World Bank experience with irrigation development : socio-economic, institutional and technical impact and lessons (Vol. 1 of 3)

Report No. 7876 World Bank Experience with Irrigation Development Socio-Economic, Institutional and Technical Impact Lessons (In Tbree Volumes) Volume 1: Overview Mexico: Rio Sinaloa and Panuco (Loans 970- and 969-ME) and Morocco: Doukkala I and II (Loans 1201- and 1416-MOR) june 15, 1989 Operations Evaluation Department FOR OFFICIAL USE ONLY Document of the World Bank 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 authorization. 1M OICI E U ONLY THE WORLD BANK WashaUgton. D.C. 20433 U.S.A. Ofre of DiaectarAeral Op*atsons Evaluathus June 15, 1989 MENORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: World Bank Experience with Irrigation Development Socio-Economic, Institutional and Technical Impact and Lessons Mexico: Rio Sinaloa and Panuco Irrigation Projects (Loans 970- and 969-ME) and Morocco: Doukkala I and II Irrigation Projects (Loans 1201- and 1416-NOR) Attached, for information, is a copy of a report entitled "World Bank Experience with Irrigation Development-Socio-Economic, Institutional and Technical Impact and Lessons" prepared by the Operations Evaluation Department. Yves Rovani by Ram K. Chopra Attachment This documaet has a restricted distribution and may be used by recipients only in the performance of their offcial duties. Its contents may not otherwise be disclosed without World Bank authoriandon. FOR OFFICIAL USE ONLY WORLD BANK EXPERIENCE WITH IRRIGATION DEVELOPHMENT SOCIO-ECONOMIC, INSTITUTIONAL AND TECHNICAL IMPACT AND LESSONS VOLUME I OVERVIEW MEXICO: RIO SINALOA AND PANUCO IRRIGATION PROJECTS (LOANS 970- AND 969-ME) AND MOROCCO: DOUKKALA I AND II IRRIGATION PROJECTS (LOANS 1201- AND 1416-MOR) TABLE OF CONTENTS Pase No. VOLUME I: OVERVIEW PREFACE ......................****........**. **........ I. BACKGROUND ......*................***...........A..... 1 A. Introduction ..... . ................. 1 B. The Projects ...................................... II. AGRICULTURAL AND ECONOMIC IMPACT ....................... 3 III. THE DEVELOPMENT APPROACH ................... ...... ..... 5 A. Land Tenure ........ .................... ........ 5 B. Institutional Effectiveness ............. ......... 7 IV. PERFORMANCE OF THE IRRIGATION AND DRAINAGE SYSTEM ...... 9 A. Status of Infrastructure ............*.............. 9 B. Water Efficiency ............ *4**.... ....... . 9 C. Operation and Maintenance .......................... 10 0. Irrigation System Design and Project Performance ... 13 V. SOCIAL IMPACT ............................ 15 VI. IMPACT ON ENVIRONMENT .............................. . 17 ANNEX 1 - Main Features of the Sinaloa, Yaqui, Panuco, Doukkala, Sidi Smail and Sidi Bennour Irrigation Projects MAPS: IBRD Nos. 21643, 21186 VOLUME IIt MEXICO PANUCO AND RIO SINALOA IRRIGATION PROJECTS VOLUME III: MOROCCO DOUKKALA I AND II IRRIGATION PROJECTS This document has a restricted distribution and may be used by recipients on!y in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. WOML BANK= :IT IRRIGATION DEVELOuBnT SOCIO-ECONOMIC, INSTITUTIONAL AND TECHNICAL IMPACT AND LESSONS OVERVIEW MEXICO: RIO SINALOA AND PANUCO IRRIGATION PROJECTS (LOANS 970- AND 969-ME) AND MOROCCOs DOUKKALA I AND II IRRIGATION PROJECTS (LOANS 1201- AND 1416-MOR) PREFACE While agriculture has been the most important sector supported by the Bank over the past fifteen years, irrigation has been the leading sub- sector of the agricultural portfolio with 312 of total lending for the sector. OED is currently engaged in a series of iDact evaluations of irrigation projects as part of a wider study analysing World Bank experi- ence with irrigation development and the effectiveness of different types of irrigation development throughout the world. The Agriculture and Rural Development Department of PPR provides support through participation of staff in missions, review of the technical engineering aspects, and assistance in dissemination of key findings to Bank technical staff. The methodology employed in Impact evaluations is particularly appropriate for Irrigation projects which are difficult to evaluate at loan closing when irrigation infrastructure is newly completed and agricultural production is in its early stages of development. Similarly, the long-run effect on poverty alleviation and equity of land consolidation and agrarian reform, which are often part of irrigation projects, is not apparent at project completion. Returning to the project at full development, there- fore, allows the technical, agricultural, social, environmental and insti- tutional Impacts of investment in irrigation to be measured and lessons drawn from project experience. At the same time, and despite the extent and range of Bank invest- ment in the irrig-ition subsector, little analysis has been undertaken to date comparing different engineering approaches to irrigation system design in similar physical conditions; more particularly in determining the rela- tive strengths and weaknesses of these varying approaches as regards actual project performance. Accordingly, OED has selected four countries for impact evalua- tion: Mexico, Morocco, the Philippines and Thailand, all of which are Important in terms of irrigation development. For the purposes of com- parison, Mexico and Morocco are reviewed together, since both characterize the use of advanced technology in semi-arid conditions. The second group will comprise the Philippines and Thailand, which represent the use of less intensive technology in tropical countries for rice production. Two proj- eacts have been identified in each country, eight projects in total. Simi- Diraopvrms Prmees 18-J-WH 14:41:00 - ii - lar impact evaluations are also under way for groundwater development in India and Bengladesh. This particular report, therefore, comprises two impact evalua- tions for Mexico and Morocco (Volumes II and III respectively) and this overview (Volume I) which synthesizes the findings of both impact evalu- ations regarding the extent of, and reasons for, sustainability and also compares the impact of irrigation system design on project performance. In Mexico, the two projects selected for impact evaluation were the fifth and sixth of sixteen Bank loans to Mexico totalling US$825.5 million for irrigation development. Both projects were approved by the Board on the same day (February 19, 1974) in the sum of US$77.0 million for the Panuco project and US$47.0 million for Rio Sinaloa. A supplementary loan of US$25.0 million was approved in July 1978 for the Panuco project. This impact evaluation also reviewed the follow-on project in Sinaloa (Rio Fuerte/Rio Sinaloa II Irrigation Project, Loan 1706-ME) approved by the Board in May 1979 for a total of US$249.7 million, which is now nearing completion. The Yaqui irrigation project in the nearby Sonora State, although not financed by the Bank, has also been reviewed for purposes of comparison. The impact evaluation for Mexico is based upon a review of the Staff Appraisal Reports of Sinaloa I and II projects (No. 281-ME dated November 14, 1973 and No. 2454 dated May 7, 1979) and of the Panuco project (No. 300a-ME dated February 4, 1974); the Project Performance Audit Report of the Panuco and Sinaloa I projects (OED Report No. 5191 dated June 29, 1984) and other relevant documents. A survey of 262 project beneficiarIes of the Panuco project was undertaken between November 1987 and March 1988 by the Colegio de Postgraduados of Chapingo, Mexico to determine the com- plex social and financial impact of the project on the different categories of beneficiaries. Two missions were undertaken by OED in August-September 1987 and February-March 1988 at which time mission members interviewed government officials and project authorities in Guasave and Tampico as well as farmer organizat.ons and project beneficiaries. In Morocco, the two projects were the fourth and fifth of seven loans for total lending of US$241.0 million for the irrigation sector. Doukkala I was approved by the Board on February 3, 1976 in the sum of US$30.0 million while Doukkala II was approved on May 3, 1977 for US$41.0 million. As in Mexico, the impact evaluation has also reviewed the experi- ence of the Sidi Bennour and Sidi Smail projects, not financed by the Bank, but also part of the ongoing On er R'bia Water Master Plan. A follow-on project, the Large-Scale Irrigation Improvement Project (Loan 2656-MOR) in the sum of US$46.0 million and approved by the Board on February 25, 1986 was also included in the review. The impact evaluation for Morocco is based upon a review of the Staff Appraisal Reports for Doukkala I and II respectively (No. 915a dated January 6, 1976 and No. 1431a dated April 12, 1977) as well as that for the Large-Scale Improvement Project (No. 5846-MOR dated January 22, 1986); the PPAR for Doukkala I (OED Report No. 4236 dated December 22, 1982) and other relevant documentation. Two socio-economic surveys were undertaken sul- taneously by the Institut Hassan II, Rabat, between September-December 1987. The first interviewed 196 project beneficiaries, while the second D rovervem/as Prefece:no 18-JU4-89 14:41:00 - III - interviewed 152 women, of whom 114 were wives of the above interviewed farmers and 28 were heads of household. The surveys were to provide infor- mation on the social Impact, changes in land tenure and to determine how the projects had changed the status and role of women in the project area. Two missions were undertaken by OED, the first in June 1987 and the second in November of the same year. Visits were made to the project area and interviews conducted with project beneficiaries, project officials in 91 Jadida and government officials in Rabat. The valuable assistance provided by the governments and all per- sons contacted by the mission in the preparation of this report is grate- fully acknowledged. Dirovervm/s* Preface:no 18-AW-89 14:41:00 - 1 - WORLD BANK EXPERIENCE WITH IRIGATION DEVELOPHENT SOCIO-ECONOMIC, INSTITUTIONAL AND TECHNICAL IMPACT AND LESSONS VOLUME I OVERVIEW MEXICO: RIO SINALOA AND PANUCO IRRIGATION PROJECTS (LOANS 970- AND 969-ME) AND MOROCCO: DOUKKALA I AND II IRRIGATION PROJECTS (LOANS 1201- AND 1416-MOR) I. BACKGROUND A. Introduction 1. Irrigation is important to agricultural and economic growth in both Mexico and Morocco. Mexico ranks the sixth country worldwide in terms of irrigation development; irrigated areas provide about one half of the value of total agricultural production and about 65% of agricultural exports. Productivity has improved only slowly, however, while cropping intensities have been below expectations. In Morocco, area under irriga- tion has been growing rapidly in recent years; the subsector supplies about 45% of agricultural value added and 60% of agricultural exports and con- tributes to import substitution objectives. Agricultural performance in irrigated areas has generally been good. Output has increased 5.5% p.e. between 1970-85, the period in which the projects were implemented, although cropping intensities varied widely as between the large irrigation schemes. 2. In both countries, existing legislation allows for changes in land tenure in irrigation development. In Mexico, Gcvernment has considered investment in irrigation as an important tool in achieving equity and income distribution objectives through agrarian r.orm. Legislation per- mits Government to expropriate land for irrigation development wLich is then granted in perpetuity to ejidos--communal organizations of landless peasants--each of whom is given on average a 10-ha irrigated plot which is to be cultivated individually or collectively depending on local arrange- ments. In Morocco, changes in land tenure are intended to both reduce land parcelling and achieve technical objectives: a highly structured and closely managed strategy in which land consolidation is closely integrated with the design and layout of the irrigation system and, at the same time, with strong support for, .and control of, on-farm management. B. The Projects 3. The four projects were designed to ensure efficient water control, with adequate water regulation systems and a relatively sophisticated irri- gation infrastructure, typical of arid and semi-arid countries. The systems were designed for gravity irrigation in Mexico and sprinkler in DIR:0VBie/einln9e spaced overview eng 14:10:00 29-APR-0- - 2 - Morocco. The four Bank loans were approved in the 1973-76 period, and were to be fully disbursed in the early eighties. 4. The two projects in Mexico, Rio Sinaloa on the western Pacific Coast and Panuco on the Gulf Coast to the east, constituted large-scale surface irrigation development of more than 100,000 ha each and included the construction of storage and diversion dams; main canals with upstream water control, lateral and sublateral canals, land levelling, drainage systems, and provision of operation and maintenance equipment and of agri- cultural services to farmers. Irrigation was previously practiced on almost half of the project area in Sinaloa, a semi-arid region with an average annual rainfall of 550 mm. Irrigation was new for Panuco, where livestock was an important activity in an area with an average annual pre- cipitation of about 900 mm. Agrarian reform objectives were to be achieved through expropriation of about one quarter of the project area of Sinaloa; in Panuco, almost 80% of the total area was to be expropriated and simi- larly redistributed to landless people. Both projects were approved in 1974 to be implemented over a six-year period by the Ministry of Agricul- ture. Its district offices would be responsible not only for construction and operation and maintenance of civil works, but also for provision of extension services to farmers. They would also liaise with the Ministry of Agrarian Reform. Project costs were estimated at US$145.6 million and US$197.4 million respectively. 5. Implementation of the two projects experienced substantial delays, largely due to lack of detailed engineering designs at appraisal and bud- getary constraints. Both projects were substantially redesigned. Addi- tional financing of US$25.0 million was provided by the Bank in the case of Panuco with an area expansion of 7,000 ha. For Sinaloa, the project scope was reduced to the construction of storage and diversion dams and the con- struction of an irrigation system covering only about 29,000 ha. A follow- on project, Sinaloa II was approved in 1979 to complete the irrigation and drainage system of the entire 105,000 ha. Agrarian reform was successfully accomplished in Panuco; almost 70% of the area was redistributed to ejida- tarios. The maximum holding size which could be retained by existing land- holders and colonos (settled in the mid sixties) was 20 irrigated ha. In Sinaloa, however, where 25% of the area was expected to have been expropri- ated, this did not take place; most farmers already had established water rights in the project area. The settlement of 600 families displaced by the dam was completed as planned. 6. At project completion of Sinaloa I, yields and cropping intensi- ties on the 29,000 ha equipped with irrigation and 16,000 ha benefiting from increased water availability were higher than expected at appraisal. Progress in agricultural development was poor in Panuco, mostly due to the poor performance of the newly established ejidos. Only about 35% of the total project area was irrigated at loan closing in 1982 and both yields and cropping intensities were lower than expected. Cost overruns continued to be significant for both projects at closure: 46% in real terms for Sinaloa and 126% for Panuco. Total project costs were US$250.0 million and US$446.7 million respectively. The combined project costs of Sinaloa I and II were similarly revised upwards, attaining US$440.0 million when the project was to be completed in 1985. The revised economic rate of return (ERR) for Sinaloa I was calculated at 11%, slightly below the appraisal DIR:0VWmesingle spaced overview e 14:10:00 24-AFR-8 - 3 - estimates of 12%. The ERR for Panuco was recalculated at 11.4% compared to 20% at appraisal. 7. The two projects in Morocco, Doukkala I and II, were smaller in scope than those of Mexico--an average of 15,000 ha each, and part of an ongoing water master plan for development of Morocco's second largest river basin. Previous irrigation development in the Doukkala region comprised the Sidi Bennour and Sidi Smail gravity systems, constructed in the mid and late sixties, but both Doukkala I and II (Zemamra and Tnine Gharbia dis- tricts) were designed for sprinkler irrigation because of topographical and soil conditions. The projects, therefore, provided for renovation and adaptation of the existing main service canal and provision of a downstream water control system; construction of pumping capacity and a tertiary net- work closely linked to the land consolidation system; equipment for opera- tion and maintenance as well as supporting services to farmers. Land con- solidation of almost 57,000 individual plots into about 6,600 farms in the two projects would facilitate provision of irrigation infrastructure and the supply of agricultural supporting services from the project authority. Both projects were located in an area where average rainfall was less than 300 mm. p.a., and winter wheat and livestock were the principal agricultural activities prior to the projects. Doukkala I was approved in 1976 to be implemented over a three-and-a-half year period; Doukkala II, approved in 1977, was expected to be implemented over a four-year period. Total proj- ect costs were estimated at US$94.4 million and US$85 million respectively. The projects were to be constructed and administered by an agency of the Ministry of Agriculture which was responsible for the provision of all supporting services to the farmers. 8. Both projects were successfully implemented, although Doukkala II encountered substantial delays, largely due to national budgetary difficul- ties. At completion of the two projects, the irrigable area (32,220 ha) was slightly larger than appraisal estimates. All lands had been success- fully consolidated in the project areas prior to construction of civil works. However, the number of holdings was about double and their average size (2.1 ha) significantly lower than anticipated due to subdivision of landholdings before consolidation. The smallest were grouped into 5-ha joint ho3dings to be farmed collectively by their owners. Project results were considered highly satisfactory, with agricultural production and yields higher than forecast at appraisal and cropping intensities in line with estimates. At project completion in 1981 (Doukkala I) and 1986 (Doukkala II), total project costs were slightly less than expected: US$85 million and US$77.4 million for Doukkala I and II respectively. Production forecasts to full development were revised upwards and the revised ERRs were considerably higher: 20.5% (vs 11.4%) for Doukkala I and 18.7% (vs 11.6%) for Doukkala II. II. AGRICULTURAL AND ECONOMIC IMPACT 9. At the impact evaluation stage, a few years after completion, a striking feature of all four projects is the lower agricultural and economic performance than estimated at the time of audit. Weaker commodity prices, an inadequate development approach in Panuco (Chapter III), lower than expected water efficiencies, insufficient attention to operation and DIR:OVOWN1faingle avaced overves eng 14:10:00 29-APR-eg - 4 - maintenance in the four projects (Chapter IV), a&:. damage to environment and natural resources in Sinaloa (Ch-apter VI) have been the main factors affecting performance. However, with the exception of Panuco, which has performed poorly, the degree of success in the transfer of technology in respect of both irrigation practice and agricultural productivity has con- tinued to be sustained in the three remaining projects, yet to a lesser degree than expected. 10. In Mexico, the two projects have evolved quite differently. In Sinaloa, agricultural development has progressed well. Doublecropping is usual practice and irrigation is used in all seasons. Cropping intensity and average yields have been higher than appraisal estimates but lower than at project completion due to growing salinity and waterlogging affecting about 17% of the project area. As a result, the project ERR is now reesti- mated at 9% (112 at completion). Despite time and cost overruns, the ERR of Sinaloa II is expected to be above 10% due to the considerable sunk costs of Sinaloa I. In Panuco, only about 30% of the project area is actu- ally irrigated leaving a large proportion of the crops fed by rain. The cropping intensity has declined since project completion and yields have been significantly lower than anticipated. The main reasons for poor proj- ect performance are the incomplete irrigation infrastructure, insufficient levelling and a development strategy not adapted to the majority of expected project beneficiaries. While private farmers have performed remarkably well, most ejidos went bankrupt and lost access to credit, resulting in reduced use of agricultural inputs and low yields. The proj- ect ERR is now reestimated at 72. 11. In Morocco, appraisal production targets have been fully met, but actual production is below the revised, upward forecasts made at comple- tion. Irrigation is fully utilized, but cropping intensities did not pro- gress as expected, partly due to the higher proportion of winter crops in the cropping pattern and the inadequate technical package for summer crops. Yields of sugarbeet, the main cash crop, have been higher than completion targets, but yields of cereals have been lower due to poor evolution of the technical package. Livestock and milk production have not progressed as expected because of changes in land tenure and the attitude of small farm- ers towards livestock. The ERRs of the two projects are now reestimated at 11.72 for Doukkala I and 10.5% for Doukkala II, about the same as at appraisal but lower than at completion. 12. Some conclusions can be derived from the experience of the four projects: (a) The tendency at project completion to over-estimate yields and cropping intensities at full development in irrigation Proiects when uptake of irrigation is successful and agricultural devel- opment is rapid, as was the case of the Sinaloa and Doukkala projects. In these three projects, unexpected internal and external factors have contributed to agricultural development remaining at a plateau or slowing. (b) The need to pay close attention to minimizing risks in the transfer of technology in climates which permit rainfed culti- vation. The most rapid transfer has come in the projects with low annual rainfalls Doukkala with about 330 mm per year and DIMt0VtieRaingle apaced overview eg 14:10:00 29-APR- - 5 - Sinaloa with about 550 mm per year. In Panuco, where the aver- age rainfall is about 900 m per annum, the introduction of Irrigation has been accompanied by an increase in rainfed cul- tivation and a steady decline in irrigated crops since the peak at project completion: overall cropping intensity reached its highest level at loan closing (106% in 1982) and has since declined to about 872 for the past four years. Irrigation in this project is seen as a high risk enterprise; there has been a steady and sustained increase in livestock production due to its higher profitability and lower risk. (c) Project experience highlights the advantages of a flexible and diverse cropping pattern. supported by an effective research program, for both winter and summer crops. In Sinaloa, the cropping pattern has varied each year depending on the availa- bility of water and market prices. Farmers cultivate up to ten different crops and seven types of vegetables, and thus are able to respond rapidly to changing price incentives. In con- trast, the cropping pattern in Morocco was closely linked to land consolidation and based on a four-year rotation of a limi- ted number of crops. This has resulted in reduced flexibility when increased area to a winter crop such as wheat displaced summer crops under existing technological conditions. No research was undertaken to fill this identified gap in the cropping pattern. (d) The relation between agricultural performance and farm size, which varies from one country to ar>.ther. In Sinaloa and Panuco, where all farms are relatively large (12 irrigated ha on average), farmers, well organised, using capital-intensive and advanced agricultural techniques have generally been more successful than the smaller farmers. In contrast, in Morocco where landholdings are small (2 ha on averapt and agricultural operations are mostly labor intensive, the a.allest farms have significantly higher cropping intensity and yields ti%n the large farms. III. THE DEVELOPMENT APPROACH 13. In both countries and the four projects, the development approach was based on two fundamental features: (i) a change in land tenure to make project benefits more equitable (Mexico) or the irrigation system more efficient (Morocco), and (ii) strong and autonomous project institutions providing a number of services to farmers in addition to the supply of water. This section analyzes the degree of success of the development approach. A. Land Tenure 14. In Mexico, land tenure has evolved significantly since project completion. In Sinaloa, there has been some progress towards agrarian reform in terms of transfer of ownership; 14% of the area formerly owned by private farmers has been transferred to ejidos. Concurrently there has ODR:0VMWaing1e spaced overview eng 14:10:00 29-APR-9 - 6 - been a blurring in the distinction between land owned and land cultivated. The area cultivated by private farmers has increased; many rent in land from ejidos. Many private farmers in Sinaloa and in Panuco have escaped land reform by dividing their holdings among family members, thereafter regrouping these individual plots into single companies of large, modern, highly-mechanized irrigated farms. In Panuco, hoqever, the success noted in agrarian reform has been arrested: there is now a greater area de facto to private farmers than expected. Average size cf private farms at project completion was 38 ha but at impact evaluation tne survey shows that this has increased to more than 200 ha, mostly due to land renting from ejidos. 15. The finding that agrarian reform objectives in Mexico have been only partially achieved at best is not a result of lack of government com- mitment. Such objectives have recei-7ed considerable attention through policy pronouncements and statutory legislation. The problem is rather that the underlying strategy was inappropriate. The average 10 ha of irri- gated land given to each ejidatario is large both by Mexican and by world standards; there is little tradition of communal cultivation on large blocks of 500+ ha in Mexico; newly settled peasants in Panuco, some with lit tle experience in agriculture and almost all with no experience in farm ,eament, received insufficient technical assistance; intensity of the extension agents was light; ejidos were expected to be established but similarly were provided with little or no training in finance/accounting/ marketing, whil, credit in cash for 100% of inputs was readily available and unsupervised. The results of this poorly adapted stratugy were sev- eral. First, explizit goals of economic growth have not been achieved in Panuco where uptake of irrigation is poor. Productivity has not increased and most ejidos are bankrupt, which has further exacerbated the vicious circle of low use of inputs and low yields. Implicit goals of equity have not been achieved in either Panuco or Sinaloa siwte large farmers have captured most of the project benefits (Chapter V). Second, the strategy has contributed little toward employment generation but rather, and despite the 10-ha size of holdings, many ejidatarios in Panuco, discouraged by the concept of collective block farming, sought off-farm employment. 16. In Morocco, the trend towards land subdivision and fragmentation, already noted at project completion, has continued; the average size of holding is now 1.9 ha. The success of the projects in transferring tech- nology has resulted in all land rights being claimed. The survey illu- strates that while, prior :o the project, there was rural out-migration, the project has resulted in in-migration from nearby cities such as Casa- blanca and even from overseas. The financial benefits of the project (irrigated land in 1988 is worth US$5,500 vs US$950/ha rainfed) together with the strong and continued adherence to Islamic inheritance laws have proved more compelling than public efforts to consolidate land and arrest lard subdivision and multiple ownership by civil means. I.. is worth not- ing, however, that land consolidation in Doukkala has reached its technical objective of integrating the layout of the irrigation system uith the land tenure system. All farms, consisting of rectangular plots, are serviced both by the road and the irrigation network, a remarkable achieve4ent in smallholder irrigation development where average h"ldings are less than 2 ha. 17. The findings related to the evolution of land teaure iin the four projects have several implications: DOR:0VERsiaingl spaced overview eng 14:10:00 29-APR-89 - 7 - (a) It was expected at project completion that once agrarian reform was introduced in Mexico and land consolidation in Morocco, the land tenure system would remain essentially unchanged. At the impact evaluation stage, however, it is clear that there is a tendency to revert back to the situation which prevailed before agrarian reform or land consolidation. Project experience showed that in irrigation projects, which result in increased value and higher productivity of land, the land tenure system is continually evolving. The implication is the need for a flexible development strategy, more particularly if the irriga- tion system and infrastructure are designed on the premise of a static land tenure situation (Chapter IV). (b) Project experience has shown the danxer of introducing collec- tive farming where there is little tradition of communal culti- vation. Both the large collective ejidos in Panuco and the grouping of small holdings to be farmed collectively by co- owners in Doukkala have failed. Most of the ejidos in Panuco have moved from the original collective farming system to individual or small group farming. In Doukkala, contrary to expectations, the smallest farmers continue working their land individually. (c) In countries with high unemployment, there is a need to develop a smallholder policy based upon high productivity and high employment per irrigated hectare. The smallholder approach in Morocco has demonstrated the success of this approach, result- ing in a large number of beneficiaries, fully employed on their farms, with a relatively low investment cost per family (US$10,500). In contrast, the smallholder approach in Mexico was questionable. The average area allocated to ejidatarios was large compared to that prevailing in most developing coun- tries in irrigated conditions. Agricultural activities were entirely mechanized, resulting in low employment opportunities, and total investment cost per family (equivalent to US$58,000) was extremely high. B. Institutional Effectiveness 18. Both countries have a long tradition with construction and manage- ment of large irrigation schemes. Distritos de Riego in Mexico and Offices d'Irrigation in Morocco provide a number of services to farmers and are known for their generally good performance and efficiency. 19. Both public irrigation services, therefore, have designed a devel- opment strategy to take account of the perceived needs of their respective farmers. Thus in Mexico, where irrigation is traditional, farms are rela- tively large and farmers generally progressive, the i.nstitutional system design is relatively light with a low ratio of extension agents. Training activities, provision of inputs and marketing are not considered essential. In consequence, the settlement of ejidatarios has been in the form of large ejido blocks patterned after the large farms. By contrast, in Morocco, the system is designed for the small and inexperienced farmers who require a larger range of services, including provision of credit and :tnputs, assis- tance with marketing, and high coverage of extension and social services. DIR:0V8&wM/single opacod overview one 14:10:00 29-APR-89 - 8 - 20. In Mexico, the project institutional arrangements and development approach have proven fully satisfactory for the Sinaloa project given that the land reform objectives were not achieved and only a few ejidos were newly established. The Gasave District has remained a well-managed organ- ization, staffed with competent and motivated people. Performance, how- ever, has been hampered by budget cuts and low cost recovery rates result- ing in sharp reduction of personnel and equipment. The government policy is now to disengage progressively, to restrict the District's activities to a small number of services, mostly operation and maintenance (O&M) of the irrigation system, and to promote farmer participation in all development activities. Farmers have reacted positively by taking over services such as agricultural research and extension, storage, marketing, and processing of products. More participation of farmers in 0&M is also contemplated. 21. The same development approach, however, has proved inappropriate to Panuco, where project beneficiaries were mostly inexperienced settlers with no experience in irrigation. Although the three Irrigation Districts in Panuco are also well managed, the example of Panuco is a counterpoint to the experience of Sinaloa. Government support in Panuco will be necessary for a number of years if project objectives are to be fully achieved. 22. In Morocco, a central feature explaining the success of the Doukkala projects has been the formulation and efficient implementation of an appropriate development strategy. The transfer of technology has been very successful in a region where farmers traditionally cultivated winter wheat, with livestock the main cash activity. The change to year-round irrigated cultivation of sugarbeet, wheat, and, to a lesser extent, summer and winter vegetables, all with high yields, is a result of an integrated package for sugarbeet and wheat whereby farmers receive (i) the necessary inputs on credit--two thirds of farmers had not had any contact with the credit bank prior to the project; (ii) strong assistance from the extension servicest (iii) advice from ditchriders when conflicts arise among water users; and (iv) a guaranteed price for their sugarbeet which is collected and delivered to the sugar mill. The project authority's (ORMVAD) involve- ment was also welcomed by the sugar factories since it ensured smooth pro- cessing of sugar from timely and planned planting and harvesting to deliv- ery to the mill gates. Despite the relatively high cost of these services, ORMVAD has achieved a highly satisfactory cost recovery rate and has become financially autonomous. While this development strategy was undoubtedly appropriate for Doukkala, it has been criticized as having failed to evolve. The strict control exerted on farmers has resulted in low levels of participation and self-help. Under the follow-on Large-Scale Irrigation Improvement Project, ORMVAD is now withdrawing from a number of its past activities with the risk of dismantling its integrated development approach, which was the main reason for project success. To be successful, this disengagement process will have to be gradual, carefully monitored and based on a more open dialogue with Doukkals farmers. 23. The lesson is that the development aproach has to be adapted to the type of beneficiary and not applied uniformly to all regions and all projects in a particular country. In this respect, the Bank's insistence on quick structural adjustments, government disengagement and privatization of public enterprises is certainly justified on many occasions, but should be applied selectively, with caution and in accordance with need. DM0V1tm%/.ingle spaced overview eng 14:10-00 29-APR-89 - 9 - IV. PERFORMANCE OF THE IRRIGATION AND DRAINAGE SYSTEM 24. This section analyzes the extent to which irrigation system design and water manag-ment have met appraisal expectations as measured by the current status of infrastructure, degree of water efficiency, operation and maintenance, and cost recovery. Main project features are detailed in Annex 1. The section concludes with an analysis of the impact of differing types of system design on subsequent project performance and outcome. A. Status of Infrastructure 25. In Mexico, the main feature of the two projects is the incomplete- ness of civil works. The irrigation infrastructure has been completed in Sinaloa I, but of the drainage system only the main drains have been con- structed, and groundwater has not been used as planned. The result is that the water table is rising rapidly with soil salinization and waterlogging on about 17% of the project area. In Panuco, the tertiary irrigation net- work has been left incomplete, and land leveling has been insufficient, resulting in low uptake of irrigation by ejidatarios. Iu both projects, construction has been of medium standard, and maintenance has been poor as evidenced, mostly in Panuco, by breaks in concrete canal linings, encroach- ment of vegetation and the non-operational status of the sliding gates. Six years after completion of the main works, the need for rehabilitation is relatively high. 26. In Morocco, the main project works are generally in good condition except at the field level where hydrants and mobile equipment have been damaged by farmers. The drainage system has been completed and proven efficient. As in Mexico, insufficient maintenance and delays in repairs have adversely affected project performance. Of note is the good status of the Sidi Bennour/Sidi Smail gravity irrigation systems, constructed in the 1960s and still fully operational. B. Water Efficiency 27. In both countries, water efficiency--which measures the extent to which water diverted from headworks actually reaches the crops to be irri- gated--has been significantly lower than expected: 50-77% of appraisal estimates, as shown by the following table which provides actual versus expected achievements of the four projects, as well as of the two non-Bank- supported projects, Sidi Bennour in Morocco and Yaqui in Mexico. Table 1: WATER USE EFFICIENCY Convevance and Dlstribution Efficiency Field Efficincy Overall Water Use Efficiincy Appraisal Actual Actual as Appraisal Actual Actual as I Appraisal Actual Actual as S ffSL"J& _ ofAgaraisal E&imat ofAppraisal Estimae of Aporaisal Sinaloa 74-80 671 87 705 555 791 SPm 371 711 P*.uco 7a S4 721 7s 48s e 52s 26 s01 oukka Ia 861 73-7811 e 751 871 753 4 491 771 Sidi Benour ass 7a as as SeS 971 aos 42s e Yaqui 61% i1l 1006 70-751 55 76 43-481 881 o5 DIROVEVW&ingle spaced overview eng 14:10:00 29-APR49 - 10 - Although lower than expected, the overall water efficiency of the four projects is relatively satisfactory when compared to many irrigation systems in developing countries. So far, the lower than expected water use efficiency has not resulted in water shortage, but adequate water supply will be uncertain if and when the optimal cropping intensity is reached. 28. In Mexico, the design of the irrigation network in both projects was a simple, conventional gravity system, with upstream water control and low density of farm turnouts (1:58 ha), reflecting the large farm sizes in the project area. In Sinaloa, lower water efficiency than expected is mostly due to poor maintenance (para. 32) and mediocre construction stan- dards, but also to over-optimistic appraisal assumptions. These short- comings, however, were compensated by excellent water management based on arranged demand which required a high degree of training and dedication on the part of project staff. Weekly water allocations and adaptation of the cropping pattern to water storage in the reservoir at the beginning of the cropping season resulted in flexible, equitable, reliable and timely water distribution. In Panuco, water efficiency has been adversely affected by the de facto distribution of land among ejidatarios and the fact that only a fraction of beneficiaries, scattered all over the project area, made full use of the water. The low density of farm turnouts, originally designed for large farms, proved poorly adapted to such a "patchwork" type of water distribution and low cropping intensity, demonstrating the low flexibility of the irrigation system. In addition, the incompleteness of the civil works has affected project ejidatarios more than private farmers, who have generally been able to correct these deficiencies with their own resources. As a result of these shortcomings, equity, reliability and timeliness of water distribution have been poor in Panuco. 29. In Morocco, both Doukkala projects consisted of sprinkler irriga- tion, not only more suited to the soils and topography of the project area, but also expected to be more efficient than gravity, with 30% more irri- gated area for the same amount of water diverted. The irrigation system was technologically advanced, with automatic upstream and downstream con- trol, an underground distribution network, and water allocation based upon demand--whereby farmers have access to water whenever they wish. The high density of turnouts (1:12.5 ha) reflected the small farm size in the proj- ect area. Despite its sophisticated design, the overall water efficiency of the Doukkala projects is lower than expected due to the inadequate main- tenance of field equipment and the much smaller than expected farm size. These factors, combined with farmer reluctance to cooperate at the block level, resulted in substantial over-watering of fields and water losses. The irrigation system, however, proved remarkably flexible and equitable as the reduction in average farm size has been accommodated without any reduced access to irrigation. The automatic water control system resulted in reliable and timely water distribution. However, because of the increasing fragmentation of land, the sprinkler system as designed no longer accommodates the current land tenure situation. C. Operation and Maintenance (0&M) and Water Charges 30. While the basic operational system of the four projects is gener- ally adequate, reflecting a long tradition in water management in the two countries, it is an important finding of this impact evaluation that (i) maintenance is poor in the four projects, and (ii) actual O&M expenditures DIR:0VSV/iogle spaced overview eno 14:10:00 29-APR-89 - 11 - are much lower than estimated at appraisal. In the four appraisal reports, O&M cost estimates were roughly expressed per delivered volume and per irrigated hectare with little or no supporting data; detailed O&M programs were at most poorly specified. Hence any comparison between planned and actual O&M programs is difficult if not impossible. In the four projects' Loan Agreements, provisions for cost recovery were in line with Bank policy requiring recovery of at least O&M costs and part of the investment costs. Therefore, water charges were to be adjusted or new water charge structures introduced in the project areas. At full development, cost recovery is actually treated very differently in the two countries. Operation and Maintenance (OW) 31. The following table shows appraisal estimates and actual O&M expenditures and water charges of the Sinaloa, Panuco and Doukkala I proj- ects.l/ For the purpose of comparison, actual O&M expenditures are also given for the Yaqui and Sidi Bennour projects. Table 2: APPRAISAL ESTIMATE AND ACTUAL O& EXPENDITURES AND WATER CHARGES (1987 US$ Equivalent) Sinaloa Panucc Yaqui Doukka a I Sidi Bennour Appr. Actual (M) Appr. Actual (S) Appr. Actual () Appr. Actual (S) Appr. Actual (S) Esoendituree erD vred Yolu"e (1.000 ON Operating Expenditures 6.4 1.9 (35) 6.4 8.2 (152) n.s. 2.1 20.8 5.3 (26) n.s. 10.0 -- Energy -- --- = - - aA AA m :-- - -- Total 6.4 1.9 5.4 8.2 2.1* 84.6 15.1 (44) 10.0 Overall Water Use Efficiency (S) 52 37 (71) S2 26 (50) 45 38 (84) 84 49 (77) 50 42 (84) Exmenditures and Water Charea ner Irrieated Ha Operating Expenditures 49.1 17.3 (86) S5.1 87.7 (68) n.e. 16.9 - 181 41 (23) n.s. 81 - Energy -- - - - -- -= ----U I1 ---- -- Total 49.1 17.3 S3.1 87.7 15.9* 302 116 (38) 51 Water Charges 54.0 6.4 (12) 51.S 3.8 (8) 12.5 403 232 (68) 65 Water Charges as I of OM Expenditures 871 10% 79% 2001 127 Easenditures per Irxilatin. Farmer 42S 155 (36) 154 409 (265) n.e. 180 1450 106 (7) n.e. 158 Are Served by Fare Turnjut (ha) 66.5 58.0 48.6 12.5 4.3 a In Yaqui, 0bf of the water distribution system is operated and maintained by water user easociations. 32. Per-hectare expenditure in the Bank-supported projects has been 302 to 65% lower in c.astant terms than expected at appraisal. Expenditure per delivered volume is also much lower than expected with the exception of Panuco, which illustrates that low utilization of irrigation water results in high O&M costs on a per-water volume basis, due to the relative impor- tance of fixed costs, mostly personnel, in the O&M budget. In the two countries, O&M departments are severely under-staffed and lack adequate equipment. In the view of this impact evaluation, staffing needs to be strengthened by an increase of 36% in Doukkala and 53% in Sinaloa to oper- ate and maintain adequately the irrigation systems. Inadequate staffing is ]/ The newly developed Doukkala II is not considered in this analysis. DWs:OVtWeingle &paced overview eng 14:10:00 29-APR-89 - 12 - partly due to budgetary constraints and government priority to new invest- ment rather than recurrent expenditure. 33. In both countries, it was expected that responsibility for 0&N would be shared with farmers through the establishment of water user asso- ciations. Such organizations have not developed in the Bank-supported projects, in contrast with Yaqui in Mexico where such associations are successfully maintaining the secondary and tertiary distribution system. It is worth noting that despite the existence of such associations 0&M expenditures in Yaqui (Table 2) remain higher then in Sinaloa, both on a per-ha and per-delivered volume basis, illustrating the degree of low bud- getary allocations for O&M in Sinaloa. Water Charaes 34. The relation between water charges and adequate 0&fM has been a subject of discussion for many years within the Bank as well as between the Bank and its Borrowers. The Bank policy assumes that if the full cost of 0&M is recovered, then governments will allocate a major portion of these revenues for the proper O&M of the irrigation systems. Conversely, effi- cient 01M of irrigation systems is a prerequisite to cost recovery.!/ 35. In Sinaloa, poor OM and non-completion of the drainage system haves resulted in strong resistance from farmers against any substantial increase in water charges. In constant terms, water charges have decreased over time; in 1987 they covered 37% of 01M expenditures, amounted to 12% of appraisal projections and were equivalent to only 1% of gross production value. The project is a clear demonstration of the vicious circle created by--and the interaction between--low water charges and poor 0&M. The only remedy, which appears now better understood both by Government and the farmers, would be for the farmer to take a much more active role to replace Government, and the creation, as in Yaqui, of water user associations fully responsible for O&M of the secondary and tertiary irrigation systems. 36. In contrast, there is no relation in Morocco between adequate water charges and OM. Water charges have been increased continuously in constant terms, following strict government policy, fully supported by the Bank. Water charges are collected by sugar mills which deduct the requi- site amount from farmers' revenues on sugarbeet. The total amount recov- ered in 1986--the equivalent of US$232 per hectare in Doukkala I or 10% of the gross production value--was somewhat below appraisal expectations, but was about 100% higher than actual OM expenditures, which were kept at a low level. Doukkala is the example of a project where high and increasing water charges have resulted in full financial autonomy of ORMVAD, but have not resulted in adequate O&M, a drawback which, it is hoped, is being addressed by the Bank-supported follow-on project (Large-Scale Irrigation Project). 2/ See World Bank Lending Conditionality--A Review of Cost Recovery in Irrigation (OED Report No. 6283). DM:OVMM/singIe spaced overview eng 14:14:00 29-APR-89 - 13 - D. Irrigation System Design and Project Performance 37. Within Morocco, and comparing Doukkala with the old Sidi Bennour and Sidi Smail projects (both gravity systems with a fully lIned distribu- tion network and a combination of upstream and downstream water control), the expected benefits of sprinklers have not fully materialized. The con- veyance and distribution system in Sidi Bennour/Sidi Smail is as efficient as Doukkala even after 16-22 years of operation; the field efficiency is in the range of that experienced in both Doukkala projects. 38. When comparing irrigation system design across countries, however, the field efficiency, conveyance and distribution efficiency and overall water use efficiency are all higher in both Doukkala and the old Sidi Bennour/Sidi Smail gravity systems than the never gravity systems of Panuco and Sinaloa in Mexico. When Doukkala and Sidi Bennour are compared with Sinaloa, the better performer in Mexico, the overall efficiency in Morocco is about 25% higher (32% for Doukkala and 14% for Sidi Bennour) than that of Sinaloa with only an 11% additional cost for Doukkala and a lower cost for Sidi Bennour, illustrating the superiority of the advanced technolog- ical design and water control system used in Morocco. 39. When comparing qualitative performance of the operational proce- dures in terms of relative advantages as between arranged demand with up- stream control and an on-demand system with downstream control, project experience shows that while in Doukkala a modern, technologically advanced system intrinsically fulfilled these requirements, good water management in Sinaloa was able to compensate for less sophisticated system design and to assure satisfactory qualitative performance. This is all the more note- worthy given the light staffing intensity for operating the system in Sinaloa, each ditchrider controlling an area of 2,751 ha, large even in comparison with other projects on the Pacific coast (Yaqui 2,115 ha) and in comparison with Panuco or Doukkala. 40. It is difficult to compare the relative merits of different irri- gation system designs with costs of operation and maintenance because actual expenditures for O&M and cost recovery have not reflected the real needs of the four projects. O&M expenditures per delivered volume and per hectare have proved greater in the Moroccan projects where the density of canals and turnouts is much higher than in Mexico. However, expenditures per farmer have not been significantly different among the projects, with again the exception of Panuco, reflecting the underutilization and "patch- work" distribution of water. In Morocco, downstream control and the on- demand system of Doukkala permit a 50% saving in operational expenditure, mostly staffing, when compared with the Sidi Bennour gravity system, but this saving is largely offset by the cost of energy. 41. In examining the relative merits of irrigation system design and performance of aaricultural production it is only in respect of the crop- ping pattern and yields that there is any inherent advantage in terms of one system over another. Thus the on-demand system enables a much wider choice of cropping pattern although somewhat ironically in Doukkala, while the cropping pattern changed this was from higher to low.r cropping inten- sity. In examining the relative impact in terms of agricultural produc- tivity, output per ha is sigtificantly higher in Doukkala (US$1,425/ha) compared with Sinaloa (US$987/ha), in economic terms, based upon a 2-ha Dt:0VEWeeingle spaced overview eng 14:14:00 29-APR-89 - 14 - farm in Doukkala and a 9.5-ha holding in Sinaloa. This is mostly due to higher-value crops in the rotation in Doukkala and not to average yields and cropping intensities, which are slightly better in Sinaloa than in Doukkala. This relatively greater emphasis to high-value crops in Doukkala is in turn a reflection of the small farm size, and an irrigation system well adapted to smallholders. When the respective benefits of sprinkler and gravity are compared within Morocco, net value of production per irri- gated hectare in sprinkler is about 14% higher than in gravity, but when O&M costs are taken into account, the not production value per irrigated hectare in Doukkala is only 6% higher than in Sidi Bennour with 40% more invested. 42. Measurement of profitability in terms of investment costs as between the different system designs reveals that Doukkala, which is the second to Sinalos in terms of costs/ha, has double the gross production value or 70% higher when adjusted in economic termst Table 3: INVESTMENT COSTS AND AGRICULTURAL PRODUCTION VALUE (US$Iha) Doukkala Sidi Bennour Sinaloa Panuco Investment Costs 4,988 3,040 5,780 4,850 Gross Production Value 1,693 1,295 832 610 Economic Production Value 1,425 - 987 - At the same time, it is difficult to see the benefits of almost US$500.0 million invested in Panuco when only 28% of the area on average has been irrigated during the past five years and production value per hectare is only US$610. 43. Some conclusions can be deciived from the performance of the irri- gation systems in the two countries: (a) Project experience has demonstrated that projects left unfin- ished do not sustain their expected benefits. The lesson for the Bank and the Borrower is that if financing is not secure it is preferable to phase a project in stages, each stage to be fully completed, rather than to disengage from incomplete investments. (b) The old gravity irrigation systems in Morocco have shown the importance of good construction standards for maximizing both project benefits and the useful life of civil works. (c) The gap between appraisal estimates and actual water efficien- cies was larger in the four Bank-supported projects than in the two non-Bank-supported projects, suggesting a tendency in the Bank to make over-optimistic projections as regards water use efficiency. (d) Project experience shows that 0&M plans and costs should be more carefully developid at appraisal than they have been in the past. Project experience also demonstrates that these may vary greatly from one project to another, depending on farm DIR:OVOWW/inefe apaced overview eng 14114100 29-APR-89 - 15 - tenure, density of canals and turnouts and the water management system. Calculating O&M costs as a rough percentage of invest- ment cost, a frequent practice of project designers, is cer- tainly not the best vaf of determining true O&M costs. Person- nel and equipment stould be carefully estimated at project appraisal. The Bank should insist that the Borrower provide a plan to assure that funding for O&M be provided and such requirement should be covenanted in the Loan Agreement. This important issue has been recognized by the Bank, which, in 1986, prepared new guidelines for estimating 0&M costs of irri- gation and drainage. (e) While the Bank policy is to promote a cost recovery system which covers full O&M and part of the investment costs, there is a need to ensure that recovered funds are used primarily for 0&M, and not for other purposes, and maintenance is kept ade- quate at all times. V. SOCIAL IMPACT 44. An important objeccive of the four projects was increasing incomes and improving the living standards of the farming population. In both countries, farmers' incomes were expected to increase by about 4 times and living conditions were to be improved through construction of village infrastructure. Contrary to Morocco where land consolidation was not intendad to bring changes in farm size, the two projects in Mexico had an explicit goal of equity through expropriation of large farms and land distribution to landless people. 45. In Mexico, a common feature of the two projects is that the expec- ted narrowing of income distribution between private farmers and ejidata- rios has not materialized. In Panuco, despite the agrarian reform, large private farmers are, in fact, the main beneficiaries of the project because of significantly increased area cultivated and rapid uptake of technology. Although incomes were expected to reach parity at full development, large income disparities are now observed among the two categories of beneficia- ries; average yearly incomes of private farmers (US$40,000) are about 30 times those of ejidatarios (US$1,320). Average incomes of ejidatarios have remained lower than the poverty level (US$3,000 in 1987 terms). Large variations also exist among ejidatarios, from a net loss of US$2,600 to a net income of US$13,750 in 1987 illustrating how individuals who settled at the same time and in similar conditions have evolved differently over 10 to 15 years. These findings show that, overall, 20% of project beneficiaries in Panuco have captured 85% of project benefits. Private farmers, with their large holdings and 1aigh-value crops have created a substantial demand for seasonal labor. For ejidatarios, farm incomes generate only about 55% of total family incomes and hired labor is the second most important source of earnings. Despite these income disparities among groups of beneficia- ries, the majority of ejidatarios consider themselves better off now than before the project because of ownership of land rights and improved living conditions in villages equipped with electricity, potable water, education and health facilities. DR:tOVtWW/&inte epaced overview ag 14:14:00 29-APR-89 - 16 - 46. In Sinaloa, average incomes of both ejidatarios and private farm- ers have increased by about three times, to about US$5,480 for ejidatarios and US$15,300 for private farmers. However, the group farming system found among private farmers together with their renting of land from ejidos means that family incomes of private farmers are probably considerably higher than they appear. When compared with average incomes of rural families in Mexico, incomes of Sinaloa farmers are fairly high, reflecting the excel- lent potential of the project area and high productivity of farmers. 47. In Morocco-Doukkala, incomes of small farmers have increased in line with appraisal expectations, reaching the equivalent of US$2,200 for 2-ha farmers and more than US$6,000 for 5-ha farmers. Incomes of 25-ha farmers (US$15,300) have only reached 72% of appraisal targets. Although not planned at appraisal, improved equity in income distribution has been attained in Doukkala. This has been achieved by a greater number of small farmers than expected (72% of farmers have a farm size under 2 ha compared to 43% at appraisal), the higher productivity of small farmers, and a sys- tem of water charges which penalizes the large and less productive farmers. It should be noted, however, that even with the higher yields, higher crop- ping intensity and higher-value crops which characterize small farm culti- vation, farmers with 2 ha or less have incomes below the poverty line. The projects have generated year-round employment for the whole family. Seventy-three percent of male family members over the age of 15 work full- time on the farm and the incidence of work has increased among women and their daughters. The projects have provided sufficient incentive to slow the process of urban migration, many farmers returning from urban areas to take up land rights. 48. There has been little progress in terms of literacy and education in Doukkala because of the increased incidence of work and the need to utilize family labor. The environment remains very traditional: families remain large (seven children on average), girls marry between the ages of 12-14 and women rarely venture beyond the homestead. Almost all families have been reluctant to move to new villages equipped with electricity and potable water because there is no place for their animals. Living condi- tions in the old villages are generally poor, with all the problems ane diseases related to a lack of potable water. Nevertheless, the majority ot Doukkala farmers expressed satisfaction with the reduction of climatic risks due to irrigation, improved incomes, better food supply and employ- ment generation at the village level. 49. A few lessons can be learnt from project experience: (a) as shown in Mexico, the expected narrowing of income distribu- tion among beneficiaries is not attainable or sustainable in the absence of a smallholder development approach. The coun- terpoint is Doukkala, where equity of financial benefits, though not specifically intended, were largely achieved due to an efficient smallholder strategy; (b) the four projects contributed to stabilizing the rural popula- tion. limiting migration to cities and alleviating underemploy- ment, a common feature of many irrigation projects; DZR:0V6t4i/*ingle spaced overview eng 14:14:00 29-APR-89 - 17 - (c) ownership of a land title is considered an important benefit by smallholders in irrigation projects, more so than direct finan- cial benefits, and yet can result in increased incomes when both the value and productivity of land have risen considerably; (d) the Doukkala experience confirms OED findings in other countries 3/ that successful agricultural development entails increasing the workload more particularly for women and chil- dren, and often at the expense of school attendance in tradi- tional societies. In Morocco, the need for improved access to education is partict ly acute given the relatively small degree of social changt in the project area despite its eco- nomic and financial success; and (e) although improvement of social infrastructure was planned in both countries, the lack of success of the new villages by Doukkala farmers, contrasting with the satisfaction of Panuco settlers, illustrates the need for planners to design new vil- lages retaining some elements of the traditional style, a fea- ture already noted in a number of Bank-supported projects.4) VI. IMPACT ON ENVIRONMENT 50. At appraisal of the two projects in Mexico, it was estimated that projece. investments would have a beneficial impact on the environment by reducing risks of flood and propagating fish and wildlife in the reser- voirs. At the Impact evaluation stagt, these objectives have been achieved in Panuco through the construction of dams in project rivers. Despite the incomplete drainage system, no salinity or waterlogging have been encoun- tered so far as water use has remained far below expectations. In con- trast, Sinaloa is a typical case of heavy salinisation and waterlogging due to poor drainage, a frequent phenomenon in Mexico. Salinity and water- logging increased from about 800 ha in 1973 to about 11,000 ha in 1987 on the left bank of the Sinaloa River, and started developing on the right bank. It is estimated that about 850 families have incurred losses due to progressive abandonment of land and/or low production. A number of houses and buildings have been damaged by salinity. The situation is not irre- versible, as demonstrated by the Yaqui project which was, for the same reason as Sinaloa, affected by salinization in the 1960s and has since been rehabilitated. 51. In Morocco, no adverse impact on environment has been noted. The drainage system, whereby excess water accumulates in natural depressions, has proven successful. No elevation of the water table has occurred. The decanting system of the sugar mills has been adequately designed and opera- I/ See Impact Evaluation of Cotton Development Programs in Burkina Faso, COte d'Ivoire and Togo (OED Report No. 6878). A/ See Impact Evaluation of the Jengka Triangle Projects in Malaysia (OED Report No. 5988). O2R-0VaW#W/ng1e spaced overview eng 14:14:00 29-APR-89 - 18 - ted. A bilharsia control program has been successfully implemented. The Doukkala area, formerly arid sand drought prone, is now a region where cli- matic risks have been considerably reduced, living standards have improved and food supply is assured without adverse effects on environment. In the old Sidi Bennour/Sidi Smail projects, after 20 years of irrigation no nega- tive impact on the environment has occurred so far, but could result from the poor status of the drainage system if the planned rehabilitation is not undertaken soon. 52. Project experience in the two countries has shown the importance of fully completing the construction of the drainage system before irriga- tion starts and the need to maintain adequately the drainage network so as to keep it operating at all times. DM:ottWife gle seV ed overvie eng 14:14ioo 29-APR-89 Mflf2[0 вnд MOROCGO р�ки �лтиг�s oF п� svuAlлA. rваиг. Рмвко. ооиогил. sтк я►иS1.l1Ni�L�я vмгксвпаи аю.lвсгs 8iм1а I snd I2 Yaqui Рмиео РгоjвеЪ ��аиlгkвls I snd II Sidi Sввi1 sud 8idi 8мпоаг Qгavt�y �,ггtааltan бпv:вr i.rir.:м {jкjritv iм3тl7т $nrin4lвг fwг3�ian CrвrlLr lмiавЕ�� �� COY11bry 1Mt1C0 Мв11iа MвsicO lbгоесо 1lогоtсо QwyrмAie СоогдlмЪав 109° МО. 2В°И 109° УО. 27°1i 98° УО, 2'1°!t I2° 110, еб°Ч 12° УО. 86�11 �� C1sиltiwbim 8u►-6ropial вввi-вгiд Sub'6ropipl вееi'вгiд Tгopicsl еиi-sгlд ТмрвиЪв, виiчгiд 7вврвnЪв, вpS-ssгd Аrегвбв Амиrвl Rвintsll (в� В60 Z70 000 80S 805 Аrвnув Аыии1 Тирвntяи (°С) ю 48 '1д 38 1В Аиnбв Апnивt Рм Qrвpora0ion (вв) 1l.7J0 1,а68 1.бю 1,186 1.1ю rra.г в„лл г,. 8еига $iмlм Rivвг Ypui Rivвг Owyslвja, Тмрвои snd СА{nум Rivвгв Qw вг (�iв Riмг бив м АЬiв Rivвr Тррв Апмtвt Rp. 1lвbгв IпЪвгааwвl Rp. lfabи Амии1 Rв6ulsbd snd UпгвуиlвЪвд МвЧn АриlвЪвд аbп Rврмlвiвд nbn Аrвгрв Алмst Avsi 1вЬi 1 isy (!4� 1,880 2,40Т 6.680 660 be0 1f�Ъвг Ои.12ы. Riwnidв Clвpitinbiм Cl-$1 (в/) C2-81fC8-81 CS-S1 С8-81 � МwЬвг 1 8 ? 1 1 Та0в1 8iогвбв Сврвеilсу [11в� 1,610 6,180 PW 2,8ю 7.82D 1 Ригреов IFp/fLC/9Е W Il1R/ВЕ ПdtlflC IiBt/9Е W IPRIHE О/ t-� S� � ТоЪв1 LмуlЬ (Iм) 11Z 690 89В 1�8 146 1 т1,р. са,сгве. 11мд s Сакиы 1:мд в сапсиы 1:п.д а ип.д авпмl �/ме wм.� бrbh ам1 бгЪh евмl 6гИ Смв1 snd мгЪА дмt мд мгвЬ cмst lbвiвив Оiвдагбв Cspaeiby (в8/в) ю 1г� бо ю м F1ав аnЪго) lbОЪгои ' lbnuft UрsЪгевв ' Ммив1 t7увЪгмw ' l4гwвl 1w6oмbie ирвtпви АиtеввЪiе .рвЪгqв Q�:Ьыbiвп ИвLюгk sяд дамвWмв апд давrцfивв 8игlвев Irriяв4ian Спкгвlи linвd (88� Сеnеиtв 11ned (16>� СопеnЧ liмд (1� Vпдвгбгпипд ввЬиЪав 4iмд а ргвсввt бreh амtв (87'$} 6гЧ емьls (Т� 6гiА Свмtв (8bi) аипЪ рьра ариt• 2ггiвв�iм Ави 8игlаев IпiбвЪlоп (М►) 1С6,000 �20,000 141,100 82,I?9 Ib,648 $�f Attwrfвl м71в. A1luviвl мi1в, УвНiмtв, A1luviвl еоilв, Вгеим аlаиаw 8им alaгww � Ьгам мi 1в Ьгом мi 1• 1tыМкiпи (Iiвibвd dopbAj wгЪiмl• мд двву ввrЪiмl• � 1Ып Сив� УЬмЪ, 8оргЬиn. lАивЪ, CoOban. $оуЬвм Мв1tв. 9егбЬив. 8rувгЬввЪ, tA�иt 8ибвгЬввЪ, УЬвsЪ � $огбЬw. УрвЪвЫи $sftlаввг, Ри/wив УрвЪвЫи, Wибв УвбвlлЫи. Fогвбв N r �1tt tand амвиhiу Ргlvsев (4!�) E,jido (7l3� Ргiиb (� Ejido (68>)) Ргivвbв (87� Fjido (86� Privвb ('90ij to11вe6iv< (10l� РгiмЪв (� Collвcbiw (16'� Аwгвб. F:и sгвв (Ь.) 18.Т б.е п1.в о.о вел v.1 п s ббМ Sвwier РеЫ 1а Adвiпissrвbim РиЫ 1а Adrinisbиbiaп PuЫia Adьlnia0гвbi0n Pubtia Adafпisbrв4im PuЫie Аlвiпiв.4гвЪiт Vв�jм llвлп� Qиив /b УМ Na ND ND om:ov�vlel�r.ь1. о1д .пп.:» 1:зо-АРя-ю - �Q - Anneп 1 Page 2 . � � � � �s � � �$ � г а � �ь � � ь • ! � " �`�$ $g _ � � � и1 о � � � i3 � � , .. а � � ii � � .i S"а � += 8 S � 8 � г1' ьi д' ю 8 � � � � _� к � !- � ц � � � ы � � $ � й О о 3 � � . . � w е1 ië :. ё'i � :в � � � � 8 � � , tf � � tk � � � � 3 ° р � �_�� �� � � 4a о •Е v /� g� а в[ � � �.'7 v w � У � О � д в� t О! е9 � ti О �'�+ Ф О i � � О!t � 6 1� : � � � � � 9 S g 3 д �(O` �в Р а 4�j 4jj ь Ya ♦ уΡ�j ♦ }� � О�� �в О w 3 v' � А . � • � v v.�i v dF� �.�в У tf � О� б 1� О t� i � � �У�� � � - а н ` р 4 �g ь ь ь ь а ��w����� � �� ....7В � ���д��Яа $ л � 8дд� �S � д� . � • � � s� � �� к О � • • А рg V • i � � Е � i ;� о ё 8 s � � а � � v � �� � ~ � .� У а LL � Z А Е �b` � � {� � S' " " s ° � �i о \ • ¢ ` ь • < � � � � U � � • � й � � в S � �; � �_ а� � " s $ „$ � � � s g � � ; в� i�� �$� � ��g����������_ ��� ��z � _ � � : Р � ъ � g � �j tl ; � S �j � g � � ♦ � . � � �� ������� ' ������..о�.,��$�•� �� ������ � � �Q ' о ·21一觔議e×1 Pa妒3 !!,,,!,,,,,!!,，無一”: ！、！, 巒i審：，彥簷彥畫I簷！ 。！‘〕’&&&&&&&&&.&&&&”一‘&: 二望，唱，要． 轎禹芝召露。華：斗 &d名壇果渥萬。。蒲邊 l&&&&&&&&&&&&&&&&&&&&&& 啊。1 開廈1屋 U蔆1登奢登登召、rr d勰，偶屆多潺，J撻鄴 I，排，！。，。，·矓排，,”孫，。·織 &！〕――嗡’―〕喘 •,�l�м к •!•и�ааа /т; •.a.ws ..м .пи уо Соог а�) в�п. ог!sв.•ао ввlv�i�гэ /v . •1Ув1lг�г �М /а •Ая.ыg . �g :lм4� 0••1д ■ Ti !иеlNbl�вк • Мё[ /Ч 'в1Vг1lгм sвtl /* � tя sot !па >�'Т st> tt� �lsa•v+d «ыD:rмич� +пм � �nl4ьа'т w/�•с w/отя w/я�е �ч/гw tяяN м!•л �•:в�го�•д «ыа ss оо �т . �т тя +ат пет (f� дi!вwsек яи!аео,� «� а � . . ��я �.�от оьs'яs /г о►t'sп (tSN R�вlаlN'МВ ввд аво� вег..м� oro'а /3 we•s ою•в /я ояг'в (яs<а я+•� ••!s�мо•••� leooi уп у�т От6 /я 064 (�N !г!ввw!яи� � � яря /я яяТ.•Т :1SN п+•� �WiO �, �q � п сн (вsNs�юl�э s в•п�+о+г+wк !w«»а ояС�L 0[О�ьеягl �sгЧ "°!аюыs*!а отт 1tLf'9:9wимидвк lwва 900�L Л 0➢'C�LLtS%ТiвпвRВ �lOnVl�Овlр рие swв+tввw� _ _ у� /я яав�т <ssN •го �•+п8 � 1 сяелт яsги � м аи ви <ял) s.� ви � 1 f0ят< ><оят< 111 1PBL fL6 п� Мо%�+�О �пв11 о!9•!/ � � М а11 pi пЧi R.овоава 9•°J б•••0«�гк � � tояес � >�о пw п..м..а м+в�р .пм OD'� �яя!'t) о0'%яL в0 00'16 яL'в t4'Lт 61"9 �я8П: вV Уг9•я!ик aN � 0►'9т W 0я'от i9BOtJ "а'Lя вЧ oL'� Oi'0 AL'т т['0 itяN ieиwll�i $в000т �•д � сяеет) ••ы«р ,пм �.я �•я сяяN •а�+да в+ч в• о�+м! 1м я�ооот � • �.� gк•я i{8N ►пи•!о 4врярт и0 • �,� уя.яд (яяlli Чпв•И /п.в!..I я+М .м � А0.1011Э я+tМ1а9 Ч!0 •�'••�пlО�гд �.я /я �•q� gd'8 00'S тб'L :яяПi М•о•+^1 иаJ ев Рг+м! 1•0 8�o00f +� � �•� /я tf"t я►'f tZ'т яг't (яяги мимt0 4�ооот �«� � �•оу /г аы•ятт ы•�е вл•ят et•п (яяN иав•и мs•вк••r .м! •м • м..в� я+:м1•�а иго � м+м:v»�з msг�a�t�!'�к�lг.� ттlв>�!�!'�!>т,6 тs�т� � tвl9 МК 1г�8 1вlв кк М� I•l�1iв01о 1.вiмд �д !^аА кк WN к p1WlS �