supervision and labor-intensive public works projects struction and service programs monitoring. AGETIP II also in- with transparent, streamlined proce- quickly and efficiently. Thus, the cluded additional pilot activities dures, and for significantly improving government created the Agence for the application of AGETIP's the management of donorfinances. Its d'Execution des Travaux d'Interet experience in government proce- approach has been widely replicated. Public (AGETIP). dures; the program's expansion into rural areas; increased grass- Between 1989 and 1996, the World AGETIP was granted special roots participation; a feasibility Bank supported AGETIP with two legal status, with its own charter study for a mutual guarantee com- credits. An OED audit* of the two and bylaws and with exemptions pany; and a lending instrument projects highlights AGETIP's success from government procurement and for municipalities. and notes that the Bank worked very disbursement procedures, as well effectively to pilot, monitor, and adjust as from staffing and salary policies. Implementation the projects before expanding them. The basic operating principle of the agency is delegated contract Preparation and appraisal of The audit, which included a spe- management (see box 1). Over time, AGETIP I successfully followed cial independent assessment by bene- AGETIP has acquired a virtual mo- a learning cycle. A first project ficiaries, sheds new light on the sus- nopoly in construction management. preparation facility was set up tainability of the subprojects and in late 1988; the project was ap- warns that, in spite of AGETIP's suc- The World Bank supported praised in March 1989; and a pro- cess, grassroots participation by local AGETIP with an IDA credit (AGETIP cedures manual was prepared and communities will have to increase if I) of $20 million approved in 1989 discussed with the government. benefits are to last over the long term. and a second credit (AGETIP II) Beneficiaries, in their assessment, said of $39 million in 1992. A third theyfelt left out of the identfication project, aimed at increasing the ca- 'Performance audit report: "Senc- and planning process. Without their pacity of local governments, is now gal-AGETIP: Public Works and full involvement, the maintenance of under consideration. The main ob- Emnploymnent Projects,'" b4 Robert subprojects and even the sustainability jectives of AGETIP were to create van der Lugi and Thomas Kuby, of the program may be at risk. temporary new jobs, increase the hirthicoming. .4vailable to Bank capacity of the construction indus- executive directors and slaff from In the early 1980s, despite a try, demonstrate the feasibility the Internal Documents Unit anid decade of structural adjustment, of its approach to delegated con- frorn regioonal information service Senegal continued to face serious tract management, and invest in cetnters. Pr&cis written by Robert economic problems. These were economically and socially useful van der Lugt and Stefano Petrucci. aggravated by political unrest subprojects. AGETIP today Box 1: The AGETIP approach to delegated contract management Delegated cnrcmagmn OniAGETIP was established as a Deleg,ated contract management Once it receives the money . temporary agency. But its encour- works as follo% s- the originator of a AGETIP awards a contract tor de- subproject (in most cases munici- tailed engineering. It then carries out aging results stimulated an in- palities) submits a proposal to all bidding procedures, awards the crease in activites, and the Bank AGEllP, which appraises the sub- contract for construction, super ises and international donors eventu- project and deterrmunes its eligibilitv execution, and arranges payments to ally gave AGETIP responsibility If the subproject meets AGETIP's contractors and consultants. AGETIP for construction management in criteria, it enters the work program. also contracts specialized consultants a large number of other projects. AGETiP enters into an agreement for quality control. When the project AGETIP also manages develop- wvith the originator, assuming a]l the is completed, AGETIP arranges for ment projects by assuming func- rights and obligations related to the the transfer of project assets to the tions that go beyond contract execution of the subproject. AGETlP originator. AGETTP also offers fre- management. For example, it takes no action, ho%%e% er, until it quent seminars for municipalities, is managing a French-funded receives the originator's counterpart contractors, and consultants to ex- microenterprise credit fund and financing tpresently 10 percentl. plain its procedures and mandate two separate IDA-funded projects in nutrition and female literacy. Then, a second project preparation Outcome AGETIP has become a highly facility was created in May 1989, visible and powerful organization. which fine-tuned the process by Both projects were carried out It is lean and efficient. For example, hiring staff and carrying out five successfully and in line with ex- accounts are paid in an average of pilot subprojects. Only after this pectations. By February 1996, three days and the procurement phase was completed did full-scale AGETIP had executed more than process is completed in less than implementation begin. 1,250 subprojects and created al- two months. About 90 percent of most 18,000 person/years of em- contracts awarded by AGETIP are AGETIP I became effective in ployment. If all subprojects are completed within the specified time. March 1990. AGETIP II followed completed, AGETIP expects to in 1993 and expanded into rural ar- have created a total of about Beneficiary assessment eas, establishing three regional of- 35,000 person/years of employ- fices to decentralize decisionmaking. ment. At the time of the audit, OED collaborated with the Ger- Difficulties in obtaining counterpart the ratio of labor costs to the total man donor agency GTZ to carry funds from the government and costs of most subprojects (indi- out an independent beneficiary as- municipalities delayed startup of cating how labor intensive they sessment. The assessment took AGETIP I. The situation worsened were) varied between 23 and 28 about four weeks in 1996, includ- in 1993-94 and IDA almost sus- percent, above the threshold of ing three moderated workshops in pended disbursements in 1995. 20 percent but below the ex- Dakar, Saint Louis, and Diourbel. To ease the funding crisis, caused pected level of 33 percent. Sub- The assessment team visited 23 in part by a dire shortage of mu- stantial capacity has been built in project sites and held 33 semi- nicipal finances, the share of mu- the construction industry. The structured interviews with benefi- nicipal counterpart funds was number of engineering firms rose ciaries. In all, the team listened to reduced from 50 to 10 percent, from about 80 in 1991 to more 159 individuals. Although the and donors agreed to provide than 300 in 1996, and the number findings of this rapid assessment credits to fund municipal contribu- of registered construction enter- cannot be statistically representa- tions. The project had hoped to prises soared from 350 to 1,900 boost the municipalities' sense of during the same period. -d --- - -- ownership of the projects by in- Agetipiser creasing the municipalities' finan- Pilot activities had varying de- cial contributions. But this goal grees of success. AGETIP suc- ! AGETI has had such an impact was not reached. cessfully expanded into rural that a new verb has come into be- .and its procurement expe- Lng agelipiser, meaning to use the areas, Anc waS procuremew of AGETIP approach to make things In spite of these problems, rience was used in a review of in the public sector more efficient, AGETIP enjoyed strong political government practices and legis- to improt e them. or to fi\ them. support and benefited from out- lation. Training and study activi- Bank and Senegalese government standing performance by both the ties were carried out success- officials often tell each other that executing agency and Bank staff. fully. But efforts to create a joint in order for a project to be done Especially noteworthy was the venture with municipal credit i properk. "i faut agetipiser" (it continuity of Bank staff working funds to finance infrastructure must be agetippred). on the project. projects failed. _ May 1997 tive, they offer a reasonable indi- cation of AGETIP's strengths Box 2: From AGETIP to AFRICATIP and weaknesses. WVhile labor-intensi% e projects the aegis of specific executing W'hile opinionis differed, the need are not new. AGETIP has been suc- agencies, AFRICATIP organizes cessiulk replicated in sex eral Afri- meetings that ha%'e resulted in sub- for improved maintenance and par- can couintnes. Ln order to benefit stantial cross-f'ertilization. There ticipation by local people was a irom each other's experience. Afri- are now 14 member agencies, and dominant concern of almost all ben- can AGETIPs Ln 1993 created a re- nine potenhal new members at- eficiaries. The assessment found gional association, AFRICATIP tend meetings as observers. A link that three groups of intended ben- (Association Regionale des Agences has also been established with eficiaries-municipalities, small d'Execuh'on des Travaux d'lnt&ert social funds in Latin America, construction firms, and consulting Public), as a formal networking or- where I funds haze lormed their engineers-have been directly in- ganization tor the programs. Under own nehvork. volved in subprojects. But the fourth and largest group of in- tended beneficiaries-the local had not been understood and that come is abundant, there are few in- people-feel left out. This has im- beneficiaries had not been able to dications of the economic and social portant implications for the future: participate sufficiently in project impact of subprojects or of the per- since local communities do not feel identification, resulting in little in- manent employment created. they have ownership of the projects centive to maintain assets. Those and municipalities have few if any subprojects that did originate from The sustainability of benefits resources for maintenance, the up- local initiative were in fact rela- created by the projects is uncertain, keep of assets created by sub- tively well maintained. mostly because of operation and projects remains in doubt. maintenance problems caused by Impact low participation by local communi- Many municipal administrators ties and lack of municipal funds. felt that the subprojects were in AGETIP has become a perma- A number of subprojects, such as line with the people's needs, but nent, fee-based institution (fees roads and community centers, other groups pointed to the lack of fully cover operating costs) with have already deteriorated and consultation and to projects that good financial performance and need rehabilitation. Others do were not in line with priorities. management. It has created jobs, not seem to meet the priorities of (For example, one respondent com- some of them permanent, and indi- the communities. mented that "neighborhood residents cations are that construction activ- would have preferred a sanitation ity in Senegal will continue to The projects' institutional impact project instead of sidewalks.") Re- increase in the short to medium was substantial. They improved the spondents acknowledged, however, term. AGETIP has significantly im- management of financial resources, that some projects clearly origi- proved the management of finan- developed new and innovative solu- nated from people's initiatives or cial resources from donors and tions to public works contract man- reflected community priorities. proven itself to be an efficient and agement, and replicated regional transparent management company. capacity building. There are several Engineers and entrepreneurs The agency has proved that proce- reasons for AGETIP's success. It noted that AGETIP could expand dures for bidding, contracting, and was set up as a nongovernmental appraisal procedures to include disbursing small public works can organization with a results-oriented social assessments and take into ac- be made competitive, simple, and management culture; it enjoys count earlier studies or proposals. flexible. Its procurement practices strong political support; it follows At present, only technical engineer- seem to be leading to reforms in high standards of efficiency and ing studies are included in the the government's procurement transparency; it has a highly moti- preparation process. But some ben- procedures. AGETIP's approach vated and dedicated staff with strict eficiaries expressed the concern that has been successfully replicated in ethical and professional standards; expanding AGETIP's mandate and several African countries, and the it has financial decisionmaking au- activities could reduce its efficiency. agency has forged networks with tonomy; and it provides training other social funds in Latin America. for its partners (enterprises, consult- Most beneficiaries agreed that ants, and municipal managers). operation and maintenance needs AGETIP included a substantial to be improved. Beneficiaries criti- monitoring and evaluation pro- Lessons cized municipalities for poor main- gram but little is known about the tenance but officials from munici- impact of AGETIP subprojects be- * To ensure sustainability, benefi- palities pointed to the lack of re- yond the number of workdays cre- ciaries must be closely involved in sources. The assessment concluded ated. While information on the use project identification and planning. that the real needs of the population of inputs and of outputs and out- But the question of how to ensure OED Precis such participation remains. Some overburdening their capacities. * AGETIP's success has important believe that AGETIP should be What is needed is capacity develop- implications for its future, as well granted greater responsibilities ment at the municipal and commu- as for the future of other similar for project identification and for nity level. agencies in Africa. Construction operation and maintenance. But management is carried out mostly experience with integrated rural * Careful project preparation and by the private sector. Thus, allow- development does not support sequencing of actions (piloting, ing AGETIP to hold a preferential this idea: transferring government monitoring, and adapting before monopoly status may not be an ap- responsibilities to project units expanding) were crucial to the suc- propriate institutional solution. A does not lead to sustainable insti- cess of AGETIP I and II. So were distinction may need to be made tutional development, and experi- the persistence and continuity of between commercial construction ence with rural development Bank staff, as well as political back- management and management of funds points to the dangers of ing by the government. social projects. OED Precis is produced by the Operations Evaluation Department of the World Bank to help disseminate recent evaluation findings to development professionals within and outside the World Bank. The views here are those of the Operations Evaluation staff and should not be attributed to the World Bank or its affiliated organizations. This and other OED publications can be found on the Internet, at http: / / www.worldbank.org/html/oed. Please address comments and inquiries to the managing editor, OED, tel: 1-202/458-4497, fax: 1-202/522- 3200, e-mail: eline@worldbank.org May 1997