Tunisia - Northwest Rural Development - Phase I - Project

* ml3X .2 0 il I:! et i 1-40 li 0 li i li i i' ti w hr I i' g; i i 0w CURRENCY kOUfVALENT YEAR US$ TUNISIAN DINAR (D) 1981 1.0 0.4938 1982 1.0 0.5907 1983 1.0 0.6788 1984 1.0 0.7788 1985 1.0 0.8345 1986 1.0 0.7940 1987 1.0 0.8287 1988 1.0 0.8578 1989 1.0 0.9371 1990 1.0 0.8000 IT AND MEASU Metric System AMD = African Develop mt Bank BDPA - Bureau pour le développanat de la production agricole BNA = Banque nationale agricole BTO • Back to office CES = Conaertion des eaux et des sols CNEA = National Center for Agricultural Studies CRDA Ca i rgional au dveloppement aguicole FAO/CP - Food and Agricultural Oganition/ World Bank Coopeti Prog GTZ Des Geselschaft f5r Tecnisch u aam IBRD = International Bank foa Reconstruction and Development IFAD = International Fund for Agicultural Developent IRR ~ Interal ute of return M&B = Monitoring and Evaluation NWRDP = Northwest Rural Development Proect ODESYPANO Office de développement sylvo-patoral du Nord-Ouest OED - Operations Evaluation Departnent PAR = Performano Audit Report PCR - Project Completion Report ROR = Rate of return SAR . Staff Appraia~l Repor SIDA - Swedish International Development Agency TD = Tunisian Dinars USAID - United States Age~cy for International Developm et WFP . World Food Program -O-VERNBME FISCAL YEAR January 1 to December 31 NOR OWICIAL USB ONLY THE WORLD 9ANK Washington, D.C. 20433 U.S.A. offio of Director1General Operados "val June 22, 1993 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Tunisia Northwest Rural Develooment Prolect (Loan 1997-TUNI Attached is the Performance Audit Report on Tunisia - Northwest Rural Development Project (Loan 1997-TUN) prepared by the Operations Evaluation Department. The project area is a hilly, over-populated pocket of relative poverty where over- grazing and unsustainable farming are causing serious soil degradation. The project was to improve livelihood by changing land-use systems: reforesting hill tops; intensifying stock- raising through pasture improvement and dairying; and concentrating farming on lesser slopes. A range of classical soil-conservation measures were envisaged. Public services: access roads; drinking water; schools; milk-collection centers; extension; and credit, were to be provided. Slow uptake of the new technology was anticipated; high risk was recognized. In soil conservation works, the project did best where no changes in land-use systems were required. Where such changes were required, works were resisted or were not sustained. The project had almost no success in persuading farmers to change their farming or pastoral practices. Overall, the project outcome is rated as unsatisfactory, even though the social infrastructure objectives were largely achieved. While preparation was protracted (over six years), the Audit finds that warnings from Tunisian and Bank technicians over lack of a proven technical package were not accepted. Supervision was uneven. The Audit finds that the ecological characteristics of the successful pilot project area were not replicated in the project area. Consequently, expected diffusion of system changes did not take place. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorlation. FOR OFICIAL USE ONLY NORTRWEST RURAL DEVELOPMENT PROJECT (LOAN 1997-TUf TABLBECNT PREFACE . .. .. .. .. .. .. .. .. .. .. . .. ... .. .. .. .. .. .. .. . .. i BASIC DATA SHEET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii BVALUATION SUMMY. ........ . . . . . . . . . . . . . . . . . . . . . . . . xiii I. BACKGROUND . . .. . . . . . . . . . . . . . ... . . . . . . . . . . . . . . . . .1 I. PROJECT DESIGN . . . . . . . . . . . . . . . . . . . . . ....... . . . . . . .2 IV. IMPACT .. .. .. .. .. .. .. .. .. .. .. .. . ... .. .. .. .. .. .. . .. 9 V. ISSUES AND LESSONS . . . . . . . ... .... . . . . ... . . . . . . . . . ... 14 Attachn~ t 1 . .. . ... ... . ... ... . ... . ... . .. . ... . ... .. .. . .. 19 MAP IBRD 15395 IBRD 15396 This document has a rustuctd dstribution and may be ud by ~cipint only In the performanc of the official duti~s. 1t contents may not oth~rwis be discised without Wofid Bank authoration. PERPOMANCEAUDITRPR NORTHWEST RURAL DEVELOPMENT PROJECT (LOAN 1997-TUN 1. This is the Performance Audit Report (PAR) on the Northwest Rural Development Project for which Loan 1997-TUN in the amount of US$ 24 million was approved 15 May 1981. The original Closing Date, 30 June 1987, was extended by two and one-half years to 2 January 1990. A sum of US$ 6.5 million was canceled in 1987 and a further US$ 1.1 mililon was canceled at Closing. The final disbursement was made on 10 May 1990. 2. The Borrower obtained project co-financing for this project in German (GTZ) technical assistance valued at US$ 4 million and World Food Programme (WFP) rations valued at TD 1.2 million (US$ 1.6 million). This support was part of broader GIZ and WFP projects. (See WFP's comments on this audit in Attachment 1.) 3. The PAR is based on the Project Completion Report prepared by the FAOIIBRD Cooperative Programme (Parts I & III) and by the Borrower (Part II), on the Staff Appraisal Report, the President's Report an. the legal documents, on the transcripts of the Executive Directors' meeting at which the project was considered, on study of project files, and on discussions with Bank stafE 4. An CED mission visited Tunisia in January 1992. The mission discussed the effectiveness of the project with Government officials from the Ministries of Agriculture (including the project authority, ODESYPANO (Northwest Forest and Pastoral Development Authority) and CNEA (National Agricultural Studies Center), Planning and Regional Development, Finance, and National Economy, as well as private citizens and representatives of international organizations. The mission met with representatives on GZ WFP, and FAO/CP in the field, as well as with representatives of WFP and FAO/CP in Rome headquarters. 5. The PCR provides a generally accurate and perceptive account of project experience, including performance of the Bank and of the Borrower, including executing agencies and other participants in the project. It correctly concludes that the project was successful in providing public services and in executing works on public land, but unsuccessful in getting farmers to change their farming and grazing systems. The PAR probes the underlying causes of this outcome. 6. Following standard OED procedures, copies of the draft PAR were sent to the Borrower and its executing agencies and to cofinanciers for comments. Comments from the World Food Programme are Attachment 1. Appropriate changes have been made in the text, PERPORMNC AUDM REPOR NORTHST RURAL DEVELOPMENT PROJECT (aan 1997-TUl BASIC DATA SHlEET 1. Related Bank Loan 2234-TUN 1983 Central TNiean Irrigation 13.7 Under implene~ 2502-TUN 1985 Northwest Agricultural 8.8 Under implhm~nn Production Project 2573-TUN 1985 Iigation Manaement and 22.0 Under impl.mentati n Improvement Projet 2605-TUN 1985 Gabes Irrigation Project 21.7 Under implem~ntatin 2865-TUN 1987 Fourth Agricultur Credit 30.0 Under impe Projeat 2870-TUN 1987 Forestry Development Project 20.0 Under ipanentain 262-TUN 1988 Sector Adjusent Loan 150.0 Under implam~ntation 3078-TUN 1989 Second Sector Adjument 84.0 Under implementat n LOMan 3217-TUN 1990 Research and Extension 17.0 Under impleention Projet 2. KeS PoM e Data TOta Proje cost (TUMIllion) 61.5 46.725 76 Loan Amount (US$million) 24.0 16.446 68 Cofinning - Total (US$milon) OTZ 1.7 3.991 235 WFP 1.7 1.56 92 Economic Rate of Rturn 7%* 6% 86 Phiyseal Components Completed as of Closing 67% overall, ranging from 18% to 1209% by component. • Ratbasedon S-yerprojeot Apprisalexpced that, if two snhq~=nvt enproj~tworvimplme~t BROR for tm 1S-year p~ogman would be 16%. 3. Project Timetable Reonnniananc --VI 1978 Tantifintion- X 1978 Preparation X 1978 VII 1979 Pre-appraal - - VI 1980 Appraisal V 1980 Vi 1980 ix 1980 Negotiations III 1981 - iV 1981 Board Approval V 1981 - V 1981 Loan Signatur V 1981 - Vi 1981 Loan Effetiveness VII 1981 - II 1982 Loan Cosing IX 1987 - XH 1989 Loan Completion Im 1988 - VI 1990 4. I1a Didbrm ents A. o ~nnlativo Esinited and Actual Dibuwen 1982 XE 1981 Vi 1982 VI 1982 500 500 -0 1983 IX 1982 550 1050 0 XD 1982 1050 2100 437 437 3 21 1 1983 1050 3150 2101 2538 15 81 VI 1983 1050 4200 521 3059 19 73 1984 IX 1983 1040 5240 766 3852 23 74 XI 1983 1040 6280 - 3852 23 61 111984 1040 7320 957 4782 29 65 VI 1984 1040 8360 627 5409 33 65 1985 IX 1984 1160 9520 230 5639 34 59 Xi 1984 1160 10680 609 6248 38 59 11 1985 1160 11840 388 6636 40 56 vi 1985 1160 13000 877 7513 46 58 1986 IX 1985 1240 14240 457 7970 48 56 X 1985 1240 15480 179 8149 50 53 M1 1986 1240 16720 1042 9191 56 55 vi 1986 1240 17960 800 9991 61 56 1987 IX 1986 1510 19470 6 9997 61 s Xf 1986 1510 20980 582 10579 64 50 1111987 1510 22490 518 11097 67 49 vi 1987 1510 24000 91 11188 68 47 1988 IX 1987 - - 1114 12302 75 s XI 1987 - - 232 12534 76 52 1 s 1988 - - 750 13284 81 55 vi1988 - - 289 135'r3 83 57 1989 ix 1988 - 693 14266 87 59 XE 1989 - - 1186 15452 94 64 I111989 - - - 15452 94 64 vi 1989 - - - 15452 94 64 1990 ix 1989 - - - 15452 94 64 X11 1989 - - 534 15986 97 67 In199 - - - 15986 97 67 vi 1990 -1- 460 16446 100 69 4. 1 laDibøremnt (CoWtd) R. Diburements by Category Category ry FY FY fy FY FY FY FY TOTAL SAIR Atd 1s 1982 & 1983 1984 1985 1986 1987 198 1989 199E I %0 i l i i l l I l EsBimatt ........... US'0 ............. % Comstruclon of 319 522 282 345 197 1036 196 651 3548 3 100 114 bulidings Vebles,euIpmnat •2460 699 328 313 372 183 21 174 4550 7300 62 and ivestock Soil conservation, 233 907 659 1187 495 512 229 482 4 704 470 100 forst and pastures M. and L. term credit - - i10 77 - 269 91 122 669 800 37 facrmental ST credit - - 293 195 - 292 288 479 1 547 400 387 Conulants' service$ 26 163 351 203 113 94 60 85 £ 095 2 400 46 Slodift 2£ 59 81 158 20 - - - 339 000 34 Agricultura resrch - - - - - - - N 500 0 Unalmocated - - - - - - - - - 2800 Sub-totai 3059 2350 2104 2478 1197 2386 886 1993 16453 24 000 69 *Fund* - i TOTAL 16446 Uk 111111 I~Tj[fit[JiII[[!IifI1I1I~i1 i.Ii~.i I. I zz zzz z zzzz I ~ hjul II rImrrII irimim sirrir ri u r r~rwwr ~d. - *d. ål i I 5 .~o.oWWæ ~j ISb ~ viii - 6. Prgå Cot itsg SAR Eat .J/ Acual CossAsm C~u* LT Loå F Total Infrastran 2 S0 2039 4 619 5363 2 357 7720 167 supla traco SoilcoMuevaiom: 6278 1252 7 530 2420 3675 6095 81 p r fore y and wo~dlot Eqnipmmu 3677 6258 9935 1362 3493 4 855 49 Co~nutm OMic 198 156 354 839 606 1445 408 S 24/ 51 202 253 655 70 725 287 R s 289 199 498 3 14 17 3 Studies W 398 498 48 272 320 64 Teci.inaliseam • 1953 1953 21 3518 3539 178 O-arm kamn* 1064 1067 2131 502 551 1053 49 Rvoving ed for 124 495 619 991 1333 2324 375 Opradomand 11978 3482 15460 7849 - 7849 2m TOTAL 26349 17501 43850 2005 15889 35942 82 1/EBaed om rsen yoam of projeced -yse pognams. y/ Aced ons lud. purchasing cos of bWDs and mans, är m m cosa and cosm of plam wheas SAR *dma ouly oar e & ie km. / Sin of o0-an * ··- Saam~ed wikoout agricakural cr-mE*, ~/ Inclig Word Food rpaum ~ adom to te val. of D 1.2 llio disuib~ted by e Offis to esaatbasutsupemu teinasa - :Lx - 7. Pat Reut IUes~c and Crop Prodcton Increase - Mk '000 2,717 6,800 5,000 67,930 5,000 Jitres -Beef tons 354 885 700 4,465 700 -Lamb tons 280 700 300 2,780 300 - Wheat tons (412) (1,030) 6,000 (10,300) 6,000 - Chick Peas tons 108 270 500 2,700 500 Increased Student Number 1,260 2,160 1,930 - - Population Roads km. 745 1,065 657 1,940 - Improved buIls1 Number 115 221 232 923 - 15-year programm ' Assuming no second phase project 2' Total mer supplied, not in use at oe tie. 8. bak Rou A. Staff Inputs (Staff Weeks) · Tase ·· gs · · 30.7 30.7 propmaadn 106.7 106.7 Appraisa 64.7 64.7 NegaduLens 2.6 2.6 ub-total 203.7 203.7 Supervlen .0 14.1 12.2 6.7 6.6 6.6 17.4 9.6 1.4 1.4 77.3 Project .6 .2 1.2 .0 .2 1.6 • • 4.1 Adminleytaon Anual •- - .6 .2 - .7 lieview .ub-ota 14.3 13.4 6.8 8.8 6.6 19.7 10.0 1.4 1.4 L.1 TOTAL 204A 14.3 1A .8 6.6 5.6 1.7 10.0 1.4 1A 2.6 - xi - 8.Use of (Coat'd) B. Mi Data - 11rse~Ym4 Pre-appraisa" VI 1980 2 28 E,L Appram"e M 1880 8 132 E.,A,PR Qupvismi i 1981 1 4 E 1 2 - &pervsin 2 0191 1 6 F 1 2 - Sperin 3 8 82 2 4 E.L 1 2 - Supenwdeln4 M1982 2 7 L.A 1 2 8upeien6 183 1 2 A - - Supendlslm 6 Vi1983 4 7 L.A,.D i 2 - Suprvson 7 11984 3 6 L.E.D 2 2 M &p ~ 8enn IV 1984 1 2 F - - &uperv0elen X1984 3 6 L,L,E 1 1 M Qpedmlon10 IX18 2 8 L.D 2 2 M Superon11 X s 1 10 L 3 SuprWIon 12 M 1987 1 4 F Supndeln 13 Vi 1987 4 10 L,E,E,O 3 - Supnlen14 IV 19M 2 11 L, 2 super!lon16 119 2 2 L,A 2 Af E: Eoonomet L:- stock SpLvel et A: Agronomst L: fural Engieer P: Paetu pecialet R: Road Spea~ F- F:orter 0: Othere. kl 1: Prolem ree oorm proem; 2. Moderate prolem.; 3: Major pro~em af 1: bmproin 2 st~onarr 3: Deteviadn i F: pnen~- M: Managerlab T: Tedicab P: P .odo 0: Od~her. PERFORMCE AU R NORTHWEST RURAL DEVELOPMENT PROJECT (LOAN 1997-TN) EVALUATION SUMMARY was known technically to justify investment in 1. The hilly project area gets enough the hilly area with such limited potential. rainfall to grow wheat, barley and chickpeas, but human population is well beyond what 4. During 6.5 years of project agriculture will support. The average farm is preparation, the project evolved towards the not big enough for subsistence. The common Bank Programs conception Drawing on grasslands and remaining forests are being exience from the successful GTZ-assisted degraded by grazing sheep, for which available Scjnane dairy project, the Northwest Rural forage is insufficient. Cropping has expanded Development Project (NWRDP) was to to areas of steep slopes. Because plots are persuade rural communities to intensify vertical and narrow, cultivation is often along livtock raiing by improving pasture on the the greater slope. Erosion and soil middle slopes. This would free the upper and degradation result This was accelerating steeperslopesforreforestation,wilecropping siltatli of Sidi Salem reservoir, a Bank- would be concentrated and intensified on the supported project on the Medjerda plain which lesser slopes. Clssical soil conservation splits the project area. measures would be installed, mobilizing as much participation of the local population as 2. The system is not sustainable. It does possible In cae this did not brke not serve the area's population well, even in of Sidi Salem reservoir, a border the short run. forest belt was to be planted. 5. Techniciams recognized that the OBJECIES prescried lanuse system changes would not be immediately profitable to farmers and 3. The Planning Ministry in Tunisia and pastoralists, hence that their adoption would the Programs Department in the Bank wanted be gradual To provide incentives for a project that would do something about this participation (and because they were part of situation. The former envisaged a group of the Bank's vision of integrated rural punctual interventions; the latter, an development), the project was to provide integrated, area-based project. The public services which the project area didn't Agriculture Ministry in Tunisia and the Bank's get: drinking water, access roads, schools, milk agriculture d on were skeptical that enough collection and agricultural Vie centers, - Ipv extension, and formal credit. Even so, the reduced Level by Government fun&s GIZ appraisal mission anticipated that the ultimate assistance, and 'bridgin finance from economic IRR would be only 7% if the project under the 2nd Agricultural Structural ended after the five year agreement, rising to Mjustment Loan. 16% only if two subsequent investment phases of a 15-year program materialized. Moreover, the project was recognized to be risky. 6. In addition to Government and Bank 1& Despite severe dat deficiencies, the finance, the project was supported by GIZ general picture of the projects results Is quite and the World Food Programme. clea It divides neatly ito three parts. 11. First, In execution of works to buil IMPLEMENTATION EXPERIENCE public facilities on publi laudg ODESYPANO did well. 'Me schools drinkinig-water systems, 7. The Borrower initially resisted having milk collection stations, and acss roads to a new, autonomous project authority, which douar (hamlets) got built Some very sensible the Bank promoted. Getting this authority were made in the drining-water (ODESYPANO) approved delayed loan componen4 replacing the small number of effectiveness by eight months to March 1982. expensive systems planned with a much larger Implementation was hampered by a hiring number of simpler, more cost-effective freeze that caught ODESYPANO with plenty systemL ODESYPANO built 12 times as of central staff but not many field personnel, many drinking-water systems as anticipated. as well as by a reorganization to decentralize Forests on public lads, especially the fors Agriculture Ministry operations. Organizing belt around Sidi Salem reservoir, got planted rural communities and eliciting their and generally grew. Access roads made quite participation proved more difficult than a difference to isolated rura people. In shoa% anticipated. Disbursements lagged. The loan the poor, hilly area got some public services it closing was extended by three years to end had been deprived of for millennia. 1989. By that time, targets had been exceeded in some areas (drinking-water points, gully 12. Second, results we mixd to correction) but were far from reached in construction of works by force account an others (pasture improved, cropping private buds and In their develont intensified). Overall, probably about two- result Ravine control by gully plugs and thirds of projected works were constructed, plantings on banks exceeded targets by two 76% of projected project costs were spent, and and one half times. Erosion control works - 68.5% of the Bank loan was disbursed. The principally contour rock terraces - met targets rest was canceled. Pasture wa only one4furth of targt. Freswould not permit tree & Bank supervision was slow to pick up planting on private lands. Acceptance vared ODESYPANO's problems. Later missions, inversely to disruption of the frmig system. however, are highly critical of administrative Except for rine control, which doesn't overstaffing, over-centralization, and the disrupt because ravines are already lost to failure to get monitoring and evaluation going. farming all measures were resisted. 9. Since Loan 1997 closed at end 1989, 13. Ibird, where ODESYPANO tried to ODESYPANO has been maintained at a change, an mud stock-rearing sysbms - av - results ranged from neglgible to modest. ODESYPANO's soil & water conservation Farmers were not willing to stop cultivating engineers about the aspects of steeper slopes. There was little or no the project "But of course they have intenslcation of production on gentler sl participated said the engineerso pwe paid Little improved grazing land has stayed that them, and they carried the stones and planted way. Since forage supply did not increase, the seedlinge project efforts to increase livestock numbers may have been counterproductive. No genetic 17. Inatitutional development. improvement was achieved in sheep. Farmers ODESYPANO proved to be an instrument rejected tree-growing on their own land acept to its complicated task. There for olive and almond trees. All vegetative were too few field staff particularly contour bunds disappeared. Stone contour tsandcentral-officeorientationwas terraces were accepted on non-cropped land so strong that even those few extensionists because de-stoning the adjacent land made it devoted less than 20% of their time to a arable. Milk and livestock production extension and kept having their vehicles taken continued to expand in Sejnane and expanded by higher authorities for other uses In 199Z modestly into adjacent areas. ODESYPANO was still overly centralized for an organization with its charge, though less so 14. Omen Animal products than at PCR time account for 75% of the project's incremental benefits as per the PCR. While the M An additional institutional problem is agricultural results point up failure to change that, with decentralization of public the cropping system, the livestock targets are agricultural support to governorates, not far from what the SAR expected after 7 ODESYPANO has become an anomaly. years. The PCR calculates an economic IRR However, in fact, the decetralized institutions of 6% for the 7-year project, as compared with do not deliver services to the poor and remote the SAR's 7% for a 5-year project, worse but hills. Some sort of special institutional not disastrously so. like ODESYPANO, will be needed to do so. 15. The PCR's critical finding is that the projects innovations will not diffuse further, 19. ODESYPANas ethos is based on soil leading to a steady state from 1989 through and water conservation, regional planning and 2012. Thus, the economic IRR is expected to civil engineering. It is topdown. Many of stay at 6%. On this basis, the Bank's Regional ODESYPANas engineers beive that with Office (Part I) finds the project marginally patience, this approach will work but wasatisfactory. For reasons explained below, Government of Tunisia, 017 FAO/CP and the Audit concurs. The Borrower, as reflected the Bank agree on the need to re-orient in Part II, expecting slow systemic change as ODESYPANO to elidt peoplels participation. forecast by appraisal, finds the project Government hired consultants who produced satisactory studies on partnership and participation in 1991. Seminars have discussed what to do. 16. ODESYPANO's infusion of cash and Whatever happens, it is unlikely that future World Food Programme food were the main public efforts will be as directive and as top- factor in the 11% annual consumption increase down as those in the past Participation is in the project area between 1985 and 1990 (cE ly to involve more than putting stones 6% for all of Tunisia). This infusion was not whe you ae told and getting cash and food a simple give-away. The Audit askted rations in return. Therefoie t does not seem - Zvi - too sanguine to epet that the institutional Converting them to dairy farmers, slw as it development that ODESYPANO represents has been, has involved less systemic change will be converted Into something better than for the ret of the project ar. Tlere, adapted to regional development needs. bill people ae mainly wheat-arley-chickpea farmers who also keep sheep In areas where there is still abundant forest farmiers can graze, SM 1L l their stock on State forest land. Since there is a commons to graze, there is less urgency to 20. The Borrower's, Bank's and OED's improve pasturv, hence little hope of selling different assessments of the project are based the projects system change. on different expectations about sustalnablity of what has been achieved and future 23. Where forests are gone and diffusion of the core technical package. Data mountaintops are exs rock devoid of on sustatnability of project works is virtually vegetation, pastoralists have no commons to non-existent. Almost none of the research graze. It is in such regions that milk component that would have answered such collections are starting to pick up. Necessity is questions was implemented; the M&E system the mother of invention. Where the commons began operating seven years late and then was has been destroyed, livestock raising has to not effective. Some ravine training appears to change; improved pastures on the upper sopes be maintained, but others have suffered from may vesy slowly be accepted. grazing. The verdict on contour stone terraces is less obvious. Improved pasture is not 24. But if promoting change in livestock sustained except in a few instances where it is systems promises to be difficult, how much fenced. ODESYPANO reported to the more difficult, thn, the cropping system completion mission that only 1330 (19%) ha of changes. Their adoption has not been the 7000 ha improved actually still existed as demonstrated nearby, as the dairy system has of 1990. Now the figure is much less. in Sejnane. Farmers will not give up farmland to .stone contour lines. No formula for ISSUES AND LESSONS sustainable vegetative bunds on farmers conditions has yet been found. Contour 21. Diffusion Optimists believe that, plowing or en plowing along the lesse sope despite its problems, the project is on track is difficult under project-area tenure patterns, and that diffAsion of its farming-system core horizontally-narrow vertical strips that give will take place; patience is required in the face each inheritor a portion of land of evy of the slow-gestation and risk the appraisal fertility and moisture, No conservation anticipated. The Audit is less sanguine. The sytem has yet proven acceptable to improved livestock and farming practices that farmers. Prospects for systemic change based succeeded in stabilizing the Sejnane ecosystem on the original project conception ae remote; will not diffuse simply to the rest of the area they depend on further, sociologically-sensitive because the project sub-areas are ecologically mentation. In short, the projects different. for farming and livestock system changes would have been seen to be 22. Sejnane is ecologically unique. Its unrealistic if basic ecologial conditions in the residents were traditionally more pastoralists project's micro-regions had been understood. than farmers, grazing animals and growing sorghum on the bed of Seliane's former 25. ado dW intermittent lake when it dried in the spring. the Bank considered this project for eight * xvii - years before the loan became effective. The 27. Leson may be summarized as four Bank hardly rushed into this project. Yet, as points: indicated above, the long processing period is marked by a distinct reticence about the - however laudable the desire to project on the part of technicians. The deviate regional povert, without a Borrower was reluctant to have a new project sound technical package, thee can be agency too. These reservations were over- no sustainable povemnt; ridden because the Bank, in particular, and the Borrower wanted a poverty-based rural - there is no substitute for a thorough development project. NWRDP passed of ecological and agro- through Bank reviews with little concern about nomic conditions when projecting agricultural technical matters. technical diffusion; 26. If the core technical package had - for famers to fee part of an diffused even a modest amount, all enterprise, participation has to involve shortcomings would have been forgotten. But more than getting paid for prescribed is has not In retrospect, Tunisia would be manualwork; and better off if NWRDP had concentrated on building rural infrastructure, on some palliative - it is easier to provide rural public soil-conservation works, and on action research works and services than to get famers to explore what land-use-system changes and pastoralists to change thei farmers would accept That could have been Systems. done at half the actual cost and with a lot less building of an institution that is now difficult to down-size. In view of the clues and hints, why did the Bank, in particular, push for the actual project design? Firstly, there was genuine hope that the land-use system changes at the core of NWRDP would catch on, helping people, bringing good economic returns, and putting an end to the destruction of the hilly Northwest's ecosystem. Secondly, the Bank was going by preconception of what an integrated rural development project should be. PERORMANE AUI E UIISIA NORTHWEST RURAL DEVELOPMENT PROJECT (LOAN 1997-M&Di L BACKGROUND 1.1 Tunisia as a whole is day. Its northern area, though, has enough winter rainfall, when coupled with its water-retentive soils, to support a thriving agriculture. The flat northern plains of Tunisia have grown wheat, barley and chickpeas since before Roman times. With the Ancient World's population density, there was no need to farm hillsides, which were forested. Before and during the period of French rule, the powerful dominated the agricultural plains; those who labored on their large farms increasingly lived in the surrounding hills, where they had micro-farms of their own and where they kept their own sheep and goats. Ownership of the large farms of the plains has largely changed since Tunisia regained its independence; they have been large cooperatives, and now they are public domains, leased to all kinds of entities, but, Invariably, still large, highly-mechanized fars. 1.2 With rapid population growth, the hill population has mushroomed. Holdings are smaller, below what is needed for subsistence on average. Almost all agricultural land in the hills is held under traditional tenure arrangements unrecognized by the state; tenure, therefore, is insecure and the most common subject of disputes. The commons are scarce, therefore getting degraded. Most of the forests have disappeared. Even when they are protected, they are grazed and there is little re-growth. So the hills cannot sustain their population now, and they are being further degraded. 13 The ecological tragedy that is proceeding in the northwest Tunisia hills affects the human part of the ecosystem. Surveys show that more than half of family income comes from working on public-works projects and on largely menial jobs outside the region; less than half from agriculture and animal husbandry. Ecodisaster is human disaster. Undoubtedly, there is no prospect for supporting the present population and its descendants in the region on agriculture and livestock- rearing. Out-migration is in prospect and is certainly desirable, both for the non-human ecosystem and for the people who better their lot by leaving the northwest hills.' But, even as out-migration proceeds, the regional patrimony, which still provides the northwest's population a partial living I Governmt and the dBankboth disagreewith this point. Indeed, oeof the objectives of the project was to slowand atop out-migration from thS area. The Audit is impressed by the fct thst to am's population is well in excess of what it could support from the Iad and by the nonagrlundajobs that there is a reasonable prospect of dveloping. The Audit sees no humane snson to confine people to tml poverty by trying to discourag othem from electing employment and a better standard of living in Tuis or in Rome. 2 through agriculture and animal husbandry, is being degraded. Erosion and disappearance of the remaining pockets of forest are only the most visible signs of that degradation. More important and more insidious are degradation of the quality of soils, of natural pastures, and of forests through over- use and inappropriate use. 1.4 The project was designed to reverse this secular trend of ecological destruction and human impoverishment. IL PROJECT DESIGN 2.1 Design was a long and complicated process. From the first mention of the project in Bank files to appraisal, 6.5 years passed; to loan effectiveness, 8 years. At first blush, this indicates a commendable prudence to understand the region's problems, to make sure that the project embodied the best technical solutions to those problems, and to assure administrative arrangements to get the job done, enlisting the participation of the people in the process. That is part of the reason for the length of identification and preparation, but there are others as well. 2.2 For the Bank, NWRDP begins with review in early 1974 of the massive Central Tunisia rural development project preparation report, prepared by CNEA (National Agricultural Studies Center, an autonomous part of the Agriculture Ministry), which was then receiving technical assistance from PAO. The Central Tunisia project did not get Bank financing, but it prompted the Finance Minity to call for Bank help with rural and regional development planning in the northweslt, specifically, in Jendouba, BDja, El Kef and Siliana governorates. Two fully staffed missions led by a veteran Bank agricultural planner in September 1974 and February-March 1975 initiated the process. On the last day of the second mission, in meeting with the Planning Minister, it became clear that the Bank had misinterpreted the request. The mission's BTO (25 March 1975) reports, 'Minister of Planning...made it clear that he is really not interested in regional development planning. <He wants> concrete project possibilities, integrated projects like Sejnane or Central Tunisia' One of the central tensions of project design was already apparent. '... he stated that he is not really interested at this time in developing the admittedly limited potential of the hilly and mountainous regions of B4ja and Jendouba which are characterized by acute pressures of population and livestock on available land resources. Instead, he thought that integrated projects should be prepared for the efficient utilization of the resources of the three public irrigation schemes in the Medjerda plain of Jendouba.' 2Asually, 18% ofthe nothwet is 'fores', moh of It Govemaut owad W proteatd. he 18% is in varying stages of degtadation, ranging from Aly stooced pamels to "fneet with no trees and lil brush. -3- The Bank was interested in helping the lower 40%; the Tunisian Planning Ministry was interested in bankable development projects, looking to extending and/or up-grading irrigation investments on the northwest's plains to find them. 2.3 In late 1977, the Bank heard that the Planning Ministry was preparing a monograph on the region. A Bank mission In May 1978 found that the project was "not yet delineated, that components were "still rather vague, and that the person in Plan responsible for project preparation "does not appear to be thoroughly familiar with the rural development concepts and seems rather uncomfortable In crrying out his work.* Moreover, the Agriculture Ministry was stHl not Involved. In fact, the Agriculture Minister "expressed his reservation on Bank financing of rural development projects. Recalling past experiences, including the Central Tunisia project, in which rural development projects were rejected by the Bank and finally financed by bilateral aid, he considered it a waste of time to cope with the Bank's Isophisticated approach' to rural development and "the numerous requirements raised by the Bank" Discussions persisted nevertheless because of the hope that, as the Bank's Programs' Director said In October 1978, this project "would possibly become a type of rural development project which Tunisia and the Bank had been aspiring to for a long time.' 2.4 Most of the aspirations, judging by files, were on the part of the Planning Ministry and of Bank Programs, not the Tunisian Agriculture Ministry and not the Bank's agricultural division. Nevertheless, on the ground, the Agriculture Ministry, mainly through CNEA, was getting involved. Bank representatives complained, and continued to complain until appraisal in September 1980, that CNEA's work was not good, that there was no clear project concept, that the project's elements were not integrated, and that administration of project implementation had barely been thought about. From the beginning the idea had been to have CNEA prepare the project; assisted by "bus-stopping' missions from FAO/CP. During the very long preparation period, there were a large number of Bank and FAO/CP missions with a large number of different staff visiting Tunisia and talking about the project. What was meant to help CNEA was not always considered helpful by CNEA. 2.5 The process continued to be protracted and contentious through the Bank's review. For instance, a participant consultant described the Yellow Cover review meeting as an 'unusually large and long' one where 'institutional and presentational aspects were discussed at length (some three hours) and only after everyone was tired and irritable did an opportunity arise for discussion of points on internal design of the project." Why? 2.6 Plthr of o With the perspective of retrospection, we can now see that CNEA, as well as the involved Bank staff4 were up against an intrinsically difficult situation, as well as some conflicting expectations. Not least of the sources of confusion was the plethora of projects in the aea supported by numerous donors. Pride of place belongs to the German GTZ which had been working to form producers' groups, get them to fence & Improve their pasture, adjust their herd to the carrying capacity, and otherwise practice improved dairy farming since 1975. This formula proved to be one of the leitmotive of the project, but how was the O'I project to relate to NWRDP? There were also existing FAO/SIDA, USAID, AfDB, and Dutch/BPDA projects in the area, some of them at the studies' stage, not to mention World Bank projects. Early on, preparation 3 BTO =port of 15 May 1978. 4 Jternal memo of 13 February 1981. *4- was complicated because IFAD was promoting a rural developmant project in the area. A Bank memos reports that ministers were Oparticularly critical" of IFAD efforts to have their own project in the northwest - to "raise their flag" - wishing that they and the Bank would integrate their efforts with those of the Germans, FAO/SIDA and USAID. As usual, Tunisia had no shortage of interested donors, but coordinating them was a problem. At one point, Bank staff commentedO that the proposed IFAD project for El Kef and Silians overlapped cmpletely with existing Bank loans (Agricultural Credit II & M and Rural Roads). 2.7 But CNEA's problems (and the problem for the many different Individuals from FAO/CP and the Bank who made short visits to advise them) went beyond harmonizing eternal development agencies with overlapping areas and goals, difficult though that was. Basically, there were three more: confusion over the area the project was to cover; disagreement over whether it was to be a "punctual" group of sub-projects or 'integrated and poverty-oriented; and confusion over the administration that would implement it. 2.8 &A Initially, Ministry of Planning proposed a project embracing four governorates. As early as 1974, the doWilde mission proposed concentrating on Bija and Jendouba, in part because of the competing presence of too many other projects in El Kef and Siliana. By the time project identification activity was revived in 1977-8, CNEA and Bank staff were well aware of IFAD's intentions to have a project in El Kef and perhaps Siliana (fomerly part of El K). If both IFAD and IBRD were financing rural development in the northwest, how would their respective roles be defined? Since IFAD didn't seem to want to finance operations jointly with the Bank, would the two finance similar operations in adjacent geographical areas? CNEA was struggling to find answers that would satisfy the financiers who, at the end of the day, financed quite different initiatives in adjacent areas. The Bank, however, did not limit itself to Bdja and Jendouba until the fifth project brief of 11 July 1980, even though this had been mooted in 1974 and hinted at in earlier project briefs. One reason was that Government was set on financing irrigation projects in Siliana. 2.9 Punctual' or "intgae. Despite Government's initial call for rural and regional planning, what even the Planning Ministry had in mind from the beginning was specific projects in the region. Early on, Government put forward five of them that were identified and partially prepared: some small irrigation projects in Sillana; catchment protection to slow erosion into Sidi Salem reservoir (also Bank financed); extension of German-pioneered pasture improvement for dairying from Sejnane to neighboring Kroomie and Mogods;* a cereal seeds plant; and improvements to Tabarka's fishing port. A lot of the hassle of project preparation stemmed from the Bank's efforts to turn this offering into something consistent with its vision of an Integrated rural development project. First the Bank persuaded Tunisia that Tabarka port did not belong and might be financed under a different loan. Then the Bank and FAO/CP axed the Siliana irrigation schemes s laternal mem of 10 January 1980. * Internal memo of 26 October 1979. 'Thm is ample evidenow in projet file that IPAD and IRD Informed each other fron time to time of dir tations. It is not clear whethew atempts to collaboata were made. Stese are "aganoies," the Tunisian sub-division of governorates. an grounds of high unit costs (US$12,000/ha) and low economic returns (2%). The seed plant, for which there was no proper feasibility study anyhow, was fobbed off to the prospective grain storage 2.10 But the remaining two were not an integrated approach. Kroumfrie-Mogods was too narrowly pasture improvement; Sidi Salem catchment was too narrowly erosion preventicn. is message is evident beginning with an October 1978 FAO/CP BIO report and from the Bank's first project brief in November. CNEA, with FAOCP help, was sent back to work to "integrate" the Kroumirie-Mogods and the Sidi Salem proposals, adding forestry, development of hill lakes (lacs collinaires), crop development, agricultural credit, building schools, access roads, and village water supply. 2.11 titutonal-issue Finally, CNEA had its hands full deciding who would implement this project In 1978, FAOICP was complaining of CNBA's lack of focus on institutional issues.' CNEA was proposing that the Livestock Service take charge of the project and coordinate other agencies as needed. Government had indicated fs wish not to create new agencies, reiterating this to a Bank Programs' mission in March 1980. Nevertheless, by July 1980 the Bank was insisting on an autonomous project authority ("office"), lending a prophetic quality to the Agriculture Minister's earlier concerns about the Isophistication' and 'numerous requirements" of Bank rural development projects which made it easier to achieve more flexible solutions with bilateral donors. 2.12 Soil conservation was the technical heart of the project that emerged. A band of forest would be planted around Sidi Salem reservoir, with gully check structures and gully plantings in mid-catchment and reforestation up-catchment. The successful Seinane dairy development through pasture fencing and improvement would "make an oil spot' into the rest of the hilly area. These "punctua initiatives would be broadened into an integrated project by persuading farmer/animal husbandrymen to change their made of operations in a more general way: concentrating and intensifying cultivation on lesser slopes and plowing along the contour, turning the middle slopes from cultivation to improved pasture; and reforesting the upper slopes. Since some of these sweeping system changes would not obviously pay off for the impoverished beneficiaries in the short run, the pill would be sweetened by throwing in some of the public goods of which the people of the project area had seen very little: schools, roads, drinking water, farm & livestock service centers, institutional credit, and lots of opportunities to work on public works projects near their own douars (hamlets). 2.13 Most of what pre-occupled the Bank's during the review process was not matters of aguiculural or sociological substance: should financing be for 4, 5 or 6 years; how should the economic rate of return be calculated and was it adequate; etc. However, the central agriculture department 'A Govermnent auempt to rplce thees with another irdgtion schme in SiUsa based on raising Lakhimes dam and exinding its brigaled command was too late. The2% economic sdem onthe original Silia d iigadon proposal is the Dank's codmat. (see Project Brief #4 of 29 Febnary 1980); CNEA's etimato was 8%. * Thee is season to beBy that erosion-reduction above Sidi Salem was not that economically attractive anyhow. Sidi Salem has a lot of dead storage. Reservoir sedmntation will have no ooot on beafts until it is f11ied. The al load from Oued Tessa and Qued Zerga would fIl Sidi Salem at the rto of 0.5%lyear. So even arresting erosion altogethar would at begin to affect the economic vale of Sidi Sale for some time, and thea only slowly. t2 Se0 IO of 10 March 1980. ・8[\！遥．1「1,-,,,: 7 2.15 Ihe US$ 24 reffion ha slated to Omer 3961o of project emts was =Ftiated in ~ 1 1981, approved by the Board in May, and signed In Y*. Goverment was M~t to autbOrizc tbc sutonomous off*= that would implmccit the project (0DEMANO - Sovt>pastoral Denk~ Authork for the Nor~ t) and to autl~ the legal dmwmtL M de~ 10 amth to March 19U Otha finncing for the US$61-5 miflion project wo «~ to c=e km Governnåent (51% or US$ 31.5 offion c~ent, InCluding 3% Or TD 12 D1111109 In World F00d Prograninte rations to participanu), 7% or US$ 4.3 million equivalent konå the PatciP=tih Ud 3% or LM 1.7 nMon equrvalent from GTZ (Ge~ TwAnic81 AUStanCeV6 2.16 ~ Uffs a~ presentation wo unusuafly acpficit in n~ g the US$61.5 raffion ptqect just the fimt 5-yow trandm of an OTlicit 15-Yex Pro~ The 15-ym progrämn and its cosg work phA staffing nq~ents and the Db wcm prw«ted fait ltus, the SM aTlicidy mm~ that the s.y= project first stage would be burdened vath huv COSU diat mild bo p~ down later Ihoc mmted of buffding up ODEMANO, tec~ auåtancc. g~g mm roa& m, w~ over the mM poopie with driåking wua, ~1s, ed~ ud subsidbed credit & input supply, and MdicL In later phases, the ~on of ~~ on wä 1,,ncaum and ~ demerapsnent could bo ina~ Mds can best be se= by comemti<n "*_~ -"- com~ the sham of a= ~ted to «6 purpm in the 5year project and in the progm= Qm~ -chm of com- LJ= r~ ii-sa kw~ wo mmervation a pa~ 13% 25% forestry, wooffiots 15% 18% ämtock demlopment 5% 10-5% b= equipsnent 7% 12% a~ tural rescasch 1% 1% maffia- -- p~ g, man gmop. 12% 6% ~ ~cal mis=ce 7% 1-5% roads 12% 6% drinkipg water 3% 2% schook 2% 1% exterksion, input supplyt credit 21% 14% studies 2% 0% M mW by WPP in A*w~ 1, Wpp p~ Tu~ 2582. Ide~ D~~ in dw Nod~ g~ fbod at US$ 15 ~. M& Wpp p~ ~ dm p~ ~ of ~ &g worm ~ ab dm Nor~ Rmd ~~ pi~ ~~* w~ b~), ag wen a ~ am, b~ ~ of Om IFAD pr~ a~ te in ~ IL IU f~ of TD 1.2 n~ is ODMWÅNO'8 09~ of tbo v~ of WPP food ~ it dh~ in tk ~P ma. ~ WFP m~. am u~ am~ to ^»P ~ wa bc* a pm of a b~ aM b~ ~~ W= in T~ ~ ~ b~ *e ~ S~ p~ ~ ~ N~P and tk aa~ Fwe~ ~ budqu~ at AM D~ -8- With commendable realism, the appraisal report recognized both that the entire program was risky and that the first 5-year phase would not give an acceptable return on resources unless followed by the rest of the program. 11L 3.1 The PCR (para. 5.01, p. 5) gives the Impression of rapid start-up after effectiveness. ".various project activities were initiated and financed for 1981 since financing retroactive to Loan signing was authoried. ODESYPANO made rapid progress in construction of project infrastructure and Initial staff recruitment. Nevertheless, no disbursements were made until the final quarter of 1982 (PCR, Table 3A, p. 31). The supervision missions which took place before disbursements began, four of them, all reported the project to be problem free or with minor problems. After big disbursements of US$2.1 million in the first quarter of 1983, disbursements averaged 43% of appraisal projections for the test of the projected disbursement period through mid-1987. 32 Just when supervision began to appreciate that the project had serious problems is a bit uncertain. There is no Form 590 supervision report in files for the mission counted as #5 by the PCR and dated February 1983.' In an April memo, the agrostologist, who constituted the mission staf% noted seriously increased unit costs for pasture establishment, roads, and other project endeavors, which called the economics of project impact into question. There was a four-person supervision mission in June 1983 but no report from it in files. PCR Table SB indicates that this 6th supervision mission rated the project as having moderate problems (*2*) but the report of the subsequent mission indicates that mission #6's rating was "I" - problem free or minor problems. Not until the 7th supervision mission of February 1984 do PCR and files agree on a rating of moderate problems, which included management expected cost, compliance with conditions, reporting, expected benefits, and rate of return. The report is in fifes too. 3.3 Rom February 1984 on, there is proper documentation of supervision, except for the 12th mission of March 1987 for which there is no Form 590 in files. For two and a half years, until October 1986, as Table 8 shows, the Bank devoted few resources to supervision. During that period, the project is rated as having from no to moderate problems; no clear picture of the nature of these problems emerges, though 'management" and procurement are mentioned. From October 1986, when a new task manager took over supervision, the project gets low supervision ratings, mainly "major problems. Most of these latter supervision reports are highly critical of project management and pessimistic about the development impact of what the project was doing. The last supervision mission in February 1989 judged that there were major problems with the project's development impact. What made it so? 3.4 What the Audit could gather generally confirms the PCR account ODESYPANO suffered from overly-frequent leadership changes. Some of the leaders were weak. Governments 17 Tabe 8, p. 31. hiring freeze of 1984 hurt ODESYPANO, which was short particularly of field technicians just when it needed them. The other side of this coin was administrative overstafWn&g what the PCR calls "too much centralization of authority and too high a percentage of administrative staW (para. 5.01(b), p. 6) In any case, each of the five governorates which had some area covered by the Office wanted to see action in some of its douars. As a result of their political urging, and despite not having the required field technicians, ODESYPANO expanded its field coverage rapidly and in a superficial way. Technical quality suffered. Participation suffered. 3.5 A loan-financed foreign technician designed an overly-thorough monitoring and evaluation system with onerous data requirements of dubious Importance. ODESYPANO found it difficult to implement this M&E system and, frankly, could see little point in doing so. Basically, M&E did not function for the project's first seven years; its utility since then is questionable. 3.6 This account of ODESYPANO's troubles and the quite negative ratings of later supervisions, coupled with the cancellations of 31.5% of the loan, might predispose the reader to anticipate that the project had little positive impact. As the following section shows, that is not the case. ODESYPANO produced some significant positive results. But the doubts raised in the Bank's review process and the risk pointed out in the appraisal report were also correct; the project failed to realize even the modest results projected at appraisal in its core area. IV. IMPACT 4.1 The contrast between the project's successes and its failures is very neat The pattern emerges clearly despite the unreliability of data Fast, public works on public land were successful. Second, public works on private land were carried out but, in numerous instances, were not sustained. Third, with few exceptions, attempts to change private farming and stock-raising systems failed. 4.2 E t, in execution of works by force account that built public facilities on public land, ODESYPANO did well Unit costs sometimes exceeded appraisal estimates, which is hardly shocking, but the schools, drinidng-water systems, milk collection stations, and access roads to douars got built. Classrooms built were two-thirds of the SAR target. During implementation, ODESYPANO made some very sensible modifications in the drinking-water component, replacing the small number of expensive systems planned with a much larger number of simpler, more cost- effective systems. ODESYPANO built 12 times as many drinking-water systems as anticipated. They * The PCR notes "a swvre defiolency of quantified or quantifiable data" (pars. 6.02(a), p. 7). It also notes *a notable difterece among data provided to supervision missions, those colected by CNEA In the procss of prearing a coMplatio and impact seport for the Office (1990) and thoes provided to the completion mission." The Audit confirms the weaknesses in ODESYPANO dela eported by the PCL Even in 1992, ODESYPANO's M&E systam was not providing useful Infomation. ODBSYPANO's Bd office was much beer. In shost, sporting on physical works constructed does not always Inspire confidence; information on the impact of those wor and of the extent to which they are being anintained is vixtually non- exis4C are appreciated. Forests on public lands, especially the forest belt around Sidi Salem reservoir," got planted and generally grew. (Forest plantings as a whole were less than half of projections.') Access roads built were 62% of target, but these have made quite a difference to the isolated douars whose access to the outside world they improved. In short, thanks to ODESYPANO, a poor, hilly area got some public services, something it had probably not got its fair share of for millennia. 4.3 A execution of works by force account on prIvate lands constructed substantial works but with vry mixed development sesults. As the PCR reports (Table 4, p. 23 and in descending order of target achievement), ravine control by gully plugs and plantings on banks exceeded targets by two and one half times; erosion control works - principally contour rock terraces - met targets; pasture improvement was only one-fourth of target, but the substantial achievement of improvement on 7,000 ha is claimed; and little tree planting was done on private lands. As the PCR explained, This activity has not been well received on private lands due to a lack of producer interest in woodots, to small holdings divided in many parcels with farmers feeling that all land is needed for other production and perhaps lack of agricultural extension...." (para. 6.02(a), p. 8) 4.4 There is sense in the relative acceptance of these four types of works. Acceptance is inversely related to disruption of the farming system. Ravine control doesn't disrupt existing operations because ravines are already lost to farming (and threaten further loss of farmland in a visible and obvious way). There was no resistance by landholders to having gully erosion on their land stopped at public expense. There was resistance to all the other measures, especially to converting to woodlots land, however steep and degraded and subject to erosion, that might still produce some crops. Contour rock terraces were never installed on cultivated land, only zn abandoned or grazing land that could be cultivated once the rocks had been removed. Resistance to pasture improvement might require some explanation; it is not a pure gain for the pasture user because future improvement in carrying capacity requires foregoing use during improvement, something poor livestock owners can ill-afford. (Furthermore, tenancy and use rights were and are complicated, and there are uncertainties as to whom would benefit from pasture improvements.) 4.5 Information on the extent to which these works were sustained is virtually non- existent The Audit saw substantial afforested areas and a lot of ravine training with gully plugs and vegetative treatment that were doing well after several years. The verdict on contour stone terraces is less obvious. Pasture improvement has clearly not been sustained in most cases. ODESYPANO reported to the completion mission that only 1330 ha (19%) of the wonderfully round number of 7000 ha reported as improved actually existed as of 1990. The Audit was taken to the model improved douars which had received blue-ribbon treatment; there, everything that had been built had been sustained except the improved pastures, which had generally been re-degraded to the level 1 Without having any quantification, the Audit wonders whether planting the lovely tree belt around Sidi Salem reervoir was a good investment. The flatter part ofthis land, where the trse grow best, also has higher value alernative uses. General lessons of erosion control suggest that a beter way to reduce siltation in the rservoir would be to start upcatchment, where the sediment comes from. The project meant to do this too, of course, but generally failed (see below). Had it not failed, would the tree belt have erved a usest purpose? = 46% according to Table 4, p. 23 whbh compares SAR estimates for the lst 7 yet with esimaed 7-year actuals. An bcorent impression is giveS by the PCR text (para. 6.02(a), p. 8 which compares actual platings at year seven (4,200 ha. here of. 4,324 in the table) with the 15-year SAR targt of 22,000 ha., giving the false inpression that only 21% of target had besn shMoved. - 11 - of neighboring unimproved pastures. ODESYPANO was able to show the Audit some remaining improved pastures; all were fenced. 4.6 Unfortunately, there is virtually no information on effectiveness of these catchment- improvement measures. As para. 3.5 above pointed out, monitoring and evaluation did not function during the projects first seven years and has not been useful in more recent years when it did function. The loan allocated US$500,000 to research, of which US$1,000 was disbursed; the SAR anticipated that US$ 700,000 would be spent on research, while US$22,000 wasO This official indifference to M&E and research on the part of ODESYPANO management, while it might be understandable in some respects, it particularly unfortunate in the case of soil conservation and catchment improvement works because individual ODESYPANO staff members wanted to cany out such research and even did so on a small scale despite distinct lack of official encouragement. As a result, there is almost no information on the physical effects of reforestation, pasture improvement; rock terraces, gully plugs, and vegetative bank holding (on their effects on sediment displacement, water infiltration, soil carbon, yields above and below the works) that would now provide useful clues to cost-effectiveness of these measures. 4.7 These public works on public and private lands pumped a significant amount of money into a poor area. Government statistics show that consumption in the ODESYPANO area increased at an annual rate of 11% between 1985 and 1990, compared to a rate of 6% for Tunisia as a whole. ODESYPANO spending was not the only factor in this differential, but it was probably the major one. 4.8 ODESYPANO's infusion of cash and World Food Programme food into the project area was not a give-away program. The Audit asked ODESYPANO's soil & water conservation engineers about the participative aspects of the project. "But of course they have participated, said the engineers; "we paid them, and they carried the stones and planted the seedlings. 4.9 Tbrd, wherever ODESYPANO tried to change ierm and stock-rearing systems, results ranged rom negligible to modest. PCR para. 6.02 recounts much of the story. Farmers were rarely willing to stop cultivating steeper slopes. There was little or no intensieation of production on gentler slopes. Less than 1000 ha of improved grazing land has stayed that way. Livestock owners also sometimes failed to keep their stock off vegetation planted along ravines, resulting in destruction of some of these works. Since forage supply did not increase, project efforts to increase livestock numbers may have been counterproductive. No genetic improvement was achieved in sheep because ewes financed by the project were no better than local ewes. Farmers rejected tree-growing on their own land except for olive and almond trees. All of the vegetative contour bunds disappeared. The 21 Prom PCR Table 3B, p. 22 and SA, p. 24. a From the 11th supervision mission of October 1986, Bank technfcians suggesled that ODESYPANO investigate the possibility of using Vetiveria ianioides lines planted at contour as a substitute for other vegetative and for stone contour barriers. ODESYPANO staff we slated to go to India to sa how this technology was being used ther, but they never did and ODESYPANO never got slps of velver for multiplication and local trial. When the Audit visited ODESYPANO, this technology transer had not yet been achieved, with each aide feeling it had made a good-faith effort, which had been frustrated. Th project ae is ilose or perhaps beyond vetiver's cold tolerano lmit. Theafob, it is not clear how well this expanding, low-ost technique for cresigainfall infiltration and slowing nn-off would do Ia the project area. Vativer might do well at low altitudes with less and less-frequent frosts, but not at higber ones. If so, in these areas, votiver might prove - 12 - most widely accepted soil conservation work that affected farming systems was stone contour bunds; these were probably accepted because the stones for the bunds were amassed by de-stoning the adjacent land, making it arable. 4.10 This litany of disappointments makes it sound as if the project achieved no system changes at all. That is not so. Milk and livestock production in Sejoane, on which the German *pilot for ODESYPANO was based, continued to expand both by intensification in its original area and by modest expansion into adjacent areas. Animal products account for 75% of the projects incremental benefits as calculated by the PCR (Table 9F, p. 36), milk for 29% and meat for 46%. The other quarter of incremental benefits were from agriculture, 7% from chickpeas and 18% from wheat and barley (production of which had been expected to fall by the SAR). 4.11 While the agricultural results point up failure to change the cropping system, the livestock targets are not far from what the SAR expected after 7 years. Recall that the SAR was extremely prudent about expecting rapid changes. On this basis, the PCR calculates an economic IRR of 6% for the 7-year project, as compared with the SAR's 7% for a 5-year project, worse but non disastrously so. 4.12 Overall ent The critical finding of the FAO/CP's PCR team, accepted by the Bank's Regional Office in Part I of the PCR, is that the innovations ODESYPANO has introduced will not "make oil spots: that they have not demonstrated their effectiveness so that farmers and stock-raisers will progressively adopt them. On this basis, the PCR projects a steady state from 1989 through 2012. No further investment - so further spread of the technology, hence of growth and development. This finding is not all that different than what the SAR projected. Unlike the SAR however, the PCR did not speculate on what would happen if there were further investments to build on the costs sunk into ODESYPANO. Part I simply said, if there were no follow-up investment, then the economic return over the long run would be moderately unsatisfactoiy, just as the SAR had predicted but a little more so. 4.13 As Part I of the PCR indicates, and as Borrowers' representatives told the Audit, many in Tunisia feel that the project, despite its diffculties, is on the track projected by appraisal - a worthwhile, slow-gestating, long-range system change. They were surprised by the PCR team's finding that, overall, the project has been unsatisfactory. 4.14 Judging a project simply satisfactory or unsatisfactory is nowhere more of an oversimplification than in the case of ODESYPANO. At one extreme, even if all its farming systems and catchment improvement ambitions fail, ODESYPANO delivered durable public services: drinking- water, access roads, some schools, rural service centers that collect milk and provide some other services. Of course, it could have been a leaner, more-fficient instrument for delivering them without its other ambitions. But then some of those ambitions have been realized. The project had modest appraisal catchment-improvement and production objectives for the early years. The PCR found after seven years and the Audit confirmed after ten that a modest part of these objectives had been achieved. superior to vegenaive contour plaUntWp ud so far, which bave zro accepwe. - 13 - 4.15 On baan the Audit confirms Part I of the PCR's findng that the project was marginally assatisfectory. The reason is that the project's core - farm and livestock system changes - - proved to be so weak. The &oil conservation measures that worked were those to alleviate the CK" of non-sustainable laud use further up the catchment, such as ravine containment, the tree barrier to catch silt on the edge of Sidi Salem reservoir, and rock terraces. What didn't work were efforts to change land use further up the catchment in ways that would prevent the erosion and soil degradation in the first place by attacking their g Without those changes, the rest of what ODESYPANO did will be uasstalnable palliatives in the long run. The complicated reasons why ODESYPANO's proposed system changes generally failed to catch on are discussed under Issues and Finns below. 4.16 Insitutional deMlgpmnt. The Audit concurs with the PCR that ODESYPANO was an inappropriate instrument to its complicated task. The PCR stresses the paucity of field staff, particularly of extensionists, and a central-office orientation so strong that even the few extensionists "devoted less than 20% of their time to md bog extension" (para. 6.02(b), p. 9) and kept having their vehicles taken by higher authorities for other uses. In 1992, the Audit found ODESYPANO still overly centralized for an organization with its charge, though apparently less so than at PCR time. The Audit was particularly impressed by the ethos of ODESYPANO, which was strongly amdnagiste," that is, infused with the spirit of the French "andeagement du territoire" tradition, a combination of soil and water conservation, regional planning, and civil engineering. ODESYPANO's CES engineers planned for the better use of the project area's resources, then implemented the works' part of the solution, using local labor where possible and paying the population to participate. 4.17 Many of the engineers at ODESYPANO's Bdja headquarters believe that, with patience, this approach will work. After all, the SAR predicted as much. But since loan 1997 closed, while ODESYPANO has been maintained at a reduced level by Government funds, GTZ assistance, and "bridging' finance from IBRD under the 2nd Agricultural Structural Adjustment Loan, the need for a modified approach has come to be widely recognized. 4.18 The institutional framework has been complicated by devolution of most of the Agriculture Ministy's local functions to the governorates and to CRDAs (Regional Agricultural Development Commssion The CRDAs, for instance, are supposed to take care of extension. Where does this leave ODESYPANO? It is a national anomaly. It is also true that the CRDAs and the governorates, whose administrative boundaries are not catchment-based, cannot easily deal with catchment problems. In theory, CRDAs are meant to provide extension service to all of their districts; in fact, plains areas get all of what little is available and the hills areas that ODESYPANO covers gets no extension from the CRDAL Now that BNA (National Agricultural Bank) centralizes formal farm lending, the project area, where legally-registered land titles are rare, doesn't get any. 4.19 Government is aware of these problems and institutional arrangements are under discussion. The Audit doesn't presume to know what institutional solution would be best for the hilly parts of northwest Tunisia and only wishes to reiterate that hill areas will not get public services unless some Institution or institutions make special provisions. 4.20 Not only Government of Tunisia, but GTZ, FAO/CP and the Bank agree on re- orienting ODESYPANO or whatever organization takes over its functions so that its "andnagiste" culture will be modified to elicit more people's participation. Government hired consultants who produced an elaborate partnership promotion study in June 1991 and a practical rural communities' - 14 - participation study in August 1991.r Since then, seminars for ODESYPANO and other ministry staff have been held to discus what to do. While some of the "partnership proposals appear to the Audit to be unnecessarily complicated, there has definitely been a "prise de conscience" about the problems of rural development in the northwest hills. Whatever happens, it is unlikely that future public efforts will be as directive and as top-down as those in the past. Participation is likely to involve more than putting stones where you are told and getting cash and food rations in return. Therefore, ft does not seem too sanguine to expect that the institutional development that ODESYPANO represents will be converted into something better adapted to egioal development needs. V. ISSUES AND LESSONS 5.1 The Audit focusses on two issues why expected farming-system changes failed to take place; and preconceptions that guided project design. The first is about the ecosystems and specific differences within them in the project area. The second is about Bank performance in its interaction with the Borrower. The Audit then draws four lessons from the issues discussed. 5.2 Why No Farming-systems Change? The big issue, to which all others are subordinate, is why didn't farmers and stock-raisers in the project area adapt their farming and livestock systems to make their practices environmentally sustainable as per project recommendations? Many Tunisian professionals and some at the Bank feel that there is nothing wrong with the prescription. They counsel patience, pointing out that the appraisal fully recognized that such changes take time, that, from the viewpoint of rural people who were only partly farmers, the proposed system changes were only marginally attractive so that getting adoption would require a great deal of encouragement. Further to that point, ODESYPANO was handicapped by a shortage of field personnel, especially extensionits. Many feel that, with a participative, partnership approach, the population could be persuaded to buy the project's vision for farming, animal husbandry and forestry in the area. 53 The Audit is less sanguine. The idea that what has slowly happened to turn Sejnane into a little Switzerland could be diffused to the rest of the area is based on a fundamental misunderstanding of the ecological uniqueness of different parts of the project area. Diffusion there may be, but it will be guided by ecological subtleties which the project missed. 5.4 Pastures around Sejnane, once improved, generally stay that way. They are fenced. Improved pastures in the rest of the project area are quickly overgrazed and redegraded, except for a few where ODESYPANO has taken control and fenced them. It is dairying at SeJnane that gives enough margin to pay for fencing. Groups elsewhere in the project area would no doubt also like to fence pastures and exclude others. They don't though because the traditional grazing they do on them isn't lucrative enough to make it worthwhile. Fencing goes with a sense of the value of pasture, 2 Molaunued Fioul. *fudedo promotion du padenarlat.'Tnis: nstibut do Gestion, June 1991, and MoncofZgidi. La pasticipation des commnatls ureles.' Tunis: Cabinet d'Eudes at do Consultation Ifuikla, August 1991. X Eves ia Sejns, not all pastues stay improved. How well land is used appeas to depend on the cohosiveness of t group to which it belongs. Sometimes a group whose land is protooled by a fence has a had time acting together because of conflict betwe fteilis that ae preseat and families all or pat of which are absent. - Is - hence with adjustment of animal numbers to the carnying capacity. Transition of grazing land from semi-open-access, common property to closed-access, small-group property, as at Sejaane, sets off a series of system changes that are good for the catchment and for the people in it This is the core on which the project was built; it had started happening at Sejnane and that development was re- enforced during the project period, and since. 5.5 The project built 21 milk collection centers to diffuse dairy successes, especially to the Kroumirie-Mogods. The PCR reports that all but 8 had been closed by 1989 because they couldn't meet operating costs and amortization, and that only one-third of the milk processed by the project was from small farmers.u In 199Z project authorities reported to the Audit that 11 milk collection centers were operating, four of them year round, a sign that dairying is well developed, and seven only part time for an average of eight months per year. Clearly, the white revolution was proceeding very slowly. Why? 5.6 The NWRDP optimists point out that the change was slow to develop at SeJuane and still is not exactly galloping, even though supported by the best of the Tunisian staff and heavy German technical assistance to help organize and strengthen groups, inter alia. But once something is proved profitable, it usually spreads rather rapidly to adjacent areas with similar conditions, often with little or no extension or technical assistance. 5.7 The key to slow diffusion is Sejnane's ecological uniqueness, a fact either not known or overlooked by appraisal and supervision missions. Seinane's people were always more animal- husbandrymen than farmers. After the winter rains, the valley used to contain a shallow temporary lake which gave way to beautiful pasture when it dried. This and relatively low rainfall for the north made Sejnane unique: more suited to stock-raising and with an agriculture based on spring sorghum, the only such zone in Tunisia. 5.8 The people of Kroumie-Mogods, Oued Zarga and the Tessa Infdrieur, when they could fully make a living from their land, were mainly wheat-barley-chickpea farmers who also kept sheep. Turning them into dairy farmers requires a much bigger system change than occurred in Sejnane. In areas where there is still abundant forest, as around AMn Draham, farmers can graze their stock on State forest land.' Since there is a commons to graze, there is less urgency to improve pasture, hence little hope of selling the projects system change. 5.9 Where forests are gone and mountaintops are exposed rock devoid of vegetation, as on the road north from Bdja to Nefza, pastoralists have no commons to graze. One sees pick-up trucks loaded with hay winding their way up to the accessible hill douars; hay is expensive. It is the Audit's impression, based on a short field visit, that it is in such regions that milk collections are starting to pick up. Necessity being the mother of invention, it is where the commons has been destroyed that livestock raising has to change, which may make improved pastures on the upper slopes a possibility. 25'w iee was taken from stats and large, private, commeral farms to give t nter enough though-put to keep thmw open. * The Fores Servin seomingly dom an impressive job of preserving ste forests from IIIngal logging and harves of orewood and charvoal. However, these forests, which arc heavily graed, are noteworthy for the absenceof young tries in the undertory. Under such condidone, foesta do not sustain and new thmselves and have to be replaled.* - 16 - 5.10 But not an easy change to promote. Unlike Sejane residents, douar residents in Kroumile-Mogods grow wheat wherever they can. The Audit heard from farmers and from extensionists of situations where ODESYPANO convinced farmers to stop growing wheat on steep, eroding slopes until favorable rains came and they changed their minds. Incentives to group action would have to be strong too. Land tenure disputes seem to be the regional sport, one which takes place outside the courts because the tenure that counts is informal. 5.11 This discussion has been confined to reasons why the proven Sejuane dairy system changes will be hard to diffuse. How much more difficult, then, the unproven farming system changest Farmers will not give up farmland to stone contour lines. No formula for sustainable vegetative bunds on farmers conditions has yet been found. Contour plowing or even plowing along the lesser slope is difficult under project-area tenure patterns, horizontally-narrow vertical strips that give each inheritor a portion of land of every fertility and moisture. There is no proven system to diffuse. Prospects for systemic change are remote and dependent on further experimentation 5.12 In short, the project's expectations for farming and livestock system changes would have been seen to be unrealistic if basic ecological conditions in the projects micro-regions bad been understood. 5.13 Bank Performance: Proict Desian Preconceptions and Retrosoective WiTdoM Since eight years passed during which the Bank considered this project before the loan became effective, the Bank hardly rushed into this project! Yet, as indicated above, the long processing period is marked by a distinct reticence about the project on the part of technicians, both Tunisia's Agriculture Ministry and the Bank's agriculture division, in marked contrast to the enthusiasm of the policy mavens: Tunisia's planners and the Bank's Programs' division. In project files, the teclnicians' reservations are rather subdued. However, technicians involved in the process in Tunis, in Washington, and Bank retirees whom the Audit contacted expressed stronger reservations. Their reservations were over-ridden because the Bank and the Borrower wanted a poverty-based rural development projecL When a new and articulate task manager took over project preparation from the rural engineer who had preceded him, NWRDP passed through Bank reviews with little concern about substance. 5.14 It would be churlish to throw stones at the technicians who designed this project. The design shows much better understanding of the area to be developed than most area-development projects. There was a core technical package, and it is a plausible one. Its only marginal attractiveness to rural people who live as much by remittances and income from rural works projects as from working the land was recognized. If the core technical package had diffused even a modest amount, all shortcominp would have been forgotten. But is has not Therefore, we look more carefully for clues ignored and hints not taken. 5.15 The Borrower and a number of Bank technicians recognized the "limited potentiar of the project area (para. 2.2). The Borrower resisted the Bank's desire for an integrated rural development project (paras. 2.3,2.9-10). Nevertheless, it was the Bank's vision of such a project that 2 Bank per ance on anaers o=8 er t a project design se discussed elsewhere in his Audk notably puas. 3.2 & 3.s oan suprAvilon and pa.m 4,12 on PCt ating ofre projecL - 17 - gradually prevailed at the issue of lengthy Bank and FAOICP contact with CNPA. The Bank has a template into which it made the project area and the Borrower's wishes fit. 5.16 The Borrower resisted the Idea of an autonomous project agency (paras. 2.11, 2.15), but ultimately agreed to one. CNEA had wanted to use the Livestock Service for project implementation. The Bank's complaints about too much centralization and too much administration, as well as the Borrower's delays in giving ODESYPANO its full complement of field staff (paras. 3.4, 4.16) are best seen in this context. 5.17 In the light of disappointing adoption nf the project's core systems package, during the bridging period and in discussions of possible follow-on development financing (para. 4.17), the Bank has been urging Tunisia not only to make ODESYPANO more participative (para. 4.20), but to lighten and shrink it and to transfer some of its functions to other institutions (para. 4.18). 5.18 For the regional land-use transformation it was expected to preside over, ODESYPANO's size was appropriate enough. The public facilities and palliative soil-conservation works which it built could have been built by a leaner organization. Based on actual results, the Borrower's reluctance to create a large new agency that would be hard to dismantle later was appropriate. The Bank has executed rural infrastructure projects elsewhere exclusively through line agencies without creating new institutions." But this is wisdom based an what we know has happened over the last twelve years, a luxury NWRDP's preparers and appraisers did not have. 5.19 In retrospect, Tunisia would be better off if NWRDP had concentrated on building rural infrastructure, on some palliative soil-conservation works, and on action research to explore what land-use-system changes farmers would accept. That could have been done at half the actual cost and with a lot less buiding of an institution that is now difficult to down-size. In view of the clues and hints, why did the Bank, in particular, push for the actual project design? The Audit concludes that it did so for two reasons. Firtly, there was genuine hope that the land-use system changes at the core of NWRDP would catch on, bringing good economic returns and putting an end to the destruction of the hilly Northwest's ecosystem. Secondly, the Bank was going by the book, squeezing the Borrower's initial presentations into the Procrustean bed of its preconception of what an integrated rural development project should be." 5.20 Lessons may be summarized as four points: - however laudable the desire to alleviate poverty in a region, without a sound technical package, there can be no sustainable improvement; - there is no substitute for a thorough understanding of ecological and agronomic conditions when judging technical diffusion; a BxampsaeTHAAND- Nozthastanamldvo project (Lu. 1198), audited by OED Report 5990 of December 1985, and KOREA - Rural Infrastruture Project (Las. 1216 & 1218), audited by ORD Report #3925 of May 1982. * At the time of NWRDP's procoeag and approval, it was common for tie Bank to lead for infrastructumo costruction in povaty-orlated asban projects. Povay-orlAnted rnal projects, however, we expected to find a technical cos that would generate enough resourmes so that they would Opay their way.* - 18 - * for farmers to feel part of an enterprise, participation has to tavolve more than getting paid for prescribed manual work; and - it is easier to provide rural public works and services than to get farmers and livestock- kespers to change their systems, but the latter Is more important - 19 - ATACHMENT 1 Page 1 of 2 WORLD FOOD PROGRAMME TUNIS FAX NO. 68 20.05.93 TO: MR. GRAHAM DONALDSON CHIEF, AGRICULTURE AND HUMAN DEVELOPMENT DIVISION OPERATIONS EVALUATION DEPARTMENT, WORLD BANK WASHINGTON, D. C. U.S.A. INFO: MR. QUAZI HAQUE REGIONAL MANAGER, ODM/WFP/ROME FROM: AUGUSTIN GUY GUNUBU DIRECTOR OF OPERATIONS, WFP, TUNISIA SUBJECT: TUNISIA NORTHWEST RURAL DEVELOPMENT PROJECT LOAN 1997-TUN Thank you for your letter of April 20, 1993 by cover of which you sent a copy of the Bank's draft Performance Audit Report on the Tunisia Northwest Development Project. As requested below, please find our comments on the document: WFP assistance to project Tunisia 2582 "Integrated Development in the Northwest" was originally approved for 18.8 million dollars (45,750 tons of wheat, 1,830 tons of vegetable oil, 2,745 of DSM, and 1,830 of sugar). The value of these commitments at the end of the project was 15.0 million dollars. The figures of 1.6/1.7 million appearing in the text and under "Appraisal/Expectation" in Table 2 "Key Project Data" of the draft report should therefore be amended.1 The WFP interim evaluation report (November 1986) of project Tunisia 2582 draws attention to many of the weaknesses noted in the PAR, namely: weak realization of physical objectives; difficulties of changing cultivation systems owing to systems of inheritance favoring long, narrow parcels of land; worknorms which didn't take sufficient account of specific physical conditions in different zones; similar criticisms of ODESYPANO such as the lack of effective extension work and monitoring; and the limited participation of farmers. WFP is therefore in agreement with the auditor's findings. WFP, however, supports the view that in time farmers will come to appreciate the value of I WFP project Tunisia 2582 covered a much larger area than that covered by the World Bank project supported by Loan 1997 and including, notably, the area of the parallel IFAD project (see par. 2.8). Therefore, only part of WFP project Tunisia 2582 constitutes a cost of and a financial contribution to the project audited here. - 20 - ATTACHMENT 1 Page 2 of 2 the changes in land use patterns which the project has tried to introduce. WFP is also fully aware of the importance of adopting a participatory approach in order to ensure project sustainability. Consequently WFP, in collaboration with the government, other donors (World Bank, GTZ, EEC, KfW) and various NGO's, is currently providing assistance, valued at 5.4 million dollars. In support of a new five year project, TUN 4783 "Rural Development in the Mountainous Regions of the Northwest", which is based on a fully participatory approach. We trust that these comments are of assistance. TUNISIA NORTHWEST RURAL DEVELOPMENT PROJECT e MVediferranean 5ea Project ara componmnts Kroumirie - Mogods Tesw Infr0 10 20 30 40 50 BIZERTE Governorate boundaries Mometes - Commune boundaries ---Cheikhat boundaries Governorate headquarters Menel Commune headquarters L e --- Paved roads -- Unpavd roads; - Jgt Railways T mSeinane Rivers teu - - International boundary o M e 3737· -~Tabo ez to I *;mn STo Tebourba Aïin Draham KSSEB - DAM C HammamBougolba SENMEfTRDAM ' BEJ Ao- ALGERIA BOU HERTMA DAM I ejez el Bob * Fernana Bou Testour -36°30' - 63 TEN U eboursouk Ghordimaou Kr'b i GBOU Ara MELEGUL EL 50t El KR 8 SakietSi Ybø ef Ohw EDt *n b *** t K EF bw esson W$ -ap enot ømpfy, on ed the *ZAAF N£ jWaddentatflite& any aenry ~fenfdoreent Se s - SILIANA 9* 30' o acc.~neo ucaudr zerme 34r00* 3 *00 TvS AR EA Dahmani* OF Ain Mosria MAP Mokhto TUNISIA Koloat Es-Sonam ALGERIA\ Shb YSbibA 3513 S .30 rer se 30 \KToKeinø # To5beit/ø 83 9wG -TUNI SIA, NORTHWEST RURAL DEVELOPMENT PROJECT ' 1k KOMETERS RAINFALL Avera Annual Rainfall M eters, CID,Project ara componenft 0-300 Paved roade 0 500-350 Unpaved roads 350-400 -+-+- Railways 400-500 Governorate headquarters 500-600 --- oenrt boundarles-- 600-800 -- International boundar 800-1000 1000-1200 1200-1500 Over 1500 7001 - - - -Tobourba' 00 j A nODraham < BEJ ALGERIA ci OMoez el Bob Bou Smef 0/ Tosfour ENDOUB Ye36rX- OTeboursouk SGbardimao^ 100 ' 'El Krb 0 uradhé - 0 Gaafour (Sakiet Si me con of me r of Youssef Me ~ t w~ it is at~. Jb~wdis~mhotos~ a EL KEF d~ ,» od do not ~mly oitett offi &he SILIANA _30' aw~_ of____ b&~. ) •O 36p00m- Tun se Dahmani . r MAP OMakhtor erissa Kalaat Khosbo .og. a ~Rouhia ALGERIA \ osbiba (.LIBYA 35*30 353~ To robess s. \ ro Kosserine a ådTo Sheile