China - Second Loess Plateau Watershed Rehabilitation Project; First and Second Xiaolangdi Multipurpose Project; and Second Tarim Basin Project

Documentof The World Bank ReportNo.: 41122 PROJECT PERFORMANCE ASSESSMENTREPORT PEOPLE'S REPUBLIC OF CHINA SECONDLOESSPLATEAUWATERSHED REHABILITATIONPROJECT (LOAN 4477 & CREDIT 3222) AND XIAOLANGDI MULTIPURPOSE PROJECT I& I1 (LOANS3727 & 4200) AND TARIMBASINI1PROJECT (LOAN 4341& CREDIT 3093) October 4,2007 Sector, Thematic,and Global Evaluation Division IndependentEvaluation Group (World Bank)  Currency Equivalents(annual averages) Currency Unit = Yuan (Y or M B ) , IO0fen = IYuan = I RMB 1994 US$l.OO Y8.70 1997 US$1.oo Y8.30 2000 US$1 .oo Y8.27 2004 US$1 .oo Y8.26 2006 US$l.OO Y7.83 Abbreviations and Acronyms AAA Analytical and advisory services BLM BasinLevel Model GoC Government ofChina EPH Electric Power ofHenan ESW Economic and sector work IBRD International Bank for Reconstruction and Development ICR Implementation Completion Report ID Institutionaldevelopment IDA International Development Association M & E Monitoring and evaluation M & I Municipal and industrial MWR MinistryofWater Resources NDRC National Development and Reform Commission O & M Operationsandmaintenance PPAR Project PerformanceAssessment Report SAR StaffAppraisal Report SLCP Sloping Land Conversion Program VAT Value Added Tax YRCC Yellow River Conservancy Commission YRWHDC Yellow River Water and Hydropower Development Corporation GWh gigawatt-hour (1 million kWh) TWh terawatt-hour (1billionkWh) M W megawatt (1000 kilowatts) FiscalYear Government: January 1-December 31 Director-General,IndependentEvaluation : Mr.VinodThomas Director, IndependentEvaluationGroup (WorldBank) : Mr.AjayChhibber Manager, Sector, Thematic, andGlobalEvaluationDivision : Mr.Alain Barbu Task Manager : Mr.GeorgeT. KeithPitman  1 IEGWB Mission:Enhancingdevelopmenteffectivenessthroughexcellenceand independencein evaluation. About this Report The IndependentEvaluationGroup assesses the programs and activitiesof the World Bank for two purposes: first, to ensurethe integrityof the Bank's self-evaluationprocess and to verify that the Bank's work is producingthe expected results, and second,to help develop improveddirections, policies, and proceduresthroughthe disseminationof lessonsdrawn from experience.As part of this work, IEGWBannually assesses about25 percent of the Bank's lending operationsthroughfield work. In selectingoperationsfor assessment,preferenceis givento those that are innovative, large, or complex; those that are relevantto upcomingstudies or country evaluations;those for which ExecutiveDirectors or Bank managementhave requestedassessments; andthose that are likely to generate importantlessons. To prepare a Project PerformanceAssessment Report (PPAR), IEGWB staff examine project files and other documents, interview operational staff, visit the borrowing country to discuss the operation with the government, and other in-country stakeholders, and interview Bank staff and other donor agency staff both at headquarters and in local offices as appropriate. Each PPAR is subject to internal IEGWB peer review, Panel review, and management approval. Once cleared internally, the PPAR is commented on by the responsible Bank department. IEGWB incorporates the comments as relevant. The completed PPAR is then sent to the borrower for review; the borrowers' comments are attached to the document that is sent to the Bank's Board of Executive Directors. After an assessment report has been sent to the Board, it is disclosed to the public. About the IEGWBRatingSystem IEGWB's use of multiple evaluation methods offers both rigor and a necessary level of flexibility to adapt to lending instrument, project design, or sectoral approach. IEGWBevaluators all apply the same basic method to arrive at their project ratings. Following is the definition and rating scale used for each evaluation criterion (additional information is available on the IEGWBwebsite: http://worldbank.org/ieg). Outcome: The extent to which the operation's major relevant objectives were achieved, or are expected to be achieved, efficiently. The rating has three dimensions: relevance, efficacy, and efficiency. Relevance includes relevance of objectives and relevance of design. Relevance of objectives is the extent to which the project's objectives are consistent with the country's current development priorities and with current Bank country and sectoral assistance strategies and corporate goals (expressed in Poverty Reduction Strategy Papers, Country Assistance Strategies, Sector Strategy Papers, Operational Policies). Relevance of design is the extent to which the project's design is consistent with the stated objectives. Efficacy is the extent to which the project's objectives were achieved, or are expected to be achieved, taking into account their relative importance. Efficiency is the extent to which the project achieved, or is expected to achieve, a return higher than the opportunity cost of capital and benefits at least cost compared to alternatives. The efficiency dimension generally is not applied to adjustment operations. Possible ratings for Outcome: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, Highly Unsatisfactory. Risk to DevelopmentOutcome: The risk, at the time of evaluation, that development outcomes (or expected outcomes) will not be maintained (or realized). Possible ratings for Risk to Development Outcome: High Significant, Moderate, Negligible to Low, Not Evaluable. Bank Performance: The extent to which services provided by the Bank ensured quality at entry of the operation and supported effective implementation through appropriate supervision (including ensuring adequate transition arrangementsfor regular operation of supported activities after loanlcredit closing, toward the achievement of development outcomes. The rating has two dimensions: quality at entry and quality of supervision. Possible ratings for Bank Performance: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, Highly Unsatisfactory. BorrowerPerformance: The extent to which the borrower (including the government and implementing agency or agencies) ensured quality of preparation and implementation, and complied with covenants and agreements, toward the achievement of development outcomes. The rating has two dimensions: government performance and implementing agency(ies) performance. Possible ratings for Borrower Performance: Highly Satisfactory, Satisfactory, Moderately Satisfactory, Moderately Unsatisfactory, Unsatisfactory, Highly Unsatisfactory.   ... 111 Contents PRINCIPAL RATINGS .................................................................................................................. vi1 KEY STAFF RESPONSIBLE ....................................................................................................... Vlll PREFACE ....................................................................................................................................... IX SUMMARY ..................................................................................................................................... XI 1. BACKGROUND AND STRATEGIC CONTEXT ................................................................ 1 The Bank's China Strategy......................................................................................................... Environmental impacts were large ............................................................................................. 2 6 Bank Assistance for Land and Water Management ................................................................. 10 2. PROJECT EVALUATIONS .............................................................................................. 12 SECOND LOESS PLATEAU WATERSHED REHABILITATION PROJECT .............................. 13 Objectives ........................................................................................................................ 13 Design .............................................................................................................................. 14 Implementation ............................................................................................................... 15 Monitoring and Evaluation........................................................................................................ 18 Safeguard Compliance....................................................................................... Relevance ..... ................................... Efficacy. Efficienc Outcome .......................................................................................................................... 23 Risk to Development Outcomes ................................................................................... 24 Bank Performance .......................................................................................................... 25 Borrower Performance ................................................................................................... 26 Summary of Issues raised by the Loess Plateau Project ........................................... 26 XIAOLANGDI MULTIPURPOSE PROJECTS I & II ..................................................................... 27 Objectives ........................................................................................................................ 27 Design .............................................................................................................................. 28 Implementation ............................................................................................................... 30 Monitoring and Evaluation........................................................................................................ 30 Safeguard compliance.............................................................................................................. 32 This report was preparedby George T.K. Pitman(Task Manager) and SunilMathrani(Consultant) who assessedthe DroiectsinOctober/November2006. Soon-WonPak Drovidedadministrative s u ~ ~ o. r t  iv Relevance ................................................................................................................................ 32 Efficacy..................................................................................................................................... 33 Efficiency.................................................................................................................................. 39 Outcome .......................................................................................................................... 39 Risk to Development Outcome ..................................................................................... 40 Bank Performance .......................................................................................................... 44 Borrower Performance ................................................................................................... 44 Summary of Issues Raised by the Xiaolangdi Project ................................................ 45 TARIM BASIN II PROJECT .......................................................................................................... 45 Objectives ........................................................................................................................ 45 Design .............................................................................................................................. 46 Implementation ............................................................................................................... Monitoring and Evaluation.............. 48 Safeguard Compliance......... ....................... Relevance ...................................................................................................................... Efficacy..................................................................................................................................... 54 Efficiency.................................................................................................................................. 63 Outcome .......................................................................................................................... 64 Risk to Development Outcome ..................................................................................... 65 Bank Performance .......................................................................................................... 66 Borrower Performance ................................................................................................... 66 Summary of key findings from the Tarim Basin Project ............................................ 67 3. OVERALL FINDINGS AND CHALLENGES LOOKING FORWARD .............................. 68 FINDINGS...................................................................................................................................... 68 Challenges Looking Forward ........................................................................................ Enabling environmental management through laws and regulations ....................................... 69 70 Overlap and Conflicting Responsibilities for Water Frustrate Reform Progress ....................... 71 Government response to coordination and management issues is slow.................................. 72 China Moves toward a Water-Saving An Increased Focus on Water Savings improved land management has been .................... 77 Progress on essential reforms is being outpaced by growing water demand........................... 79 CONCLUSIONS AND LESSONS ................................................................................................. 80 Annex A .Basic Data Sheet ............................................................................................ 83  V BOXES Box 1: ImprovedIncomes-Waijau Watershed, HoglinVillage, Jining County, Gansu Province................................... 21 Box 2: Growing WUAs -the experienceof Mousonma IrrigationDistrict, Kashgar prefecture....................................... 62 Box 3: Inner MongoliaAutonomousRegion's 2000 Water Trade ................................................................................... 77 FIGURES Figure 1: Productivityof land increased................................................... Figure 2: Farmersgross per capita incomesincreased.................................................................................................. 22 Figure3: IPMwas successfulin reducingdependenceon chemical cont ............................. 54 Figure4 : Farmers'net householdincomesfrom agricultureincreased ............... Figure 5: Yields of major crops improved 1998-2004............................... Figure 6: Ownership of durable goods rose ......................... ..................................................... .57 TABLES Table 1:The Assessed projectsaddress mostof China's land and water problems Table 2: Second Loess Plateau- Objectives, componentsand costs...... ...................................................... 14 Table 3: Factors DeterminingProject Outcome ........................................ ....................... 24 Table 4: Xiaolangdi - Overallproject costs and phasing Table 5: Xiaolangdi I& II-Objectives, ......................................... 29 Table 6: Xiaolangdi ElectricityProducti Table 7: Factors Determining Project Outcome ........... Table 8: Bulkwater prices............ Table 9: Tarim II Objectives, Compo - Table IO: Crop diversificationoccurred -are cultivated Table 11: Changes in Net Household In Table 12: How water revenuesare spent- QinbuluguqiWUA, Aksu Prefecture ......................................... 61 Table 13: Evaluationof factors co Table 14: Water Use in China is Not Efficient   vii PRINCIPAL RATINGS ICR* ICR Review* PPAR ~~~ ~ Second Loess Plateau Watershed RehabilitationProject (Loan 4477Kredit 3222-CHA) Outcome Highly satisfactory Highly satisfactory Satisfactory Institutional High High Development Impact** Risk to Moderate Development Outcome Sustainability*** Highly likely Highly likely Bank Highly satisfactory Highly satisfactory Satisfactory Performance Borrower Highly satisfactory Highly satisfactory Highly Satisfactory Performance Xiaolangdi Multipurpose Project I(Loan 3727-CHA) Outcome Satisfactory Highly satisfactory Highly satisfactory Institutional Substantial Substantial Development Impact** Risk to Significant Development 0utcome Sustainability*** Likely Likely Bank Satisfactory Highly satisfactory Highly satisfactory Performance Borrower Satisfactory Highly satisfactory Highly Satisfactory Performance Xiaolangdi Multipurpose Project I1(Loan 4200-CHA) 0utcome Highly satisfactory Highly satisfactory Highly satisfactory Institutional Substantial Substantial Development Impact** Risk to Significant Development 0utcome Sustainability*** Highly likely Highly likely Bank Satisfactory Satisfactory Highly satisfactory Performance Borrower Highly satisfactory Highly satisfactory Satisfactory Performance  ... Vlll /CR* ICR Review* PPAR Tarim Basin I1Project(Loan 4341-CHA and Credit 3093-CHA) Outcome Highly satisfactory Highly satisfactory Satisfactory Institutional High High Development Impact** Risk to Moderate Development Outcome Sustainability*** Highly likely Highly likely Bank Highly satisfactory Highly satisfactory Highly Satisfactory Performance Borrower Highly satisfactory Highly satisfactory Satisfactory Performance * The Implementation Completion Report (ICR) is a self-evaluation by the responsible Bank department. The ICR Review is an intermediate IEGWB product that seeks to independently verify the findings of the ICR. **ASof July 1,2006, Institutional Development Impact is assessed as part of the Outcome rating. ***AS of July 1, 2006, Sustainability has been replaced by Risk to Development Outcome. As the scales are different, the ratings are not directly comparable. KEY STAFF RESPONSIBLE Project Task Manager/Leader Division Chief/ Country Director Sector Director Loess PlateauWatershedRehabilitationProject(Loan 4477-CHAand Credit 3222-CHA) Appraisal Juergen Voegele Geoffrey Fox Yukon Huang Completion Josef Ernstberger and Mark Wilson David Dollar Anis Wan/ Achim Fock XiaolangdiMultipurposeProjectsI(Loan3727-CHA) Appraisal Daniel Gunaratnam Joseph Goldberg Nicholas Hope Completion Daniel Gunaratnam Mark Wilson Yukon Huang XiaolangdiMultipurposeProjectsI1(Loan4200-CHA) Appraisal Daniel Gunaratnam Geoffrey Fox Yukon Huang Completion Daniel Gunaratnad Mark Wilson Yukon Huang Xiaokai Li Tarim Basin I1Project(Loan4341-CHA and Credit 3093-CHA) Appraisal Douglas C. Olson Geoffrey Fox Yukon Huang Completion Geoff Spencer and Mark Wilson David Dollar Jiang Liping  ix PREFACE This is the Project Performance Assessment Report (PPAR) for four projects that cost US$3.2 billion and included four loans and two credits to the People's Republic o f China over the period 1994-2005. The Xiaolangdi MultipurposeProject was designedas atwo-phase operation. The initial Xiaolangdi Multipurpose Project (Loan 3727), estimated to cost US$2,294.0 millionat completion o fthe secondphase, was approved inApril 1994 for a loan o f US$460.0 million andwas closed on schedule inDecember2000 fully disbursed.The Xiaolangdi Multipurpose Project I1(Loan 4200) was approved inJune 1997 for a loan o f US$430.0 millionat which time total project costs haddecreasedto US$2,855.8 million. The project was closed as scheduledinDecember 2003 when US$80.5 millionwas cancelled. Total project cost at closing was US$2,688.8 million. The TarimBasinI1Project (Loan 4341 andCredit 3093), estimated to cost US$ 272.6 millionwas approved inJune 1998 for a loan o fUS$90.0 million and a credit o f US$65.6 million.US$2.67 million of the loanwas cancelled inSeptember 2001. It closed inDecember 2005 one year later thenplanned, andwas fully disbursed.Total project cost at completion was US$287.8 million. The LoessPlateauWatershed RehabilitationProject, estimatedto cost US$252.4 million, was approved inMay 1999 for a loanof US$lOO million and a credit ofUS$50 million. The project closed after a six month extension inJune 2005 and was fully disbursed. Total project costs at completion were US$240.2 million. Thisreport is basedonthe variousMemoranda andRecommendationsofthe President, a Staff Appraisal Report and Project Appraisal Documents, loan documents, project files (both at the World Bank headquartersand inthe China Country Office), ImplementationCompletionReports anddiscussions with Bank staff. An Independent Evaluation Group (IEG) missionvisited China inOctober-November 2006 to discuss the effectiveness of the Bank's assistance with the central government agencies, development partners, representatives o fprojects' implementingagencies inthe provinces and governorates, and stakeholders and farmers inthe field. The cooperation and assistance of central government officials, managementand staff o fthe implementingagencies, other stakeholders andproject beneficiaries i s gratefully acknowledged. ThisPPAR is part of a series of assessmentsofwater and environment projects in China that were selected for detailed evaluation to inform IEG's Country Environmental Case Studythat i s a key inputto IEG's global evaluation to determine the Effectiveness of WorldBankGroupAssistancefor the Environment. The global study i s due for completion inearly 2008. Following standardIEGprocedures, copies o fthe draft PPARwill be sent to the borrowers and agencies for their reviewand comments. However, no formal response was received.   xi SUMMARY The four projects assessed inthis report provide a great range o f experience central to the Bank's assistance for sustainable landand water planning and development inChinaduringthe period 1994-2005. Eachofthe projects hadtwo phases; however, this assessment includes bothphases o fthe Xiaolangdi Multipurpose Purpose Project, the Second Loess Plateau Watershed Rehabilitation and the Tarim Basin I1Projects. The Xiaolangdi Resettlement Project that resettled 184,000 people i s the subject o f a separate IEGassessment. The four projects were designedto provide integrated management o f land and water resources withintwo o f China's 9 major river basins. They addressedthe issues o f water allocation and water use efficiency for agricultural andnon-agriculturaluses, flood, sediment control and salinity management, andinstitutions for operations and management o f infrastructure. The Loess Plateau I1project and the Xiaolangdi Projects are both situated inthe Yellow River Basin and the benefits o fboth phases o f the Loess Plateau Project directly accrue to the Xiaolangdi dam located downstream. The Tarim Basin i s located inthe far northwest inXinjiang Autonomous province and i s the only large river basin completely within a single province. The SecondLoess Plateau Project's objectives, to help achieve sustainable land development by increasing agriculturalproductionand incomes and improving ecological conditions inthe tributary watersheds o fthe Yellow River, were fully achieved with few shortcomings. Relevance o fthe objectives was high.However, the designthat uses afforestation as a secondary instrumentfor sediment control (after terraces) may not be wholly appropriate giventhe increasing shortages o f water inthe Yellow RiverBasin. Replicatingthe experience o f the first project, the second project terraced sloping land and afforested heavily eroded slopes, installedcheck dams and successfully imposedbans on grazing. Incompensation fanners were helpedto attain more productive and profitable agriculture including forest products, orchards and forage for a substantially increased number o flivestock enterprises. While fanners' incomes increased, the extentto which this is the result ofthe project is difficult to discern due to problems withthe designand implementationo fthe monitoring and evaluation system. However, more efficient and sustainable use o f landand water resources reduced sediment inflows to the Yellow River by an estimated 53 milliontons duringthe life o fthe project, andthis i s projectedto continue at a slightly lower rate indefinitely. Overall economic rates o freturn (ERR) varied from 18 percent to 21percent, close to the appraisal estimate o f 21 percent. The slightly highervalues include the benefits from sediment reduction and carbon sequestration. Although the methodology usedto calculate the ERR i s exemplary, this assessmentbelieves that productioncosts are underestimated and benefits are overestimated due to the presence o f other donor and a substantial number o f Chinese-financed local developments that fostered increased productivity and incomes. Inaddition many rural roads benefitingthe project were constructed by other agencies. This not only changed market opportunities but also provided improved accessto off-farm employment, better education and health.  xii The outcome o f the project i s rated as satisfactory. Bank performance i s rated as satisfactory and Borrower performance i s rated as highly satisfactory. Risk to development outcome i s rated as moderate. There are few local level funds to repair extensive storm damage to terraces and farmers' incentives to maintainterraces at their own expense may bejeopardized by the availability o f substantial subsidies for this purpose inadjacent areas under centrally-funded programs. The Xiaolangdi Multipurpose Projects' objectives inboth phaseswere fully achieved with cost savings. Floodprotectionwas providedto Henanand Shangdong provinces downstream and included productive infrastructure, towns and 103 million people mainly inthe rural areas o fthe North China Plain. Sediment accretion inthe 800 kmlower reach ofthe Yellow Riverhasbeenhaltedandthe riverbedhas beenlowered through the planned release o f artificial floods from the reservoir. The Yellow Rivernow flows continuously to the sea, unlikethe late 1990s when the downstream reacheswere dry for upto seven months a year. Water supply to several large and mediumcities, major industrial centers and downstream irrigation are effectively regulated by the dam. Muchneededhydropower to meetthe peak power requirementsinthe Henan Gridwas supplied, albeit slightly behindschedule-but generation now exceeds appraisal estimates. Although water supply for irrigation has been improved, it i s below targeted levels because o f overall water shortagesinducedby water tariffs that do nothing to reduce consumption, and upstream water diversions inthe Yellow RiverBasin. In addition the provincial government failed inthe second phase to increase power tariffs according to covenants and full-cost pricing o f water released for downstream use has not beenimplementedby the Yellow River Conservancy Commission. Overall efficiency o fthe project i s assessed as substantial. The ERR at completion was 13 percent compared with 20 percent at appraisal. The primary reason for the reduced ERR i s that the volume o f dry seasonflow and sediment inthe Yellow Riverhas declined since appraisal inresponseto growing upstream withdrawal and increasing frequency o f droughts. This, inturn, has lowered benefits from hydropower, agriculture and sediment control. Evenso, the benefits from power and urbanwater supply are probably underestimated becausewater and energy are underpriced and none o fthe benefits fi-om the improvedecological conditions downstream and inthe delta are included. The outcome ofthe first phaseproject that focused on construction i s rated as highly satisfactory. The outcome o fthe secondphasethat focussed on facility management i s highly satisfactory. Bank performance i s rated as highly satisfactory for bothphases. Borrower performance i s highly satisfactory for the Phase Ithat focused on construction management, financing and application o f safeguards. Phase I1performance i s rated as satisfactory becauseo f slow progress on institutional reform and water pricing. Because o f water allocation and pricing issues, Riskto development outcome i s ratedas significant. This i s due to increasing water shortages inthe river basin, an administrative water allocationprocessthat takes no account o fthe economic value o f water and i s subject to political manipulation, and water pricing that substantially undervalues water and distorts its utilization.  ... Xlll The Tarim BasinI1Project successfully contributedto increasing incomes of poor farmers through development o f irrigated agriculture and drainage infrastructure and establishment o f institutionsto ensure sustainable use, development and management o f water resources and landinthe Tarim Basin. The extent farmers' incomes were improved bythe project is uncertaindue to attributionissues not captured bythe monitoringand evaluation system. The downstream effects o f integrated land and water management and the resultingwater savings may have assistedthe resumptionof flow inthe lower Tarim River that hadbeendry for the last 30 years. The uncertainty is becausethe impact of secular changes inclimate andrainfall inthe regionwere not considered by the monitoring and evaluation system. Evenso, restored river flows rejuvenated the natural riverine forests with substantial ecological benefits.While mechanisms were put inplace to regulate water tariffs, lack o fpolitical will frustrated needed increases. The ERRfor the overallproject was 19percent at completion compared to 14 percent estimated at appraisal. This assessmentbelievesthe actual ex-post economic rates o f returnusingthe same economic and financial models could be closer to the appraisal estimate. This i s becausethe cost stream for the ex-post ERR only captured costs directly attributable to the project and fully attributed all benefits to these inputsalone. Yet we know that other international sources o f funding supported development inputswhich directly or indirectly benefitedthe project, as did China's own regional investments.In addition, there were manyregional, prefectural and country inputs that also benefitedthe project -such as China's own comprehensive agricultural development program. Project outcome is ratedas satisfactory. Bankperformance was highly satisfactory and Borrower performance i s rated as satisfactory. Riskto development outcome i s rated moderate. Agriculture inthe basin i s heavily reliant on cotton and grains and their price is subject to markedannual fluctuations due to global factors. Continued provincial government subsidieskeepwater prices low. This could underminethe water conservation efforts promoted by the project as well as the financial sustainability o f water user groups set upto manage local operation and maintenance o fproject infrastructure. These projects provide three main lessons: 0 The Bankneedsto elevateits engagement in China's water resources policydiscussions above the provinciallevelto emphasizethe perverse incentivesthat are created by extensive agricultural water subsidies. Subsidies for irrigation and land managementjeopardize water conservation effort, longer term sustainability o f irrigation and land reclamationprojects andwater availability for towns, cities, industryandenvironmental management. Despite more thantwo decadeso fBank partnershipand several state-of-the art projects piloting new approachesthese tendto be enclave projects. While Chinese senior planners and politicians at the center acknowledge that pricing should play a role inensuringmore efficient and rational allocation and use o fwater this i s still as at an early stage o f applicationand far from politically acceptable inthe provinces. Without such  xiv a change inthe near future water shortages and environmental consequences o fresource mismanagement will continue to grow to crisis proportions. Monitoring and Evaluation requirescapacity building and reorientation to measureoutcomes and impacts.There i s little virtue inimplementing development models for reclaiming degraded lands or for river basin management ifimpacts on welfare, incomes and the economy cannot be accurately measured. While this i s less o f a problem for physical achievements it i s a major problem for measurement of socio-economic outcomes and impacts. Muchmore attention has to be givento understanding the appropriate counterfactual and ensuringunbiased sampling. There i s also an issue around selection o fthe most effective M&Emodel considering the outcomes expected andthe cost o fthe projects. Appropriate training to redress the problems found inM&E i s clearly indicated. Inaddition, more attention shouldbe givento fully utilizing existingChinese resources. The Ministry o f Finance has indicated that it has a national network of monitoring that include socio-economic indicators and that consideration should be givento working with themto triangulate evaluative findings. Muchgreater care is required indealingwith issues of attribution when estimating ex-post economic rates of return. Specifically this means clearer definition o fthe project counterfactual and a systematic accounting o fthe effects o f exogenous actors and investmenton project impact. Without such attention to these exogenous factors there i s a danger that estimated ERRS give a false impression o fproject impacts. This, inturn, may lead to incorrect lessons on development effectiveness and the efficacy o fpolicy, institutional andengineeringmeasuresutilizedto achieve development objectives. Vinod Thomas Director-General Evaluation  1 1. BACKGROUNDAND STRATEGIC CONTEXT 1.1 After the establishment o fthe Peoples' Republic o f China in 1949 government put flood control and the planned utilization o f its water and land at the top o f its agenda. Together these would buildfood security and contribute to China's industrialization. Great importance was attached to establishing river basinagencies for the Yangtze, Yellow and the Huaihe rivers and inthe 1960sthree more were added-the Haihe, Pearl and Songliao river basins. Subsequently groups o f smaller riverswere classified inthe Southwest, Southeast and Inlandriver basins bringingthe total to nine.Early efforts focused primarily on large-scale, multi-purpose infrastructure development programs aimed at flood prevention, waterlogging control, irrigation, hydropower, navigation, and industrial and urbanwater supplies.Thus for example inthe Yellow River basin 156 Sino-Soviet projects were outlined by 1952andthe FirstPeople's Assembly approved in1955the first comprehensive and capital intensive utilizationplan for the whole basin that focused on power generation and flood control through the construction o f 46 large dams.' Similar multi-purpose master plans were developed for the Huaihe (1956), Haihe 1957), Yangtze (1958), andPearl (1959). 1.2 Although the capital intensive approach waned duringthe 1960s, subsequent water development proceededunder reduced central government support usinglower cost, locally-funded projects. The start o f de-collectivization and the HouseholdResponsibility System inthe 1980sgreatly increased farmers' incentives and productivity but also ledto a markedreductioninstate andlocal government funding for routine operation and maintenance o f large irrigation infrastructure as demand for municipal and industrial investmentgrew. Withthe increasing maintenance backlog many irrigationprojects became unreliable, water use efficiency declined and national irrigation area contracted evenwith newconstruction.* Despiteyield improvements, bythe mid-1980sgrain production had stagnatedas the area sown decreasedby nearly20 percent. Subsequently this was redressed by greaterinvestmentinthe agricultural sector and water mana ement, and introduction o fthe Provincial Governor's Responsibility Systeminthe 1990s! By the mid-1990s agriculturalproductivity rose to meet government's targets. 1.3 Despite these set-backs, between 1949 and 1990the number o f large and medium dams grew from 23 to over 3,000, over 200,000 kmo f flood embankments were constructed and China's irrigation area expandedfrom18 millionto 48 millionha or about 1.Guordano,M.et alia. 2004.Water Management inthe Yellow RiverBasin:Background,CurrentCritical Issues andFutureResearchNeeds.ComprehensiveAssessmentResearchreport 3. Colombo, Sri Lanka. ComprehensiveAssessment Secretariatofthe CGIAR. 2. Lohmar, Bryanet alia., 2003. China's AgriculturalWater PolicyReforms-Increasinginvestment, ResolvingConflicts andRevisingIncentives.AgriculturalInformationBulletinNo. 782. EconomicResearch Service.USDA. 3, Provincialgovernorswere requiredfrom 1995to take responsibilityfor the grainbalanceintheir jurisdictions inorderto raise localself-sufficiency.Yang, Hong. 1999.GrowthinChina'sGrainProduction 1978-1997: A DisaggregateAnalysis. WorldDevelopmentVolume27, Issue 12 ,1999 4. Irrigatedareawas 45.6 millionha in 1980and44.6 millionha in 1986.Irrigatedareagrowthresumed in 1990andby 2004 the area is estimatedto be about 54 millionha.  2 20 percent o fthe world's total. As stated by H e Changhui, Assistant Director General o f FA0 in2 0 0 5 ~"China has made thegreatest achievements in water resources development ~ and conservation.Feeding 22percent of the world'spopulation with only dpercent of the worlds water resourcesand 7percent of the worIdS arable land -as well as lifting 200 millionpeople in rural areas out ofpoverty in the short time span of 20years - is a testimony to its success. Thecontribution of China to worldpoverty alleviation andfood security is remarkable." These remarkable achievements, however, came at considerable cost to the environment. Environmentalimpactswere large 1.4 Ownership o fwater i s deemedto be the property o f all citizens and the state. Under such a property regime water was exploited as an open-accessresource on a first-come, first served basis. While more rationally plannedcentrally-funded projects led development upto the 1970s, thereafter the provinces, municipalities andtown ships increasingly utilized local water resources for their own benefitusingmainly their own resources, be it free labor and/or local finance. A consequenceo f this was the declining quality o f management and stewardship o fthe nation's water and landresources. 1.5 Between1949 and 1998 per capita water use increased 130percent andtotal water use by 430 percent. Demandfor industrial and municipal water grew more than three times faster than that for agriculture.6 The unevengeographical distribution o f water resources in China compounded the problem; the six northernriver basins contain 44 percent o f China's population, two-thirds o f its cultivated areabut only 13 percent o f its water resources. Thus inthe northwater availability, increasingly reducedby droughts andupstreamdiversions for agriculture and industry, i s a major constraint to economic development althoughthis has not always beenthe case.79 Management o f floods andwater pollution are among the major challenges inthe south-east. 1.6 Over-exploitationo fwater resources inthe Tarim Basin inthe arid northwest ledto the loss oflandresources.' Untilthe late 1990sdesertification hadincreasedby about 7,000 km2a year primarily because dams and barrageshad divertedalmost all surface water and considerable amounts o f groundwater for agricultural use. As a result o f inadequate leaching and salinization o f soils some 370,000 km2o f farmland were abandoned and the 5. He, Changhui. 2005. Water ResourcesConservation and Saving. Speech delivered to the Special Session o fthe China Development Forum: Building a ResourceEfficient Society. June 2005. Beijing. 6. Lohmar, Bryan et alia., 2003. op., sit. 7. Overall runoff from the Yellow River basin declined by a third between 1956 and 2000: from 65 billion m3 in 1956-70 to 43 billionm3inthe period 1991-00. Yellow River Conservancy Commission. Personal communication. November 2006. 8. A similar period o f reduced discharge affected the Yellow River in the period 1922-32 when the average annual flow was 24 percent less than normal. MOWR October, 2006. 9. Qi, Feng and C Chen. 1998. Current situation, problems and rational utilization o f water resources in arid north-western China. Journal o f Arid Environments. 40. 373-382.  3 Tarim River ceasedto flow with considerable adverse impacts on ecology and regional biodiversity.lo 1.7 The adverse effects of generations ofpoor uplandmanagement amplifiedby population pressures have also seriously affected China's landresources. Nationally about 5 billiontons o f soil i s eroded each year from an area o f about 3.6 million km2 or about one-third of China's territory. Over halfo f all erosion is attributedto poor water management inuplandareas augmented by overgrazing and loss o f native vegetation." Worst affected areas include the loess plateau and the redand black soil areas o f northeastern and southern China. About half o f all soil erosioni s inthe Yangtze River Basin, and a third inthe Yellow River Basin. The impacts are substantial. Flashier monsoonal runoffincreased flood hazards and sediment inflow into major rivers. Reduced soil water storage affects dry-season vegetation and agriculture. 1.8 Farmers' copingmeasures such as increasing rangeland livestock have only exacerbatedthe problem. InGansu, for example, overgrazing o f grasslands and desertification has reduced the Maqu Wetlands along the Yellow River from 70,000 to 20,000 ha.'* Since the 1960s, increased soil erosion has reduced reservoir storageby about 10% and reduced navigable waterways from 172,000 to 108,000 km. Andthe bedo fthe Yellow and many other rivers are elevated above their floodplains because o f excessive deposition o f sediments that greatly increasesthe risk o f flooding. There i s also a very high correlationbetweenthe 217 counties with a highincidence o frural poverty and dwindling soil productivity - 87 percent o fthese counties suffer from soil er~sion.'~Restoring degraded landsand their vegetation also provides a cost-effective tool for carbon sequestration o f a scale that could have significant global benefit^.'^ Thus tackling the causes o f soil erosion not only reduces the risk o f floods andextends the life o f expensive hydraulic infrastructure but also benefits China's poorest farmers. 1.9 The Yellow River is hugelyimportantto China. One innine (over 130million) Chinese live within the Yellow River Basin, and most o fthese people depend, directly or indirectly, onthe river for their livelihood. The river descends 4,450m over its lengtho f 5,500 km, draining an area o f 800,000 square kmandtraversing nineprovinces andor autonomous regions en route to the Yellow Sea inthe Gulfo f Bohai. A large area o fthe North China Plain outside the Basinalso depends on the Yellow River for water supply. 10. Yang, Xiaoping, J Dong and P.D. White. 2006. The KeyRole o f Water ResourcesManagement inthe Ecological Restoration in Western China. Geographical Research. 44(2) 166-154. June 2006. 11. Liu, Zhen. 2004. Soil and Water Conservation in China. Proceedings o fthe 9* International Symposium on River Sedimentation. October 18-21. Yicang, China. 12. Xinhua News Agency. 2004. Sand ChokesYellow River's Cistern. China Internet Information Centre July 20,2004. 13,Ning, Datong. 1997. An Assessment o fthe Economic LossesResulting from Various Forms o f Environmental Degradation in China. http://www.library.utoronto.ca/pcs/state/chinaeco/land.htm 14. Parham, W. 2001. DegradedLand: South China's Untapped Resource. Grown inblocks, trees can sequester 15 tons o f carbon per hectare.The Chinese government launched the Sloping Lands Conversion Program in 1999 and pays farmers to plant and conservetrees inwatershed protection areas. The program is ambitious: a target o f about 15 million hectaresby 2010 and a total budget o fabout $40 billion  4 1.10 The Yellow Rivercontains the highest sedimentconcentration of any river inthe world - at times and inplaces it can resemble a mudslide a mile or more inwidth. Violent storms inthe middlereaches o fthe river can leadto sedimentconcentrations o f more than 400 kgper cubic meter.The middlereach includes the Loess Plateau which contributes 90 percent o fthe river's sedimentload. 1.11 The aggraded lower reach is already dangerously elevated, inplaces more than 10 meters above the surrounding plain. Silt accumulation causes the river bedto rise inexorably and as a result flood embankments have to be raised at the rate o f one meter every 10 years. There are 1400 km o f dykes along the banks of the lower reaches o f the Yellow River. The lower reach contains most o fthe irrigated land inthe basin and i s the most productive interms o f agricultural output. However the lower reach suffers from three handicaps: inperiods o f shortage it i s last inline to receive water; it i s most prone to flooding; andit must deal with most o fthe silt deposited inthe river from upstream regions. 1.12 Since the start o frecorded history, floods along the Yellow Riverhave killed millions, disruptedthe lives o ftens o f millions, and have caused incalculable property and economic losses. Efforts to control it go back 4,000 years. Since the founding o fthe PRC, extensive systems o f dams and dykeshave beenbuiltto preventfloods, and there has not beena single serious breach inthe past 50 years. Butthese defences would have been insufficient to contain really major floods, such as that o f 1843, which i s estimated to have produced a peak flow o f 35,000 cubic meters a second at Xiaolangdi. Floods may strike virtually anywhere inthe basin, both inlate summer as the result o f intense rainstorms, or even inthe dead o fwinter as ice formations dam the northernreaches o fthe river inInner Mongolia and lower reaches near delta. Flood and sediment control are closely linkedand without sedimentcontrol, effective flood control may be impossible. 1.13 O fthe total historical average15annual runoff o f 58 billion cubic meters inthe Yellow River basin, about 30 billion cubic meters are drawn off for various uses, including transfers to cities outside the basin. There has been a water supply shortfall o f about 3 billion cubic meters inrecent years, compared to total demand, despite the additional suppliesmade available as a resulto fthe constructionofthe Xiaolangdi reservoir. About 6 millionhectares are irrigated by the Yellow River and its tributaries,o fwhich 2 millionha are inthe lower reaches which are the most productive due to the temperate climate that allows year-round cultivation.l6 1.14 Earlierconcerns that Yellow Riverfloods would bethe dominant management challenge inthe north also appear misplaced. A series o f seven large dams has effectively providedprotectionagainst the 60-year flood. Evenso, excessive sedimentation has seriously reduced flood storage capacity. The Soviet-designed Sanmenxia dam designed for hydropower and flood control was effectively abandoned inthe late 1980s as most o f its storage capacity hadbeenfilled by sediment-the new Xiaolandgi dam downstream has now taken over these functions. 15. Inthe last decade, the runoffhas been under 50 billion m3. 16. About a third are located inHenan province and two-thirds in Shandongprovince.  5 1.15 Droughts have recently been a greater problem than floods. While the seasonal dryingo fthe lower reach o fthe Yellow Riverinthe period 1972-1990 occurred for an average o f 50 days in21 o f 27 years, by 1997 the lower 704 kmbelow Kaifeng was dry for over 226 days. Outflow to the Bohai Sea occurred for only 27 days. L o w flow, for example, in 1995 reduced water availability to Dezhoucity by almost half forcing limited production or suspension o f 139 industrial enterprises. Reduced availability o f surface water and low levels o f irrigation efficiency inthe north increasedthe demand on groundwater that has yieldedsubstantial economic benefits givenits greater reliability more efficient use. Insome areas o fthe north China plain, however, this has ledto resource miningas demand exceedednatural replenishment, especially inurbanareas. Inthe past 20 years groundwater levels have fallen typically by 10to 50 meters causing land subsidence, saline water intrusionincoastal areas, and disrupteddrainage systems.l7 1.16 Although the issue o f flood management appears to be no longer a major problem inthe north, thisis notthe caseinthe southandcentralparts ofChina. Excessiverainfall over the Yangtze River basin in 1998 causedthe worst floods o fthe century killing 3,000, making 14 millionhomeless and costing China $24 billion ineconomic losses. Although 1999 was not as wet, further extensive flooding occurred as stressed levee systems failed. The subsequentofficial enquiry attributed the impact to the loss o fflood plain storage due to development andsedimentinfilling, reduction o fupstream forest cover by half since the 1950s and a marked increase insoil erosion.18Indeed, some experts fear that short-sighted economic exploitation o fthe land resources within the Yangtze River basin could cause it to become another Yellow River as 41 percent o fthe Yangtze basin's area suffers from sever soil erosion problem^.'^ Some o f the economic loss and human suffering was because encroachment on the floodplain following removal o f flood levees by local govenunent. 2o And 960,000 people at riskinfour provinces withinthe basinhave beenremoved from flood diversion and retentionareas since 1998. 1.17 The pollution from increased agriculturalproduction, domestic andindustrialuse also reducesthe availability o f good quality water. More than a third o f China's lakes are highlypolluted. In2004 only a quarter o fChina's riverswere inthe highestwater quality category; conversely a third were classified as highly polluted. Although industrial discharges are large and highlypolluting, this has been effectively re ulated since 199S21 but severe problems remain, particularly for drinkingwater sources?' Pollution is worst in the north and east especially incities due to lack o f urbanwaste water treatment. Agriculture i s a close second to municipal areas as a source o fpollution, specifically the difficult to treat non-point source pollution from fertilizer andpesticide runoff and livestock waste.23 17. MOWR. 2005. EvaluationResultso fNational Groundwater and Environment Investigations. 18.US.Embassy,Beijing. After the Floods: Water Control inthe Yangtze.November 1999. 19. People's ChinaDaily. Soil ErosionEndangersChina's Longest River. July 30,2001. 20. For example, inMay 1992HuarongCounty, HunanProvince, removedone km of flood dyke to open 80,000 hato agriculture: as aresult 700,000 peoplelost their flood protection. 21. SEPA. 2004. Reporton the State ofthe Environment. 22. Liu, Yingling. 2006. China's DrinkingWater SituationGrim; Heavy Pollution to Blame. Worldwatch Institute.August 3,2006. 23. World Bank.2006. China Water Quality Management-Policy and Institutional Considerations.  6 1.18 China's RehabilitationMeasuresPrecededBankInvolvement.Since the 1970s the objective o f landmanagement has changed from a primarily productive focus to one in which it i s recognized that they also provide environmental services -particularly for water management and reductionof desertification. According to MWR by the endo f 2005 almost one millionha o f erosion mitigationhadbeenimplementedincluding 0.4 millionha o fwatershed improvement and 0.6 millionha o f ecological re~toration.~~China implementedtwo complementary approachesto reducingerosion: (a) reducingthe causes o f erosion by halting the loss of uplandvegetation and forests and (b) restoring degraded landscapes. 1.19 Starting in 1978, the government o f China launched massive forestry and agroforestry programs to reverse deforestation and landdegradation, including the "Three North" project, located inthe aridand semi-arid zone o fthe northwest, north central, and northeast China. Thanks to these large investmentsintree plantation and shelterbelt development and a natural forest logging banmanagedby the State Forestry Administration, China has successfully turnedthe tide o f formerly rapid deforestation. Evenso, regulationand enforcement remainproblematic becauseo fperverse local incentives associatedwith the Provincial Governor's Responsibility System." 1.20 On a much smaller scale, integrated watershed management was specifically targeted by the MWR since the early-1990s. Implementedthrough its Department o f Water and Soil Conservation activities include participation o frural households inafforestation, small dam construction and terracing o f landto reduce water erosion. 1.21 Inview ofproblem discussed above there were six broadissuesneedingurgent attention where the Bank's global experience was relevant: 1. Water resourcesplanning and allocationusingeconomic criteria and instruments; 2. Flood, drought, irrigation and sedimentmanagement; 3. Rehabilitationo fwater infrastructure, particularly irrigation and drainage, and institutionsto improve operation and maintenance; 4. Introductiono fmeasuresto reduce poverty indegraded landareas; 5. Land and soil conservation, salinity control and watershed management; and 6. Pollution control and management. The Bank's China Strategy 1.22 Improving China's environmental and water management was an important component o fthe Bank's Country Assistance Strategies (CASs) from 1995 to 2006. Initially these CASs envisioned environmental improvements as a result o f the broader, market-oriented reforms, more efficient infrastructure and improved institutions. It was expected that these reforms would bringprice changes to reflect the true cost o fnatural resourcesand encourage greater resource use efficiency. Later CASs aimed to improve the 24. MWR. 2006. op cit.,. 25. Rozelle, S, J. Huang, S.A. Husain and A. Zazueta. 2000. From Afforestation to Poverty Alleviation and Natural ForestManagement. Operations Evaluation Department. The World Bank.  7 environment by helpingto shift China from its dependence on industry, which has beenthe main source o fpollution duringChina's growth over the last three decades, by encouraging the growtho fservice sectors, enhancing marketefficiency andimproved institutionsfor water management and utilization. 1.23 The 1995 CAS sought to assist China's sustainable growth challenge through "easing long-standing constraints to growth by alleviating infrastructuralbottlenecks, particularly inthe energy, transport, water, and communications sectors; andcurbing the spreado f environmental problems amidst rapid urbanizationandthe flurry o f development activities." Specifically, the Bank Group's objectives and assistancestrategy were to assist China achieve macroeconomic stability while maintainingrapid and more efficient growth, and sustainthis growthby shiftingresourcestoward priority areas, including environmental protectionandpoverty reduction. Assistance for the rural sector focused on sustainable landuse, including the development and management o f forest resources, improved land use on degraded and marginal soils, development o f newwatershed management and protectionmodels, and safer drinkingwater. 1.24 The 1997 CAS maintained the same objectives but shiftedthe emphasis. Responding to increasing pressures onthe environment from the rapid industrial, urban growth and increased agricultural intensification, the Bank continued with a strategy o f financing investments for near-term environmental quality improvements. But it also aimed to strengthenChina's institutions, namely the SEPA and its policies. New objectives for rural investmentswere introducedsuch as integrated coastal zone management and pollution reduction were added to the older ones, such as irrigation and water management, forest conservation and biodiversity. 1.25 What distinguishedthe 2003 CAS from earlier ones was that more innovative particularly becausethe lothFive-Year Plan(2001-05) gave special attentionto instruments and a wider range o f them were neededto assist China meet its objectives, environmental issues. Environmental deterioration and the declining productivity o f many natural resources were associatedwith poverty and unsustainable productionpractices, especially inthe environmentally fragile and poorer central and western provinces. While the CAS'Sfirst pillarwas growth-oriented and sought to improvethe business environment and helpto accelerate the transitionto a market economy, its second pillar addressed concerns about regional imbalances and poverty. Poorer and disadvantaged people and lagging regions were to be addressedthrough investmentlending inrural development, infrastructure and social sectors. These included improvements to water, land and natural resource management. 1.26 The maturing o fthe Bank Group's relationship with China is reflectedinthe 2006 Country Partnership Strategy (CPS). Inadditionto continuing the Bank Group's assistance on reducing poverty and mitigating rising inequality between coastal and inlandprovinces, the CPS aims to deepen assistance for financial intermediationand improvements inpublic andmarket institutionsto facilitate China's integrationinto the world economy. Onthe environment, increasing the efficiency of energy generation and water use and reducing pollution were seen as priority issues.  8 1.27 BankStrategyfor Landand Water was formalizedin the mid-1980s.The Bank's first agriculture sectorreview drew attentionto the problems of China's degraded land and the needfor complementary approaches to land and water management. Inthat sense it laid the foundation for the Bank's investmentsinthe Loess Plateau andthe Tarim Basin. 1.28 The Bank's 1985 China - Long-term Issues and Options evaluated key issues facing China and made recommendations for the Bank's future engagement over the period 1985- 2000. The overriding concern was to achieve China's ultimate economic objective o f catching up with developed countries while ensuringthat the benefits o fprosperity are widely sharedwithinthe socialist system. While agriculture would continue to be largest andmost important sector of China's economy inthe short to medium-term,the report clearly saw the general direction o f development as a shift away from activities that are constrained by fixed resources (land and water) and a shift into improving human capital for modern agriculture including agro-industrial production. 1.29 Evenso, investmentinirrigation and drainage was suggestedto increase production and offset the highloss o f cultivable landconverted to non-farm uses. Drainage improvement was seen as a priority activity inthe NorthChina. Expansiono f irrigated area was seen as less importantthan improvedwater management inexisting irrigation systems, a highproportiono f which lackedadequate facilities for distributionto the farm level. Such an approach recognizedthat competition for available water insome parts o f China between agriculture, industry and human consumptionwill increase and make agricultural expansion more difficult. 1.30 Based onthe results o f Chinese pilot studies the Bank recommended that the productivity o fupto 20 mha o f loess soils mightbe substantially improved if economically justified after further study. Major elements inthe improvement o f such lands include the conversion o fthe most erosion-prone croplands to grass and trees, control o f grazing particularly by sheep and goats, and the plantingo fbetter adapted crop varieties. 1.31 Inthe poorest provinces agricultural development was seen as possibly the only way to increase incomes where degraded land reduced production capacity and made livestock productionunprofitable. Thus the Bank's recommendations targeted both likely interventions and regions- but with reservations. The typically resource-poor areas o f extremepoverty inthe northwest and southwest were apriority but it cautioned that investment may not bejustified on economic grounds alone. Development constraints to address were lack o f irrigation, salinization, depletedpasture and harsh climate inthe northwest, poor soil and inadequate drainage inthe southwest and weak transport infrastructure inboth regions. Inthe worst-off areas, the Bank identified out-migration as potentially one o fthe only feasible policy measures. 1.32 The Bankintroducedintegratedriver basinplanning.The Xiaolangdi multipurpose dam project was the logical outcome o fthe Bank's first major piece o f analytical and advisory work aiming specifically at water and sediment management: the 1993 Yellow River Basin Investment Planning Study. The study, using a simulation model and non-linear optimization techniques, determinedplanning priorities for the basin, identifiedwhich o fthe government's proposed investmentstotaling US$10.8 billionwere  9 economic, and proposed an economically viable development program coveringthe period to 2010. 1.33 Unlikethe earlier planningexercise carried out byYellow RiverConservancy Commission(YRCC) in 1988,26it brought all available data together ina consistent framework and put economic, rather then engineeringviability as the decision variable. The multivariate simulation model developed by the study team was subsequently mainstreamed by YRCC. The results it produced challenged YRCC's ambitious plans to increase irrigation inthe basinby 40 percent. It found that only halfo fthe proposed area o f 1.7 millionha was viable giventhe water requiredfor sediment flushing and meetinga doubling o fmunicipal and industrial demand. And even the smaller expansion o f irrigation would bejeopardized ifthere were no improvements to agricultural water use efficiency on the existing irrigatedarea o f 2.9 millionha. The study also revealed that China's basic water, landand agricultural planningdate was or poor quality, piecemeal and sometimes inconsistent. 1.34 The Bank undertook a major environmental sector reviewin2000, Clear Water, Blue Skies,and produced a Country Water Resources Assistance Strategy (CWRAS) and an Agendafor Water Sector Strategyfor North China in2002. Together their findings and strategies affirmed the continued relevance o fthe development objectives o fthe three assessedprojects and the Bank's water and land development portfolio o fprojects. 1.35 While Clear Water,Blue Skies (2001) focused primarilyon air andwater pollution andtheir relationto economic growth, itproduced recommendations for action onwater that support the objectives o fthe river basinmanagement aspects o fthe projects. Notably, the report highlightedthe mutually reinforcing nature o f higher levies and economic reforms incontrolling pollution. Specifically, inthe short-term it recommended strengthening river basinmanagement by consolidating government agencies and allocating responsibilities to fewer institutions; providing technical and financial assistance to disseminating water-saving technologies for agriculture and industries; andto protect water resources from overexploitation, introductiono f regulation and a water extraction tax. Inthe medium-termit recommended price adjustment to encourage water conservation. 1.36 TheAgendafor WaterSector Strategyfor North China andthe CWMS are less sanguine. They found that the traditional water management practices will not support continued sustainable growth inthe 21Stcentury. They predictedthat acute water shortage and pollution inNorth China will soon become unmanageable unless more significant, comprehensive and sustained commitments are made to implement strategies and initiatives to bringwater resources utilization back into a sustainable balance. While supply augmentation i s probably necessary, this cannot be achieved without pricing, management andregulatory reforms inall areas o fwater resources management. 26. YRCC. 1988. RevisedSummeryo fthe Report onYellow RiverHarnessingand DevelopmentPlanning.  10 BankAssistance for LandandWater Management 1.37 Inresponseto China's needs theBank started lendingin1981andsincethentotal lendingacross all sectors has exceededUS$39billion. 27 About a quarter o f this was for rural development and 17percent -US$6.65 billion-was to assist improved management o f landand water. 1.38 Lendingfor agricultural water management started in 1982and a series o f23 projects has enabled the Bankto assist most regions o f China. The majority o fthis assistancewas for province-specific area development projects that initially focused on extending and completing irrigation infrastructure inexistingprojects. And since the early 1990sinstitutional development became increasingly important. Similarly, lending for land management started in 1983 through a series o ften projects, each covering several provinces ina region, focusing on areas needingbetter land andwatershed management to alleviate poverty, conserve water and reduce erosion and flooding. 1.39 The three projects that are the subject o fthis assessmentspannedthe period from 1994 to the end o f 2005 and embodied important newinitiatives and more comprehensive and multi-sectoralapproachesto landand water management. All three projects deal with different aspectso friver basinmanagement incentral and western China. The Loess PlateauProject addresses the issue o f sustainable land and watershed management inthe Yellow RiverBasin. The other two projects support introduction o f integrated river basin management to the Yellow River and Tarim RiverBasins. Each o f three projects hadtwo phases. Inaddition, the Xiaolangdi Project includedinvoluntary resettlement o f over 184,000 people that was assessed at the same time andthis i s reported separately.28 1.40 Rationalefor the Loess PlateauProjects.Initially, the Bank provided its support to mitigate the poverty impacts resultingfrom land degradation inthe relatively poor red soils areas insoutheast China through its two Red Soils projects. The first RedSoils Project was largely successful indemonstrating the state-of-the-art technology for watershed management on especially favorable sites. Emboldened by the success o fthis approach in 1994 the Bank embarked on a second-phaseRed Soils Project (a typical Red Soils watershed included the steeper sloping, shallower soil areas o fthe upper area, where the role o f forestry would be more important, as well as the gentler sloping, terraced landmore typical o fthe first RedSoil Project areas.) At the same time the government expanded the watershed management portfolio to tackle a more challenging area-the Loess Plateauthat covers 640,000 square kilometers o fthe upper and middle parts o fthe Yellow River Basin.29This follows the recommendation o fthe 1993 Yellow RiverBasinInvestment 27. Over the period 1981-2005 the IDACredits were US$29.13 billion andIBRD Loans were US$9.95 billion. 28. IEG.ProjectPerformanceAssessment Report: CHINA XiaolangdiResettlementProject.ReportNo. June, 2007. 29. The Loess Plateau, namedafler the yellowish soil that coversthe area, is the biggestsuchregioninthe world. Boundedbythe QinlingMountainsandthe Weihe Plaininthe south, the Great Wall inthe north, the TaihangMountainsinthe east andthe TaoheRiver andWuxiao Mountainsinthe west, it coversall ofShanxi Province,the northernpartof ShaanxiProvince, mostoftheNingxiaHuiAutonomousRegion, centraland eastern areas o f GansuProvinceandwesternsectiono f HenanProvince.Comprising400,000 square kilometersandrising800 to 2,000 metersabove sea level, it is China'sthird-biggestplateau.Exceptfor a few  11 Planning Study (1.41.)These second generationprojects introduceda small watershed approach that encompasseda broader range o f site conditions andproven environmental protectionmethods. 1.41 Rationale for the Xiaolangdi Multipurpose Project. The Bank's Yellow River Basin Investment Planning Study demonstrated that only some ofthe major infrastructure projects envisaged by YRCC were economically viable. By carefully examining the seasonally variablevalue o fwater for agriculture and sediment control the study determinedthat construction o fthe Xiaolangdi multipurpose dam andexpansionof sediment control efforts inthe upstreamloess plateau was requiredandthat bothwould be economic. The study also demonstratedthat water was pricedfar belowits opportunity cost -agriculturalwaterpriceswereonehundredthofthemar inalvalue.Withinthebasinthe price elasticity o f water was estimated to be close to -1.0." Thus the study recommended that raising water fees should be a major part of the government's strategy to conserve water and maximize economic returns. 1.42 Rationale for the Tarim Basin Project. Situated inthe border ofnorthwest China, the Tarim Basinforms the southernhalfofXinjiang UygurAutonomous Region. With an area o f slightly over half a million square kmand a populationi s about 7 millionthis closed basini s surrounded by mountains onthree sides. Traditionally, agriculture on the piedmont margins o fthe basincentered on oasesfedbythese rivers or from Kere~esThe~total . ~ populationinthe project areain 1998 was 4.7 millionpeople o f whom 3.6 million lived in rural areas. The agricultural labor force was 0.94 million andthey manageda cultivated area o f about 0.56 million ha. By completion the total project population hadincreased by 50,000 people and the agricultural labor force by 4,000.32 1.43 Annual precipitationranges from 1000 mmover the mountains to less than 20 mm inthe center ofthe basin. The TarimRiver- China's longest inlandriver(1300 km) - drains the basin eastwards around the Taklamakan Desert into Taitema Lake. Numerous smaller rivers runoff the mountains but most disappear into the desert. Prior to the start of reservoirs and irrigation works (1957) the Tarim's waters reachedLopNor (now a salt- encrusted lake bed). Historical annual flows average over 5 billion cubic Subsequent upstreamdiversionscausedthe riverto becomedry over the 300 kmreach upstreamofTaitema Lake since 1972.34This 300 kmreach-referredto as the "green corridor" -supported a forest ofpoplar and Chinese tamarisk and related flora that formed highlands and large river valleys, it is covered with a layer of loess 100to 200 meters deep. According to historical records, most o f the plateau was covered with dense forests, lushgrasslands and fertile soil. But predatory reclamation, indiscriminate felling o ftrees and overuse o f grasslands as well as destruction by frequent wars stripped the area o f nearly all its forests. Each year, more than a billion tons o f mud and silt are swept from the plateau into the Yellow River, the cradle o f the Chinese civilization. 30. Price elasticities to demand: domestic water -0.3 to 0.6; industry -.045 to -1.37; irrigation -.037 to -1.5. 31. The traditional irrigation supply in the Tarim basin i s the Kerez system. Gently sloping tunnels collect ground water from the piedmont areas and provide oasis irrigation via shallow wells or small reservoirs. These tunnels were excavatedover 2,000 years ago and the total lengthis about 5,000 km. 32. Minority ethnic groups in China are not subject to the "one family, one child" rule. 33. Zhao Sonqiao and Xucheng Xia. 1984. Evolution o f the Lop Desert and Lop Nor. The Geographical Journal. 150(3), pp311-321. 34. Taitema Lake is located about 160 km southwest o f LopNor.  12 a barrier to the northward expansion o fthe desert.35Inaddition biodiversity was significantly reduced. Over-irrigated areas upstream became salinized and this produced poorer quality inflows to the river downstream. Together they reduced agricultural productivity and adversely affected farmers' and pastoralists' livelihoods. These problems andthe gradual loss ofthe green corridor also focused nationaland internationalattention onthe consequences o f environmental mismanagement. Thus recreation o f the lower Tarim Riverto sustainthe green corridor became a nationalpolicy issue. Its recreation also became a key outcome indicator o fthe integrated effects o f efforts to upgrade irrigation engineering, agriculturalpractices and institutions inthe context o friver basin management. 1.44 These projects were relevant to the Bank's global and country objectives andtheir objectives and components were closely matched to China's development needs, Table 1. Table 1: The Assessed projectsaddress most of China's land and water problems Developmentissue Loess Plateau Xiaolangdi Tarim Basin 1. Water resourcesplanning and allocation using 0 0 economic criteria and instruments 2. Flood, drought, irrigation and sediment 0 0 0 management 3. Rehabilitationo f water infrastructure, particularly 0 0 irrigation and drainage, and institutions to improve operation and maintenance 4. Introduction of measures to reduce poverty in 0 0 degraded land areas 5. Land and soil conservation, salinity control and 0 0 0 watershed management 6. Pollution control and management 0 2. PROJECTEVALUATIONS 2.1 This chapter systematically evaluates how effectively eachproject met its objectives andrates its outcome. Following a review o fthe relevance o fproject designand implementationexperience, the efficacy and efficiency o f each project i s assessedand outcome determined. Findingsand lessons drawn are presented inChapter 3. 35. Betweenthe 1950s and 1980s the area o fPopulus Divesfolia shrank from 54,000 ha to 16,400 ha and 132,000 ha o f farm land were desertified along the Tarim river and Kinqi river. Source: www.unu.edu.unupress/unupbooks/uu02fe/uuO2febOb.htm  13 SECOND LOESS PLATEAU WATERSHED REHABILITATIONPROJECT 2.2 Second Loess Plateau Watershed Rehabilitation Project. Early efforts to treat the Loess Plateau includednationally-driventop-downcampaigns to terrace slopes, plant trees and shrubs, and builddams inthe gullies to intercept sediment runoff. While broadly successful inreducing erosion these interventions were not integrated with efforts to raise agriculturalproductivity and farm incomes. The major issue was the poor linkage between land conservation and income-generating activities. Subsequentlythe Government's development strategy for the Loess Plateau changed inresponseto successful pilots in small watersheds (para 1.40) where it was demonstrated that a comprehensive approach involving beneficiaries could tackle both soil erosion and raise farm incomes. This approachwas implementedwith considerable success on over 700,000 ha o f land inthe Bank's first Loess PlateauWatershed DevelopmentProject (1993-2000) that covered three province^.^^ The secondproject extended the geographic scope inthe original three provinces (Shanxi, Shaanxi and Inner Mongolia) and replicated it ina fourth: Gansu province. OBJECTIVES 2.3 The two objectives o fthe projectwere to helpachieve sustainable development in the Loess Plateau by (a) increasing agriculturalproduction and incomes and (b) improving ecological conditions inthe tributary watersheds o fthe Yellow River.These were to be achieved through the introductiono f more efficient and sustainable use o f land and water resources, and the reduction o f erosion in, and sediment flows from, 12tributaries o fthe Yellow River over an area o f 19,500 square kilometers. O fthe 39 counties inthe project, 19 are officially classified as among the poorest inChina, and it was expected that the project would directly benefit 2 millionpeople. Inaddition it was expected to produce substantial benefits downstream as a result o f reduced sedimentation, and globally through carbon sequestration. There were three main components: crop-land improvement, slopelandprotection and support services and training. Details o f objectives, components and costs are summarized inTable 2. 36. Loess Plateau Watershed Rehabilitation Project. Costing US$248.7 million it was approved in 1993 for a IDA loan ofUS$lSO million (Credit 3540-CHA.) The project closed in2002.  14 Table2: SecondLoess Plateau-Objectives, componentsand costs Objectives Components Costs(US$ million) Help achieve sustainable Planned Actual development inthe Loess croplandimprovement Plateau by: 119.0 114.9 Terraces 84.2 Silt retention structure 17.3 Irrigation including water cisterns 17.5 (a) increasing agricultural Slopelandprotection 83.3 94.0 productionand incomes Affrostation and vegetation cover 38.5 Livestock 5.0 (b) improving ecological Horticulture 36.3 conditions inthe tributary Nursery 3.4 watersheds o fthe Yellow River project supPoti costs 25.6 18.3 Research,training and studytours 5.2 M&Eand survey and design 12.5 Operations and Management 7.9 Equipment, vehicles and buildings 7.9 Contingencyandfees 25.5 13.1 Total Costs 252.4 240.2 DESIGN 2.4 Withinthe Loess Plateau about 1,100 small watersheds were targetedas the basic planning unit for improvement. A typical watershed contained several villages and ranged insize from 1,000 to 3,000 ha. Crop-landwith slopes lessthan 25 degrees coveringabout 88,900 ha around 3,464 villages was to be improved through terracing, sediment control structures and provision o f irrigation. The wide and level terraces created from slopelands and old degradedterraces would enable improved agriculture and easiermanagement. Sediment control structures were designed to control flooding, create usable landand store water for irrigation andvillage water supply. Irrigation works includedthe construction o f about 30,000 small water cisterns to catch run-off (mainly from roads), small surface-water diversion structures, and small-sized irrigation schemes to serve 4,100 ha mostly inInner Mongolia. Most o fthe land leveling and terracing work was to be undertakenthough individual contracts with local operators o f small bulldozing equipment, communities and, occasionally, with individual farmers. Arbor trees and shrubs were providedby specialist forestry teams under contract to communities; orchards and grasseswere plantedby individual farmers. 2.5 Slopeland protectionwas intendedto reduce erosion and water loss and allow farmers to increase their income through diversification o f their production systems. This included 90,000 ha o f afforestation and 91,000 ha o f shrubs on land too steep for trees. Pasture grasses to support livestock development would cover 56,00Oha, and horticulture - mainly h i t and nut-bearing trees -would cover another 72,000 ha. Most o fthis work was to be done through contracts with community groups similar to that for terracing.  15 2.6 Training was to finance human capacity development through research and extension, training and study tours, survey and design work and monitoring and evaluation activities. Inadditionthe project financed vehicles, office upgrades and equipment, and project operation and management costs. 2.7 To increase ownership o f the project's conservation focus, the project demanded various kinds o f obligations such as significant restrictions on fanners' production practices inreturnfor localinvestment(e.g., stopping free grazing or cropping on slopes.) For other activities where fanners provided free labor, the project provided trees and shrubs free o f costs but requiredsignature o f "use and management" contracts to ensure sustainability. Project investmentinorchard trees onterraces, irrigation, grass and livestock, was to be on the understanding that beneficiaries repaidthe principalplusan interest rate ofbetween4 and 6 percent. Inmost cases the counties provided loans to farmers for this purpose. The cost o f inputsthat providedpublic goods, such as dams androads, was to be repaidby the counties. 2.8 Quality at entry i s rated as satisfactory. Giventhe physical coverage o fthe project (para 2.10) the decision not to loadthe components with newinstitutions or reform existing ones was pragmatic, as was excluding support for secondary activities such as improving extension services, rural credit, agro-processing or marketing.There was no independent reviewbythe Bank's Quality Assurance Group. IMPLEMENTATION 2.9 The central Ministry of Water Resources (MWR) hadoverall responsibility for the project that was implementedthrough the Yellow River Basin Commission based in Zhenzhou, Henanprovince and its Upper and Middle Reach Bureaubased inXian, Shaanxi province. Policy and coordinationat the central, provincial, prefectural and county levels was through Project Leading Groups (PLGs). The central PLGwas headedby a vice ministerofthe MWR, and includes representatives from the Ministryo fForestry, one vice governor from each o fthe project's provinces, several directors o f YRCC, MWRs' ForeignAffairs, Planning, Water and soil Conservation, Water Resources, and Finance bureaus and representatives from the our provincial Project Management Offices. The PLG's at the provincial level have the same structure andwere headed by the vice governor incharge ofagriculture assistedbythe localrepresentatives o fthe central ministriesand the Agricultural Bank o f China. Prefectural and county level PLGs had similar structure and responsibility. A central Project Management Office (PMO) was locatedinthe YRCC offices and with the regional governments looked after day-to-day management o f all aspects o fthe project, including international procurement and training. Provincial and county PMOSmanaged implementationat those levels and worked through the village level committees. 2.10 The scale o f the project was extremely challenging with as much as 1,000 km betweenproject watershedsthat includeddetailed terracing andplanting o ftrees, shrubs and grasses over an area o f400,000 ha. The overall management structure hadunits in264 xiangs (township), 37 counties and 12prefectures inShaanxi, Shanxi, Gansuand Inner Mongolia. Work was implementedin3,350 administrative villages.  16 2.11 The project became effective in September 1999but got off to a slow start because it overlapped the last three years o fthe first Loess PlateauProject and several newnational policy initiatives and programs for reclaiming degraded lands became operational (para 2.13). Together these increasing demands stretched local resources. Thus insufficient counterpart fundingwas a major problem at all levels throughout implementation and carried reputational risks for the project's environmental focus due to delayed payment o f suppliers, contractors andfarmers. This shortage was accentuated by unfamiliarity with Bank reimbursementprocedures at the township and county levels, overloaded accountants, significant processing delays withinprovincial Finance Bureaus and unwillingness (drivenby prefectural and county poverty considerations) to recover costs from beneficiary farmers. Several PMOSdiverted or withheldBank funds disbursedfor physical works to cover operating andmanagement costs, raising objections from auditors. Despiteall these problems, physicalprogress was extremelywell managed and met most output targets. 2.12 Disseminationo fthe project concept and obtaining local buy-inwas made more onerous by exogenous factors. Severe droughts inlate 1999,2000 and 2001 made establishment o ftrees and shrubs difficult and requiredmuch replanting leading to higher Over the same period, sharply lower global prices for agricultural commodities reduced farmer's profits from planting grains on the newly constructed terraces lowering interest inthe project - short-term gains from annual crops were an essential stop-gap while farmers waited for the harvest from trees and orchards to mature. Inadditionthe dissemination and adoption o f grazing bans on steeply sloping lands cut income from sheep and goats. To compensate, targets for terrace construction were slightly reduced andmore project funding was allocated to livestock at mid-ternreview. This, inturn, put a greater emphasis on forage crops and creation o f grasslands, and ledto increased tree plantingin gullies instead of less-profitable shrubs. 2.13 A numbero fnewnationalprojects andprograms influencedproject implementation. While almost all reinforcedthe conservation objectives o fthe project, their top-down approach and liberal use ofgrants underminedLoess 11's participatory and self-reliance approach that did not include subsidies (para 2.7.) The major programs includedthe `Returning Slopeland to Forest', the `ReturningSlopeland to Grassland', the `Ecological Reconstruction' and the `Sandstorm Prevention' programs. While supporting activities similar to those inLoess 11, most o fthese interventions focused primarily on resource conservation with little concern for local livelihoods. Evenso, their similarityto Loess I1activities created misunderstanding at the local level and confused project beneficiaries. The new national programs also provided grant funds for activities that Loess I1expected localbeneficiaries to pay for inkindor cash (tree and grass planting on slopes, water cisterns, and small-scale irrigation). 38 37. Thistrendwas laterreversedas wetter thennormalweather inthe last 2 years o fthe project greatly increasingsuccessfulplanting. 38. Under The Sloping Land Conversion Program the annualgrainsubsidy for farmersis 1,500 kghainupper reachesofYangtzeand2,250 kg/ha inuppedmiddlereacheso fYellow River.Cashsubsidy for farmers' health,educationandother needsis FWB 300ha annually for same durationas grain subsidy. Subsidyfor seeds andsaplingsis RMB750ha. Incomeform forests and grassesis tax exempt.Farmersare also given30-  17 2.14 Repayment obligations ofthe Bank-financed investment transferred to farmers were handled by the counties who tried with varying success to balance cost-recovery obligations between private beneficiaries and the public sector. This was strongly influenced bythe political climate that increasingly aimed to lower the burden on farmers by increasing grant-aided programs and eventually abolishing most fees, levies and agricultural taxes. Often the counties were left with unforeseen obligations that further exacerbated their cash-flow and slowed progress. For example, because key dams and warping dams39generated significant downstream benefits, YRCC,provincial and county officials argued that the cost should be borne by central government, not by them. In consequence, at mid-term review, only 15 percent of physical targets for these dams had been achieved. Subsequently, after failure to obtain a central government policy directive, each structure was paid for according to local agreements. Inmost cases the counties ended uppaying for those that were built. 2.15 Adequate and timely supply of highquality tree and shrub seedlings, a vital component of the project, was hindered by continued public ownership of project-financed nurseries that were seen as "cash-cows" by the Forestry Departments and county officials. A s the mid-term review noted: "The less than satisfactory nursery outputs are causedby inadequatesupervision of project nurseries and their outmoded managementandplanting practices. Manyprocedures are conservativeand out of date...Theirplanting regimes are haphazardand in most cases they are rarely integrated with the actual needs of the treatment areas. Managers are often administrators with nofundamental understanding or knowledge of technical issues and most of them have only received minimal training in nursery management.Downstream trainingfor the out-reachfarmer nurseries is also haphazard and, in many cases, non-existent.''And to make matters worse, the lack of independent oversight created a conflict-of-interest for state nurseries thus ensuring that all seedlings regardless of quality were transplanted. As a result, they were less able to survive the drought years and about a thirdrequired replanting. This marginally improved towards the end of the project when the entrepreneurial skills of the private sector were unleashed bythe agricultural reforms of 2002 and some private nurseries were established. Farmers andofficials interviewedby IEGfour years later confirmed that the supply ofhighquality seedlings remains a problem for many o f the same reasons.40 2.16 Project closing was extended for 6 months to mitigate crop failure caused by drought and the loss of a construction season incurredby an outbreak of SARS. Reduced interest payments duringconstruction led to project cost savings of US$12.2 million. The IDA credit and IBRDloanwere fully disbursed. year contractsfor plantingand caringfor trees andgrasslands. Farmersreceivingsubsidiesare also be responsiblefor plantinggrasdtreesontwo or more muofbarrenland.Farmer's subsidypackageshave durationoffive years, but are adjustable.Since2005 all subsidies were lumpedunder a single payment: RMB 2,400ha inthe YellowRiverbasin, RMB 3,450ha inthe YangtzeRiverbasin.Source: Bennett,M and J. Xu. 2005 op cit., 39.Keydams are locatedclose to the outlet ofwatershedsthat haveareas o f3-5 km2.Keydams (earthdams 10-20m inheight)retainsedimentsthus creatingland, and inthe interimfimctionas small flood controland water supplyreservoirs.Warpingdams are smaller dams (3-10mhigh) andtheir sole purposeis to intercept sediment and create flat landingullies. 40. IEGfield interviewsin Gansu, Shaanxi and ShanxiprovincesOctober28-November,2,2006.  18 Monitoring and Evaluation 2.17 Design. Robust monitoring indicators were developed at appraisal and fully implementedusing existingM&Esystems developed underthe first Loess PlateauProject. Progress, input, output and outcome indicators were clearly specified as were targets and milestones. Quality o f designi s rated as high. 2.18 At appraisal itwas statedthat "the project will be monitoredthrough a set of indicators that take into account the four key elements o f a sustainable development strategy: productivity, stability, resilience and equity. Specifically, project staff will monitor impacts on crop yields, household production and incomes, grazing policies and practices, long-term land use rights, and integrated watershed planning." The Project Appraisal Document (PAD) gives a very clear statement o f desired project outcome and outputs and how these performance indicators were to be measured. The primary instrumentswould be surveys ofhousehold incomes, productionwithand without the project and sedimentmonitoring inone watershed per county. The designo fthe detailed relationship betweenproject components, their linkages, the economic benefits generated and how these were to be monitored and quantifiedfor economic appraisal andevaluation i s rated by this assessmentas best practice.41 2.19 Implementation. Outcome and impact monitoring forphysical accomplishments benefitedfrom training and was satisfactory. But reporting o f incomes and economic benefitsis not reliable for several reasons. First, the methodology usedto establish "without project" control areas i s unclear (except inthat they were not random) as it was subordinated to eachprovince and subsequentlydownto county level. Second, the methods usedto select individualhouseholds were subjective andnot random, and dependedalmost entirelyonthe understanding and knowledge o fthe local official monitors. Third, no attempt was made to account for spillover effects from parallel projects on either project samples or their controls. Fourth, there were no strict guidelines to assist baseline household selection or ensure a standard process and conformity inreporting. Fifth, the interpretation o f indicators showed wide variation as didhow the indicators should be calculated. Random checks by the Bank's monitoring specialists in2000 revealed a whole range o fproblems.42However, this was their final inputto the project and it i s not known if theirconcerns were addressed. 41. The PAD overviews indicators inAnnex land details expected outcomes and impacts inAnnex 4: Economic and Financial Analysis. 42. "Unfortunately some counties have selected households from the very top o f each grouping. This means that in many cases the richest families inthe village are included inthe sample, and the sample does not include any poor families. Counties should delete these unrepresentative households from the sample and instead make sure that households typical o f each grouping are chosen. For some counties, this may mean the entire sample should be reevaluated and perhaps redrawn. Onthe Northern route, the mission checked the accuracy o f the household monitoring data in almost all counties. Regrettably, major differences were often apparent betweenthe monitoring data collected and the actual situation inthe households. Common errors included: i)recording data at mid-year and therefore not capturing the entire year's results; ii)including land invested under the project inthe baseline data; iii)omitting large pieces o f land; iv) inaccurately counting household members. Inconsistent treatment o f some line items between counties was also quite common." Supervision Report Aide Memoire. September 2000.  19 2.20 Utilization. Effective M&Esystems developed underthe project for management o f inputsand outputs were fully internalizedandhave been widely adopted inother national programs. The institutionaldevelopment impact inducedby good M&Epractices for monitoring project progress and costs was substantial according to prefecture and local government officials interviewedby IEG.Local documentation o fproject processes is impressive.Transparent andeffective accountability allowed evaluation o fthe effectiveness o fpublic sector institutions, project management and individuals, and improved local implementationcapacity. Local ownership is highand political leaders at various levels have taken personal responsibility for the project, exercising a tight system o f control and self-evaluation. As far as IEGcould determine routine process-related M&E at the county and prefectural level continues. It i s not certain, however, ifthe results are used. 2.21 O fmore concern i s that the "impacts" o fthe project on incomes and economic efficiency may be misleading because o fthe questionable methodology and implementation. Giventhe widely geographic dispersion o fthe project sites a strong case couldbe made for a more experimental approach with a much greater degree o f management by M&E specialists. 2.22 Becauseo fthe questions raised by the implementationand utilization experience o f M&E, its overall quality is ratedmodest. This rating gives greater weight to satisfactory assessment o fthe achievement ofthe income improvement objective. Safeguard Compliance 2.23 The project raised only one safeguard issue concerning small dams. While it was not raised at appraisal it was reviewedduringthe Bank's regular supervision and found to be o fno concern. Relevance 2.24 Relevanceis rated as high. The project was relevantto the CAS objectives and for the agriculture sector inChina. These called for assistanceto promote better utilization o f marginal landthrough sustainable techniques, and to develop new income-generation programs inpoorer provinces. Environment objectives o fthe CAS include promotion o f soil and water conservation that are the major components o fthis project. 2.25 China's NinthFive-Year Plan(1995-2000) placedsustained and stable growth in agriculture andthe rural economy at the top o f its agenda, and its ambitious goals include major increasesingrainproductionand farmers incomes particularly inthe arid and semi- arid zones that account for 30 percent o fthe total arable land. Following the drought and flood events o fthe late 1 9 9 0 the ~ ~ central government initiated six large-scale forest conservation programs to restore forest and vegetative cover inupstream watersheds o fthe Yellow and Yangtze River basins. Two o fthese programs, the 1999 Sloping Land Conversion Program (SLCP) - also knownas the Grain for GreenProgram-andthe 1999  20 EcologicalForest Compensation Program begantrial implementation inearly 2000 with an overall budgeto fmore than US$40 billion.43 2.26 Inthe Loess Plateaugovernment strategy emphasizesthe comprehensive and integrated planningo f small individual watersheds inclose consultation with the beneficiaries because land conservation i s compatible with sustainable and productive agriculture andthey are mutually reinforcing. This approach was piloted on over 700,000 ha o f land inthe highly successful first Loess Plateau Watershed RehabilitationProject that overlapped this project and was completed in2002. The devastating floods on the River Yangtze in2002 also heightenedthe Government's concernto address the causes o f the flooding, deforestation and erosion causedby decades o f unsustainable land-use practices. Consequently the project enabled the promotion o f integrated watershed development on a large scale inChina. 2.27 Project design was and remains relevant becausepartnership betweenthe public sector, local governmental institutions and farmers i s essential for sustainable development o fwatersheds. The Bank's most recent global critique o f watershed management reaches the same concl~sion.~~comprehensive approach adopted correctly promoted integrated The watershed management that includedforestry, soil and water conservation, agriculture, and livestock sectors. Its reliance on the public sector for implementationwas relevant because at the time o f designthe private sector was unwilling to provide financing for these types o f investments,particularly inthe poorer areas targetedbythe project. At evaluation, extensivepublic sector involvement make remains relevant becauseo fthe strong public good element o fthe objectives and the few private rural and agricultural enterprises able to participate onventures o fthis scale. Efficacy 2.28 Efficacy i s rated as substantial and i s discussed below. Agricultural production and incomes increased 2.29 The targeted 1.9 millionpeople benefitedfrom the project, 80 percent o fwhom were directly involved inproject activitie~.~'About 89,600 ha o fnew farmland were created byterracing andthe per capita area increased from 0.11hato 0.14 ha. The area o f farming on unstable sloping lands fell from 451,000 ha to 278,000 ha.46Orchards and wheat replaced corn and the flatter land decreasedunit labor requirements and encouraged greater use o f fertilizer and other inputs. According to project monitoring data agricultural 43. This covered atotal of 174pilot countiesspread over 13 provinces inthe upper and middle reaches ofthe Yangtze and Yellow Rivers. Inthe first four years 15 millionfarmers enteredthe program.Itexpandedto 2,000 counties in25 provinces and by 2004 had converted7.2 million ha ofcroplandand afforested4.9 million ha ofwasteland.Source: Bennett,M.T. and J. Xu. 2005. China's Sloping lands Conversionprogram: InstitutionalInnovation or Businessas Usual?Workshop on "Payment for Environmental Services -Methods and DesigninDeveloping Countries." Beijing. 44. The World Bank. 2007. WatershedManagement Approaches, Policies and Operations: Lessonsfor Scaling-Up. The Water-For-Foodteam, Energy,transport and water Department.April 20, 2007. 45. Populationinthe project area increasedfrom 1.988 millionto 2.138 million between1998-2004. 46. MWR. 2005. EcosystemBenefit Monitoring and Evaluation Report. Page41.  21 productionhas changed from generating a narrow range of food and low-value grain commodities to higher-value products, as the example (Box l), and summary data indicate (Figure 1.)Incrementallythe largest changes were to the other category (livestock, forest products and other related outputs), forestry and orchards. Box 1: ImprovedIncomes-Waijau Watershed, Hoglin Village, Jining County, GansuProvince This village has97 householdsanda populationof 500 and is at an elevationof about2,200 m.The project terracedthe whole areabutthe new roadwas financedbythe county. The village headmanfarms 15 mu(1 ha) and supports a family of 5. When askedto statethe factors that mostexplainedhis improvementsin income heputthe impactofthe new roadas beingmost important.The primaryreasongivenwas that surplusfamily labor couldeasilyandquicklyaccess thejob market inthe nearesttown, formerly a two- hourjourney bybicycleconsideringthe steepuphillhomewardpath.He couldalso take goods to market cuttingout the middleman;similarlyagriculturalinputsbecamemore accessible andcheaper. He also put a highpremiumonthe new primaryschool andthe clinicwith a residentpara-medic. Agricultureproductivity has risenappreciablyover the last 6 years. Wheat yieldgrew from 100kg/muto 350 kg/mu; cornfrom 150kg/muto 400 kg/mu. Potatoeswere less than 250 kg/mu and are now greater then400. Hehas livestock but itis usedfor milk and draughtpower.Familyper capita annual incomes haverisenfrom the rangeRMB400 -500 to RMB 1,200-1,300; halfis from off-farmincome. Source: IEG interview October29,2006. 2.30 Incremental per capita incomes inthe project area were Figure 1: Productivity of landincreased estimated at completion to be 58 percent higher thaninnon-project Value of 3,000 . Output control areas (Figure 2.) Giventhe Yuan M&Eand attributionproblems this (million) 2,500 - n 0 1998 02004 estimate o f incremental incomes i s 2,000 i I I not robust (cf. footnote 42.) The project contributed -the exact 1,500 extent is not certain -to the restructuring o fthe agricultural 1,m sector inthe Loess Plateau, the 500 adjustment to a market-oriented economic environment and created conditions for sustainable soil and Oops Other (e.g. Forestry Orchard Pasture INeStOCk) water conservation. 2*3 Terraces have Source: Loess Platueau I1M&E database reduced labor inputs and, together withroad access to the terraced fields, have allowed for mechanization and better communication to markets.The resultinglabor savings have permittedmany farmers to pursuenewincome-eaming activities, such as off-farmjobs, livestock productionandhigh- value fruit and nuttree production. Increasedincome and the prospect o f a sustainable natural resource base have allowed farmers to invest ina wide rangeo f enterprises and social programs which have benefitedthe communities as a whole.  22 Figure 2: Farmers gross per capita incomesincreased 2.32 Some 80,832 ha o fterraces (91 percent o fthe appraisal target) were constructed andprovide the basis for annual and perennial crops produced on terraces, including food grains, vegetables, fodder crops (e.g., maize and legumes) and fruit trees. Conversion o f slopeland to leveled fields increasedmoisture retention, reduced soil erosion and ledto higher crop yields (50-100 500 - percent higherthan on the previously sloped land) and more diversifiedcropping. Along with the terracing ofslopeland, some 24,000 water cisterns were Source: Loess Platueau I1M&E database constructed (79 percent o f the appraisal target) and are successfully mitigating adverse effects o f frequentperiods o f drought. Most investmentsinirrigation were made inInner Mongolia where the extremely low rainfall means that terracing alone would not sufficiently raise yields to bring about a sustainable agricultural system. 2.33 Farmers established about 57,000 ha o f forage that not only reduced soil erosion o f sloping lands but also provided fodder for pen-fedanimals thus reducing grazing pressure. All project counties have now adopted and enforce grazing bans, andthe provincial authorities have extendedthis practice to most other counties inthe four project provinces -this now covers about 85,300 ha. After the widespread adoption ofgrazing bans, the growing o f forage was successfully practiced on a large scale inmost project areas. Along withthe grazing ban, the project successfully provided investmentinlivestock production to help farmers to cope with the effects o fthe ban, support the transition into sustainable production systems and make full use o f incremental fodder andcrop residues produced by other activities to support over 38,000 head o f incremental stock. The introductiono f new breeds o f sheep and goats suitable for pen-feeding, such as Cashmere and Boer goats and Shandong small-tailed sheep, rapidly replaced local breedsand their improvedproductivity made livestock production a major contributor to household incomes, particularly inthe drier northerncounties. 2.34 About 63,000 ha o f "economic trees" and orchards were plantedunder the project (87 percent o fthe appraisal target o f 72,000 ha). Continuously improving road network and information access has enabled these plantations to generate attractive financial returns on domestic and export market. Over 1,100 fruit stores were built (103 percent o fthe target) andallow farmers to capture additional benefitsfrom selling off-season. Ecological conditionsin the tributary watersheds of the YellowRiver improved 2.35 This project established 109,000 ha offorest trees (121 percent ofthe appraisal target) incontour trenches or inpits on wastelands. Inareas less suitable for trees, some 71,000 ha (78 percent o fthe appraisal target) o f drought-tolerant shrubs have beenplanted.  23 Some 57,000 ha o f slopeland were closed and protected for the purpose o f natural regrowth o fvegetation. An ecological survey o f one catchment where grazing was banned indicates a modest increase inbiodiversity: from 107 to 112 varieties o fplants and from 36 to 38 families.47 The same evaluation report estimated that the total area o f newvegetation produced by the project can absorb 1.48 milliontons o f carbon dioxide each year. 2.36 The 2,100 small sediment control structures built are estimated to divert or capture about 25 milliontons o f sedimentper year, o fwhich some 45 percent i s retained by the dams and the remainder i s retained by the terraces and the various vegetative measures. To date project works retained an estimated total o f 53 milliontons o f sediment. Efficiency 2.37 Project efficiency i s rated as substantial but with some shortcomings. 2.38 Overall ERRSvaried from 18 percent to 21 percent, close to the appraisal estimate o f20.9 percent. The slightly highervalues include the benefits from sediment reduction and carbon sequestration. Although the methodology usedto calculate the ERRis exemplary,this assessmentbelieves that production costs are underestimated andbenefits are overestimated due to the presence o f other donor and a substantial number o f Chinese- financed local developments that fostered increasedproductivity and incomes. Inaddition manyruralroads were constructed by other agencies andthis not only changed market opportunities but also providedoff-farm employment access, andbetter education and health. OUTCOME 2.39 The outcome of the project is rated as satisfactory. Relevance was high, while efficacy and efficiency are rated as substantial. Table 3 summarizes the results by objectives discussedinpreviously. 2.40 IEG's field visits and interviews confirmedthat substantial economic development had occurred at sites visited. However, it was also clear that there had beenmany inputs in these areas that complemented those o f the project. Because o f the difficulties with the implementationo fthe M&E system a robust counterfactual could not be determined.In consequenceattribution o f observed benefits i s difficult and, at best, IEGcan only confirm that the efficacy and efficiency o fthe project are substantial and this leads to a rating o f outcome as satisfactory. Ifa robust counterfactual had been available and had been monitored correctly, the ICR's highly satisfactory outcome rating may have been confirmed. 47. Ecosystem Benefit Monitoring& Evaluation Report.Water and Soil ConservationResearchInstitute. August 2005.  24 Table 3: Factors Determining Project Outcome Objective Relevance Efficacy Efficiency Outcome Helpachieve sustainabledevelopmentin the Loess Plateauby: (a) increasingagriculturalproductionand incomes High Substantial (b) improvingecologicalconditionsinthe tributary watershedsofthe Yellow River High Ehbstantial Overall High Substantial Substantial Satisfactory RISKTODEVELOPMENT OUTCOMES 2.41 Riskto developmentoutcome is ratedas moderate. The mainrisk factors are institutional and environmental. 2.42 Institutionalriskarises from continued local government willingnessto enforce the grazing banand assistwith future operation andmaintenance o fterraces should they be damaged by extreme weather events. There i s a risk that farmers will not remainwilling to forgo income for 4-10 years while their economic trees mature and bear fruit, or remain willing to repay their share o fproject investmentand look after the reclaimed slopes that have sub-marginal returns when opportunities for off-farm employment look increasingly favorable. This risk i s enhancedbecause non-project land conservation activities inthe Loess Plateau area are fully subsidized though centrally-funded programs. 2.43 The SLCP has stated environmental goals o freducingwater and soil erosion and increasing China's forest cover and areaby retiringsteeply sloping and marginal lands from agriculturalproduction. Farmersare compensated with grain and cash subsidies for taking sloping land out o f agriculturalproduction, and providedwith seedlingand management subsidies for planting and cultivating trees and grassland on converted land. The program was designedto convert 14.67 millionhectares o f cropland (including a target o f afforesting 4.4 millionha o f landwith gradients o fmore than25 degrees) in23 provinces to forest by 2010 and has an overall budget o fmore than US$40 billion. In addition, it hada soft goal o f afforesting an equivalent area o f~asteland.~' 2.44 Trial implementationo fthe policy began inearly 2000 ina total o f 174pilot counties spread over 13 provinces inthe upper and middlereaches o f the Yangtze and 48. The annual grainsubsidyfor farmers is 1500kgha inupper reaches o fYangtze and2250kgfhain uppedmiddlereaches ofYellow River. Cashsubsidyfor farmers'health,educationand other needsis RMB 300haannually for same durationas grainsubsidy. Subsidyfor seeds and saplingsis RMB 750ha.Income from forestsandgrasses is tax exempt.Farmersare also given30-year contractsfor plantingandcaringfor trees andgrasslands. Farmers receivingsubsidies are also be responsiblefor plantinggrasshees ontwo or moremuofbarrenland. Farmer's subsidypackageshavedurationoffive years, but are adjustable.Since 2005 all subsidies were lumpedunder a singlepayment:RMB 2,400ha inthe Yellow River basin,RMB 3,450ha inthe Yangtze River basin.Source: Bennett,M.T. and J. Xu. 2005. China's Slopinglands Conversionprogram:InstitutionalInnovationor Businessas Usual?Workshop on "Payment for EnvironmentalServices-MethodsandDesigninDevelopingCountries.Beijing  25 Yellow rivers. Inthe first four years 15 millionfarmers enteredthe program. It expanded to 2,000 counties in25 provinces and by 2004 had converted 7.2 millionha o f croplandand afforested 4.9 millionha o fwasteland. 2.45 There is a modest risk also that government's current focus on conservation and "grain-for-green" philosophy may beje~pardized.~~Earlier government policies towards restoring degraded land were a responseto the growing stockpiles o f grain inthe late 1990s. However, now that water shortagesare reducing grain output inthe norththere i s a riskthat food production concerns may override those for conservation andthusjeopardize the "grain-for-green" philosophy. Ifso, central fundingcould dry up and farmers could abandon marginal land that i s not cost-effective to nurture. 2.46 Environmental risk. Large expanses o ftrees and brushunder drought conditions pose a highfire risk. As far as IEGcould determine this has not beena concern o f local government even though the Bank has raised this withcentral government. 2.47 The increasing water shortages inthe lower Yellow Riverbasinposes a riskthat there will be sustained objection from lower ripariansto increasing water retention and use inthe LoessPlateauareaas aresultofgreatly increaseddeep-rooted vegetation andcheck structures and small dams to retainrunoff. Current water retention due to the project i s relatively small it i s estimated to be 130millioncubic meters a year) so this risk i s probably small; This implicit transfer o fproperty rights (to water) implicit inthe greening o fthe Loess Plateau i s not acknowledged by the Chinese government or taken into account inthe inter-provincial water allocationprocedures (paras 2.107-2.1 13.) BANKPERFORMANCE 2.48 Overall Bank performance is rated as satisfactory. The project preparation processwas short because it built on the knowledge, experience and lessons o fthe first Loess Plateau Project andhadthe same task team leader. The institutional, safeguard and fiduciary aspectswere well understood as was the needto have improved arrangements for monitoring and evaluation. Inretrospect, the familiarity withthe Chinese institutions involved inthe first project may have causedunderestimation o fthe difficulties encountered innewcounties and prefectures. Feedback for the Borrower put a very high emphasis on the Bank's comprehensive approach that successfully marriedconservation practice with income-generating opportunities for farmers. Similarly, the consistent application o frigorous procurement and fiduciary policies, while at first beingseen as onerous, was later fully accepted and mainstreamed inthe more progressive provinces. 2.49 Bank supervision i s rated as satisfactory but with some shortcomings. Supervision o f such a large and diverse project area was hugelychallenging and the Bank team used innovative methods to maximize field presenceand coverage -but only supervised once a year. Problems were jointly identified, and solutions were discussed and agreed between the Bank andthe Borrower andquickly implemented.Continuity oftask management and 49. Hong Yang. 2004.Land ConservationCampaign,IntegratedManagementAnd Local Participation In China. 13" International Soil ConservationOrganizations Conference. Brisbane. July 2004. 50. Loess Plateau 11-M&E Results. 2006. Evaluation o f water and soil conservation benefits.  26 maintaining a core supervision team composed o f a broad mix o f international and local professionals throughout the project's life greatly helpedto develop and maintaina healthy and trustful cooperation betweenthe Bank and the Borrower. This continuity and trust allowed true sharing o f supervision responsibility andproblem solving ina consistent and persistent way. However, the single annual missionwas inadequate giventhe reported problems with counterpart funding and audits at the local level, and the concerns raised about the understanding and implementationo fthe M&Eprogram. There should have been more specialized follow-up on M&E, as acknowledged inthe ICR. This has now been addressedinpartnershipwiththe UnitedKingdom's Department for International Development (DFID) to helpupgrade M&Eand impact assessmentthrough parallel financing. BORROWER PERFORMANCE 2.50 OverallBorrower performance is rated as highly satisfactory.Borrower ownership was high. Preparationwas highly satisfactory. The MWRiYellow River Conservancy Commissionput inplace a strong national preparation team that was formed o f experienced project staff and external experts on agriculture, water resources management, forestry, and institutional and social development andthis enabled the project to be completed within a year. The project preparation work, adopting a fully participatory approach, covered major technical, institutional, economic and environmental aspects of the project design, and the views o f all stakeholders at central, provincial and local level were reflectedinthe process. 2.5 1 Implementationwas highly satisfactory. The CentralPLGwas effective at coordinating the project and ensuringa standardized approach to planning and management among the participating provinces. They were proactive insupporting new initiatives learnedfrom project feedback and demonstrated strong leadership backed up by significant policy and operational support. The extensive promotion and enforcement o f grazing bans and long-term land-use contracts have significantly contributed to the achievement o f project objectives. The only real problem was the difficulty o f getting a continuous flow o f adequate counterpart funding, particularly from the lowest administrative levels who hadto pay for the project. It i s remarkable testament to the good Chinese project management skills that, giventhis problem, the work was completed on schedule and there were no slippages. SUMMARY OF ISSUES LOESS RAISEDBY THE PLATEAU PROJECT 2.52 The project experience raises four issues: The importance of a soundly designedM&E systemis particularly important given the number o f other development initiatives andprojects that affect the project area; More attention should be givento buildingcapacity for M&E at all levels that moves beyondthe simple measurement o fphysical impacts and include socio- economic impacts, development o frealistic counterfactuals and methods to deal with the problem ofattribution o fcosts andbenefits. 0 Resolving the differing policies applied to financing and management o fwatershed conservation. The government's own watershed conservation projects remainpublic goods oriented and farmer's contributions are fully subsidized. This suggests that  27 the Bank's innovations (through several Bank-financedwatershed conservation projects) inwhich beneficiaries are expectedto contribute to private goods generated by watershed conservation project remain enclave activities andthat the policies the Bank is promoting have not beenmainstreamed inChina. 0 More attention needs to be givento the regional impact o f revegetating the Loess Plateauto minimize consumptive use o f scarce water resources. XIAOLANGDI MULTIPURPOSEPROJECTSI& I1 OBJECTIVES 2.53 Eachproject hadthe same objectives. There were five aimed at: 0 providing flood protection for major infrastructure and 103 million people mainly in the rural areas o f the North China Plain; 0 controlling sediment accretion inthe 800 km lower reach o fthe Yellow River for a period o f about 20 years; 0 supplyingand stabilizing water to several large andmediumcities andmajor industrial centers; 0 providingreliable irrigation for some 2 million ha; and 0 generating much neededhydropower to meet the peak power requirementsinthe Henan Grid. 2.54 Although the above objectives remained unchanged betweenPhase Iand I1o fthe project, urbanwater supply took on an increasing importance at the project progressed. Booming economic growth has translated into ever rising demand for municipal and industrial water supplies, and this trendi s likely to continue. The Xiaolangdi reservoir i s keyto maintaining urbanwater suppliesinnorth China duringperiods o f drought, as eloquently illustratedin2000-01. This key role was not anticipated at the time of project designand appraisal. 2.55 Overall project costs andphasing are summarized inTable 4. Costs classified as "other" were export credits from the American Import and Export Bank (US$55.8 million) and from the Tokyo MitsubishiBank (US$18.2 milli~n.)~' Total costs are related to objectives and components inTable 5. 5 1. The costs presented inthe ICR differ from those inthe Bank's OperationsPortal and data from the latter are presentedhere.  28 Table 4: Xiaolangdi-Overall projectcosts and phasing 1994 1996 1998 2000 2002 2004 Total Costs, US$ millions DESIGN 2.56 The physical works consisted o fthe construction o fa very large dam and power plant onthe Yellow Riverinthe transition zone betweenthe middle and lower reaches, about 8OOkm upstream from the delta. The dam was designedspecifically to create a reservoir capable o f storing bothwater as well as 0.5 billioncubic meters o f sedimentthat was expected to accumulate init over 20 years. 52 This would not interfere withwater storage, for which there would be 5.1 billion cubic meters o f capacity available. 2.57 Manytechnical and engineering studieswere done inthe decade-long period preceding construction, (including some by international experts). Geological investigations o fthe site go back as far as 1958. The Bank also provided some assistance in project preparation, via the use o fthe IDA Second TA Credit. This evaluation shares the view o fthe Bank's Quality Assurance Group that the project's quality at entry was highly satisfactory. The early completion o fthe project i s an eloquent testimony to the thoroughness and highquality o f all the preparatory work undertakenprior to construction. 52. 154mhigh, 1.7 km in lengthand with anundergroundpowerhouseequippedwith 6 X 300MW generators. Ithas createda reservoir about 140km in length, with a maximum storage capacity of 12.6 billion cubic meters.  29 Table 5: Xiaolangdi I& I1-Objectives,Components and Total Objectives Components Costs (US$millions) Planned Actual Dam Construction.Phase I: Build Xiaolangdi rockfill dam, 154 m high, with a crest length o f 1,370 m; Phase11: 1. Introduceflood extendedthe crest lengthto 1,667 m. control inthe lower Construct a common intake structure, 1,520.3 1,760.9 reacheso f the Yellow feeding nine large diameter tunnels and RiverBasin to protect surface spillway for river diversion, major infrastructure and 103 million people flood handling and sediment management, discharging into a 2. Control siltation inthe common plunge pool. 800-km downstream Facility Management. Phase I: channel o f the river and training and technical assistancefor prevent further YRWHC incontract management, aggradation so that levee project scheduling, cost control, claims heightsneednot be management, financial management, 235.5 201.4 raised further during a managementinformation systems, period o f 20 years organizational support, corporate 3. Provide water for planning and personnel management. assuredirrigation for 2 Phase11: technical assistancefor million hectares training in operation and maintenance 4. Provide more stable (O&M) ofthe dam, powerhouse and 5.1 4.6 water suppliesfor on related facilities and planning and stream cities and personnel training for operation o f the industries project. EnvironmentalManagement. To monitor, manage and offset any 10.1 13.8 negative impacts o f the project; and in PhaseI1to implement o fthe 5. Generate hydropower environmental managementplan for supplementing the Power Development. Construct a power base load o fthermal station with six associatedpower 462.7 256.0 stations inHenan tunnelswith turbine and generators Province and the Central with an installed capacity o f 1,800 M W China Power Grid (6 x 300 MW) and associated switchyards, and transformers. River Basin Management. An 1 institutional program for M W R and YRCC to support the reform process in the water sector, Le., inadapting to 14.8 14.7 Institutional reform and international accounting procedures, capacity-building that sustainableresource mobilization for applies to all 5 objectives water resourcesprojects, water pricing, water licensing, effective river basin organization, and water dispatching systems for basins. Interestduring construction - 607.4 437.4 2,855.8 2,688.8 2.58 The first Bank loanfundeda substantial amount o fcapacity-building and training for the staff of the Commission and the newly established (1991)River Water &  30 Hydropower Development Corporation's (the Corporation) both locally and abroad and providedtechnical assistanceto the MWR as well. The focus o fthe second Bank loanwas on transformingthe Corporation into an operating rather than construction management entity, while also providing considerable institutional support to the Commission inwater allocation and dispatch, reservoir operations and flood and sedimentforecasting and management information systems. The current irrigation water dispatch system for the lower reach was set upwith Bank assistance under the project 2.59 The emphasis on developing facility management and training for the staff o fthe neworganization set upto buildand operate the dam was appropriate. Training for the Yellow River Conservancy Commission (the Commission) was to enable them to undertake institutional reforms such as inthe areas o fwater pricing and facilitate coordinationand interactive management systems with the Corporationon the day-to-day regulationo fwater resourcesinthe river basinand flood forecasting. IMPLEMENTATION 2.60 By any international civil engineeringstandards, the Xiaolangdi project is amajor accomplishment interms o f its scale and complexity. The technical challenges and difficulties inimplementationwere numerous, but as these have beendescribed satisfactorily inthe ICRs, they will not be repeated here. Nevertheless, the fact that the physical works were completed slightly below cost and a year ahead o f schedule should be noted, as these are exceptional occurrences inthe case o fhydroelectric projects, regardless o f where they are built. The first power generation unit went into service inlate 1999. Downstream sediment flushing from the river bedhas beencarried out successfully since 2001 and continuous water flows to the river mouth have beenmaintained since then. Monitoring and Evaluation 2.61 Design. The designo fM&EinPhase Iwas appropriate even though it paid little attention to outcome indicators. This was because the focus was on the engineering construction that was managed by the Corporation whilst the downstream impacts were routinely monitored by either the Commission or the statistical departments o f Shandong andHenanprovinces. Thus PhaseIfocused onphysical process indicators andreporting o f physical progress and financial expendituresduring construction. However, it did incorporate extensive arrangements for dam safety monitoring and flood forecasting inline with international best practice. Thiswas independentlyoverseen by anindependentDam Safety Panelthat included leading Chinese and World experts. 2.62 The PhaseI1designo fM&Ewas a markedimprovement and clearly specified key performance, outcome indicators and targets. These were:  31 Eliminate annualflood damages ofUS$91millionayear Increase farm incomesby about 15 percentinHenanand increase labor productivity inthe farms by 20 percent.Increase grainyields by 20 percentfrom 4 tonsha to 4.8 tonsha and increase productionby 0.515 milliontons ayear. Reduce industrial downtime due to lack ofwater to 2 days per year Between200 and 2025 riverbedceases to rise between2001 and2025 andYellow River flood dykes do not haveto be raised. Eliminate catastrophicflood damage andreduce flood damagewithin the dykes for a 2-year occurrence to a 7-year occurrence. Generate 2,400 GWh ofpeak power 2001-03 increasingthis to 4,700 -5,400 GWhfrom 2004. Apart from the monitoringo fpower generation, all the other indicatorswere expectedto be part o f routine monitoring carried out by the Commission andprovincial agencies. This was true o fthe indicators related to water levels, sediment and river morphology butno arrangements were establishedto determine ifthe routine monitoring by other agencies was satisfactory or coveredthe expectedoutcomes and impacts. Although the expected outcomes were the result o fthe river basinmodeling exercise (para 1.33), this was a deterministic mathematical model andrepresentedonly a simplifiedform o f economic reality. No allowances were made for the many other factors - including changesto economic policy or the political economy -that would affect impacts on agriculturalincomes and productionor security and availability o f downstream water supplies.For example, industrial downtime due to lack o fwater was also a function o f many exogenous factors such as intra-basin or municipalwater allocation withinthe provinces, the water conveyance system below the dam and increasing industrial water demand. Inthat sense there was no credible real time counterfactual against which to measureincremental economic impact. 2.64 While the improvements inflood conditions and sedimentation could be measured downstream, this would be the result o fthe cumulative impact o fXiaolangdi and changes upstream. As far as this evaluation can determine, no mechanism was designedto monitor the differential impacts ofthe dam and its operation. Overall quality o fM&Edesigni s thus rated as modest. 2.65 Implementation.Dataon water inflows and outflows from the reservoirwere collected and this continues. This i s a sine qua non for decisions on reservoir management, water allocations and dispatch. Indicators o fpower operations are also recorded continuously. The Commission has an extensive network o fhydrological stations on the Yellow River andhas accumulated data over many years on water flows. These data provide the essential inputs for its analytical work. Datarelating specifically to Xiaolangdi are therefore just one necessary element inthe Commission's requirements for carrying out its mandate o fwater resource management inthe Yellow River Basin. 2.66 The Commissionmonitors changes inthe deptho f the main channel o fthe river in the lower reachinorder to assess the effectiveness ofthe sedimentflushing operations. However, this evaluation was unable to ascertain ifthe Commissionmonitors ecological changes inthe river or delta resultingfrom the restoration o f uninterruptedyear-round water flows.  32 2.67 Utilization. Monitoring data have beenfully internalized andhave beenusedto operatethe dam effectively according to its operating rules and the requirementsfor sediment flushing, flood control and power generation. The efficacy and efficiency o fdam operations i s discussedbelow (para 2.76 et seq). 2.68 Taking account of design, implementation and utilization, the overall quality of M&Eis rated as modest. The reasonisthat the claims for the project's outcomes and impacts on agriculturalproductionare basedupon modeled projections, not the results o f field-level M&E. This i s not to say it could not be done. Giventhe large capital investment less than 0.1 percent o fproject costs would have allowed a carefully designed and independentlymanagedimpact evaluation usingan experimental (randomized) design.53 Safeguard compliance 2.69 The project raised three safeguard issues: dam safety, involuntary resettlement and cultural heritage. The dam lies inan earthquake-prone area and could be subject to an earthquake o f magnitude 6.25 on the Richter Scale. Project design fully internalizedthis risk inthe designofthe dam andits embankments. A network of seismometers hasbeen operational since 1995. A Dam Safety Panel has independent oversight o fthe facility. Satisfactory arrangements are inplace for dam inspections andemergency preparedness. The project was less successful inimposing dam safety requirementson the Sanmenxia dam that lies upstream (theflood wave from the collapse or overtopping o f this dam would threaten the safety of Xiaolangdi.) At completionthis dam hadproblems with inadequate liftingcapacity ofthe gantry cranes at the dam crest andthis would significantly reduce its ability to control pool levels ina flood emergency. Additionally there i s some risk that the dam would be unable to maintain pool levels above 335 m. 2.70 Resettlemento f 184,040 people was subject to a stand-alone resettlement project that metthe requirementso fthe Bank's policies on involuntary resettlement (OP4.12.) EG's recent evaluation o fthis rated it as satisfactory butwithreservations on the cessation o f M&Eactivities on credit closure. Without continued M&E it i s impossible to monitor the longer term impacts ofenforcedresettlementonthe welfare and incomes ofaffected people and host communities. Lack o fknowledge precludes establishment o f measuresto mitigate hardship and accelerate a return to normality. 2.71 Cultural heritage was an issue becausethe dam was built in"the cradle o f Chinese civilization." Extensive archeological investigations, classificationand salvage programs were implementedand completed satisfactorily. Relevance 2.72 Relevance is rated as high. This i s basedon strategic relevance to China's needs, to the Bank's strategy for China (para 1.21 et seq.,) and relevance o fthe project design. 53. Such an exercise was extremely rare within the Bank at the time o f appraisal and it is only applied to a small proportion o f social sector projects today.  33 2.73 The choice o f a multipurpose dam was highly relevant as it allowed flexibility o f choice betweenvarious operational modes using an increasingly scarce resource -water. The choice o fhydropower was appropriate giventhe needfor alternative sources o f clean energy. 2.74 Flood and sedimentcontrol are closely linkedand clearly deserve top priority due to the risk o fdestruction o f assets and loss o flife andthe heavy cost that this would impose on China. Meetingthe water requirements o f urbanpopulations i s rightly considered the nexthighestpriority. Water requirementsfor irrigation andpower are not necessarily exclusive, but water releases from the Xiaolangdi reservoir would differ intime and scale if power generation was to be givenpriority over irrigation. The priority givento agriculture i s entirelyjustified giventhe large number o frural beneficiaries from irrigation andthe fact that substitutesfrom other sources can be found for electricity supply. 2.75 The adoption o f a two-phase approach primarily to reduce engineeringrisks and enable lessons from the first phase to be incorporated into the second phase was highly relevant. Efficacy Flood protectionwas provided 2.76 Preventing damage to life and property i s the priority objective o fthe Xiaolangdi project. The Xiaolangdi dam andreservoir together now provide the populationliving downstream with protection against all floods with a one inthousand year probability. Prior to the project, the existing infrastructure o f dams and dykes only offered protection against floods with a one insixty year probability. By means o f careful water flow regulation, Xiaolangdi has also successfully preventedthe formation o f `ice dams' that usedto cause flooding inwinter when downstream sections o fthe river froze. 2.77 The efficacy o fthe dam was tested in2003, the heaviest wet seasoninNorthern China since 1981. The reservoir was able to contain the inflows and store 6.3 billion cubic meters o f water and maximum downstream discharge was limitedto 2,800 cubic meters per second. Without it, flows downstream at Huayankouwould have reached 6,000 cubic meters per second with devastating consequences and danger to 18 millionpeople. The Corporationcites a figure o f Y 11billion (US$ 1.3 million) as the avoided cost o f flood damage that would otherwise have occurred. IEGcould not find out what this figure includes or how it was calculated. 2.78 On the basis o f experience to date this objective was fully achieved andits efficacy i s rated "high." Sediment accretion of the Yellow Riverwas controlled 2.79 Insome respects,the operational phaseo fthe Xiaolangdi project hasprovedto be more complex than the actual construction o fthe project. The project designand hydrological studies were done inthe 1980s (ifnot earlier) and nearly twenty years later, the actual hydrology inthe immediate post-completionperiodhasturnedout to bequite  34 differentto what was expected.54The effect o f a series o fthree dry years (2000-02) immediately coinciding with the completion o fthe dam ledto different reservoir operating rules havingto be devised and adopted. The low water inflows also meant that the annual inflow o f sediment into the reservoir was far less than expected: 0.4 billiontons on average, as against the predicted 1.3 billiontons. 55 Consequently the reservoir's capacity for flood control inthe initial years was higherthanpredicted. This inturnallowed the lowest reservoir level prior to the rainy seasonto be 5 m higherthanthe design, which inturn meantthat more water was available for supply to urbanareas affected by drought. Reservoir modeling will continue to be o f great importance becausethe past few years have shownthat the rate o fdecline o f its live storage capacity has beenquite differentto the initial planning parameters. 2.80 Itwas estimatedat appraisal that the reservoir would trap 7.5 billioncubic meters over its first 10 years o f operation. This target has not beenreachedprimarily because o f the reduced inflow o f sediment to the reservoir. The true measure o fthe efficacy o fthe sediment management objective is that downstream flood embankments(dykes) have not hadto be raised thus saving an estimated US$75 million a year.56 2.81 Sediment FlushingOperations.Managingthe highsedimentload o fthe Yellow Riverpresents the Commissionwith complex andnearly unique challenges. The experience o fthe very rapid silting up o fthe Sanmenxia reservoir inthe early 1960s provides valuable lessons for the design o fXia~langdi.~'The needto trap some coarser sediment while allowing finer silt to passthrough the reservoir was explicitly incorporated inthe design, which envisaged aprogressive filling upofthe lower levels ofthe reservoir, while always preservinga substantial amount o flive storage for flood control andwater supply* 2.82 Eachyear since 2002 inJune/July, prior to the onset o fthe rainy season, the Corporationhas carried out a controlled `flood' inorder to flush out sediment from the reservoir as well as to transport out to sea sedimentpreviously deposited further downstream. A discharge o f at least 2,600 cubic metedsecond (m3/sec) i s drawn from the lower levels o fthe reservoir and the rip current created erodes deposited silt and transports it downstream. Each 1,000 cubic meters released removes about 1.5 tons o f sediment and this is continued for periods ofabout 7-10 days loweringthe lake levelby about 10meters 54. The long-term hydrological series (1919-74) that was the basis for the planning and design studies may still prove to be valid since 2000-06 i s a relatively short period. 55. As against predicted inflows o f 25-28 billion m3, the actual inflows inthe three years 2000-02 were about 13 -15 billion m3. Since the dam was closed total sediment inflows are estimated to be 2.6 billion tons, o f which 0.5 billiontons have since been flushed out 56. Historically, between 300 and 400 million tons o f sediment have been deposited on the riverbed along the lower reaches every year, resulting in a rise inthe riverbed o f 8-10 centimeters annually. Thus it was necessaryevery decade to raise 696 km o fthe main dikes at least 1 meter, to strengthen 969 km o f levees, to raise or strengthen 5,395 spurs invulnerable sites, and to reconstruct or complete 51 channel training works and 439 spurs totaling 40.8 km.The total present value o f one o fthese decade-long comprehensive programs was estimated, at appraisal to be at least Y 4 billion ($755 million). These costs were expected to increase over time as the base o f the dike was widened and as the integrity o f dike foundations was threatened by seepage and salinity as the riverbed rises. 57. Upstream o f Xiaolangdi inthe middle reach o fthe Yellow River. Despite redesign, modifications and silt flushing, it now has a live storage capacity o f under 2 billion m3.  35 and thus creating more flood storage. Whenmixed with fresh water drawn from higher in the reservoir the highenergy o fthe combineddischarge and its large velocity enable the riverto transport a higher sediment loadthanit carries and itthus also scours the river bed downstream. 2.83 The flushing operations have beenremarkably successful. The flow capacity o fthe main channel o fthe river has beenraised froml,800 m3/secto about 4,000 m3/secas a result o fthe deepening o f the channel by about 1 m on average. Total silt transported out to seahas beenestimated at about 260 milliontons as a result o fthe flushing operations. The Commission's objective i s to restore the river's flow capacity to that o fthe 1970s which was about 5,000 m3/sec.Outside the rainy season, the average discharge rate downstream o fXiaolangdi i s maintained at 800 m3/secwith a minimumo f 200 m3/secduring droughts. Ifdryseasondischarge exceeds800m3/secthere isariskthat sedimentwillbecarried downstream from Henanprovince to Shandong province but not be flushedto the sea. If this didoccur itwould riskincreased floods inthose aggradingreaches. 2.84 Research i s beingundertakenby the Commissionto improve sedimentmanagement intheXiaolangdi reservoir andto achieve the downstream flushingeffects withlower water di~charges.~~ The sediment discharge ratio (the percent o fincoming sedimentflushed out) from the reservoir at 18 percent i s already better than the designtarget o f 14percent andattempts are beingmade to raise this further inorder to prolong the life o fthe reservoir. Tests are beingundertaken with water releases carrying sediment loads one-third higher than the highest sustained to date. The combination o fthe lower sediment inflows in2000- 02 and the successful flushing operations means that an additional five years have already beengained before the reservoir's sedimentstorage capacity of 7.5 billionm3is entirely filled. 2.85 On the basis o f the experience to date the sedimentmanagement objective has been fully achieved andefficacy is rated as "high." Water supply to downstream citiesand industrieswas delivered 2.86 Water shortages inthe lower reaches o fthe Yellow River prior to the construction o fthe Xiaolangdi project were chronic. Inthe two decadesprior to the project, it was not unusual for the riverto dry up completely for weeks at a stretch (para 1.15.) Yet despite drought conditions in2000-02, the water stored inthe reservoir andcontrolled releases throughout the year have meant that, since 2000, water flow inthe lower reaches o fthe Yellow River all the way to the sea, and has never beeninterrupted.The ecological benefits have beenconsiderable: freshwater wetland areas have reappeared, as have fish species that have not beenpresent since the 1980s. The availability o f silt-free water downstream o f Xiaolangdi i s also beneficial as it reduces the cost o furbanwater treatment and sedimentation o f irrigation canals divertingriver water. 2.87 Supplyto major cities such as Tianjinwere maintainedduringone o fthe worst droughts o fthe past century -a 1in200 year event -despitethe reduced water inflows to 58. The reservoir had an initial live water storage capacity of 12.6 billion m3, which was expected to gradually fall to 5.1 billion m3 over 20 years. This level is to be preserved indefinitely thereafter.  36 the Xiaolangdi reservoir. Indeed,the reservoir provedto bethe only dependable surface water supply for downstream towns and cities. Meetingmunicipal and industrial demand for water indry years was not givenmajor prominence at project appraisal, but i s likely to play a bigger role interms o f water allocation issuesinthe coming years. 2.88 This achievement is partly the result o fexogenous factors andnot attributable solely to the project. Itwas also the result offorce majeure by the lower riparianprovinces at the highestpolitical level. Without this interventionXiaolangdi would have been unable to supply the demand because o fthe water allocation system. This was the result o fthe Yellow RiverWater RegulationManagement Regulation Stipulation (1998)that was a responseto the 1997 drying up o fthe river. Underthis newmandate the Commission was authorized to conduct unifiedwater resources regulation inthe basinandreassign quotas to eachprovince or autonomous region. Where possible this i s achieved automatically by the Commissionvia telemetric control from Zhenzhou for 77 important water abstraction sites and key reservoirs. Quotas are proportionally increased or reduced based on the predicted total natural runoffmodeledby Commission. As a result o fthese measures annual water diversions from the Yellow River were maintainedbelow 30 billion cubic meters inthe period 1998-2001.59 2.89 Controlling the water diversion from the Yellow River was, however, only a partial solution because demand continued to exceed supply. Thus emergency closure o fupstream irrigation off-takes was requiredin 1999 to provide essential minimumflow at Sunkou and Lijinmeasuring stations downstream.60In2000 anemergency releaseofone billioncubic meters ofwater stored for hydropower generation at Longyanxia dam was neededto alleviate that year's water crisis inTianjin City on the east coast.61And in2002 a special high-level delegation from Shandongprovince secured an additional 0.8 billion cubic meters from the State Council without any compensation for upstream users inNingxia province and Inner Mongolia autonomous region.62These adhoc adjustments took no account o fthe opportunity costs o fwater so these emergency allocations are not generally the most economically efficient solution. Clearly, effective and efficient institutions are at least as important as engineeringinriver basinmanagement. Irrigationwater was suppliedwith increasedreliability -but probably not for the whole of the targeted area of2 million ha 2.90 To fully supply irrigable agricultural land inthe two provinces o fHenan and Shandong requires the annual diversiono f about 23 billionm3from the river.Prior to Xiaolangdi total water availability for all end uses was only 32 billionm3with the result that water for irrigation was always below the biological optimum to maximize crop yields. 59. Fu, Guobin, and ShulinChen. 2006. Water Crisis inthe Yellow River:Facts, Reasons,Impacts,and Countermeasures. WaterPractice& TechnologyVol 1No 2. IWAPublishing2006. 60. Zhu, Quingping.2006. PreliminaryAssessmenton the Impactso f the UnifiedWaterRegulationsinthe Yellow River.MOWR. 61. Embassyofthe USA, Beijing.Managingthe UpperReachesofthe YellowRiver.April. 2003. 62. Wang, Yuhua.2003. Water Disputeinthe Yellow RiverBasin: Challengesto a CentralizedSystem. Commentary.Woodrow Wilson Center.Washington.D.C. August 2003.  37 Evenafter construction o fthe projectthere is insufficientwater to fully irrigate the total croplandinthe lower reaches, giventhe needto supply cities and industries. 2.91 Despite the construction o fthe Xiaolangdi project and the additional water made available by it-4.7 billion m3annually on average over the six years 200-2006 -it i s estimated that there i s still an annual average shortfall o f 3 billionm3o fwater just inthe Yellow River basin.63All indications are that this cannot be covered and the shortfall may even worsen despitemeasuresto increase supply anduse water more efficiently. 2.92 Thus this objective was partially achieved and its efficacy is ratedas "modest" at best. Lack o f evidence causedby inadequate M&Ewas a major constraint injudging achievement. Hydropower generation exceeded expectations 2.93 Itis one o fthe ironies ofthis multi-purpose projectthat the electricity production from it rates last inthe order o fpriorities, while beingthe sole source o f revenues to pay off the construction debts and cover the runningcosts o fthe entire operations o f Corporation. At the same time the hydroelectric plantplays a vital role inthe Henanprovincialpower supply system. 2.94 Xiaolangdi (1SOOMW) i s the largest generation plant inHenanand accounts for almost three-quarters o f all hydro inthe province. Although its capacity exceeds the next largest (thermal) plantsby a third and it has greater operational flexibility, itnow accounts for less than 6 percent o f the total provincial generation capacity. Xiaolangdi therefore plays a valuable role inmeetingpeak electricity demand and in2005 its capacity represented over a tentho f the total demand o f 17,600 MW on the Electric Power o fHenan (EPH) system. Its size also makes it a key plant for EPHinfrequency regulationand maintaining grid stability. The fast responsetime o fhydro generation units gives it an important role inensuringuninterruptedsupply inthe event o f unplanned shutdowns by thermal plants. Annual generation planstake into account the water supply needs o f urban and agricultural users and arejointly agreed between EPH, the provincial government and the Corporation. EPHisthe Corporation's only client. 2.95 The first 300 MW unitwent into service inlate 1999 andthe final one was commissioned two years later. Although total electricity generation by Xiaolangdi inthe first three years was substantially below expectations, largely due to drought-induced low flow, it has exceeded 5,000 GWH inboth 2004 and 2005 and i s expected to continue at about this level inthe next few years (Table 6). This exceeds the targeted energy production o f4,773 GWhinthe period 2002-08. At the current average sales price o f 63. One billion m3could irrigate about 200,000 ha if it were all applied to a crop likewheat. However as about 50% or more will be lost intransit from the river due to seepage from canals actual coverage is more likelyto be 100,000 ha. Thus the total incremental supply from Xiaolangdi could potentially provide about 0.5 million ha o f surface water irrigation. Inaddition the water lost en route in seepage could be recovered from groundwater by tubewells perhaps adding up to 0.2 million ha.  38 about Y 0.32/kWh, this production gives the Corporation gross annual revenues of about Y1.6 billion(US$ 205 million).64 2.96 Since the project closed the Corporationhas had approvalto construct a small re- regulating dam, Xixiayuan, 16km downstreamo fXiaolangdi. This will provide daily water storage to iron out spikes indownstream water releases resultingfrom surges inpower generation. It will thereby give the EPHdispatch center greater operational flexibility and reduce the scope for conflicts betweenthe use o f water for power generation andnon- poweruses. The associated 140MW powerplant will also contribute modestlyto electricity productionto the tune o f 500 GWWyear. The sale price o f electricity has yet to befixed. Onthe assumption that the tariffwillbe at least as muchas that as for Xiaolangdi, the Corporationwill earn anadditional Y160millionannually (over US$20 million) from this investment. This Y2.2 billionproject i s again beingfundedpredominantly by the central government on a grant basis. The Corporationwill only be responsible for the debt service on a commercial loan of Y500 million, which funds the balance o fthe cost. 2.97 Despitethe shortages of electricity inChina duringthe past two years, the Corporation has not beenable to obtain a significant increase inits powertariff. The price it is authorized to charge is the outcome ofaprocess involving the buyer (EPH) the Henan provincial governmentpricing bureau, the NDRC65andits provincial arm, PDRC. Although the NDRC approved a tariff o f 34.5 fedkwhwith effect from 1/1/2006, the Corporationhasnotbeen permittedto levy this ratedue to provincialobjections to a pricehike. 2.98 ElectricityPricing.Onthe other hand, since mid-2004, the Corporationhas been allowed to charge differentprices for electricity according to the time o f day. Ithas a peak tariff o f 1.33 times the base rate of 31.7 fen/kWh and an off-peak rate o f 0.55 times the base tariff during the night hours o f 2300-0700. 2.99 The current level o ftariffs is substantially less thanwas expected by the Bank at the time ofproject appraisal. Evenafter deductingVAT, the Bank's projections contained in the PAD for the secondloanassume apowertariff almosttwice that actually levied. The Bankalso expected atwo-parttariffto beintroduced, with separate capacity and energy charges. This has not occurred.66Even so, the Corporationhas earned an annual return o f about 8.6 percent on fixed assets andthis exceeds the Bank's stipulation o f a 5 percent after 2007. 2.100 At the time of the first appraisal, the Bank's analysis o f Xiaolangdi's power tariffs was baseduponthe cost ofthermal equivalence: that o f a coal-fired plant operating only four hoursper day to meet peak load. With the prevailingcapital costs and coal price o f 64. Including VAT, that is levied at 17%. 65. Being under the MWR, the Corporation is subject to NDRC's authority on power pricing. 66. The Loan Agreement stipulates in Section 4.07 "Immediately upon commissioning o fthe Xiaolangdi Dam, charge and collect from all said users, water tariffs sufficient to recover: (i)the costs o f operation and maintenance o f said dam in full as they are incurred; and (ii) the capital costs o f said dam in full not later than December31,2014."  39 1993, the resultingfigure for the value o fthe firm energy to be producedby Xiaolangdi was estimated to be 52 fen/kWh. Although conceptually valid, this estimationwas never acceptedby EPHor the provincial government as the basis for pricingthe power from Xiaolangdi. The initial power sales agreement betweenthem [that the Bank requiredas a condition o fproceeding withthe second loan] merely stated that the price would be set at a level sufficient to enable the Corporationto fully recover all costs and earn a fair return on the investedcapital. Table 6: Xiaolangdi Electricity Production, Prices & Sales since 2000 2000 2001 2002 2003 2004 2005 2006* Estimated OutDut fGWh) Appraisal Phase I 2,507 3,760 3,947 4,875 4,875 4,875 4,875 Appraisal Phase I1 2,275 3,679 4,231 5,115 5,115 5,115 5,115 Actualproduction 642 2,111 3,286 3,684 5,015 5,048 5,300 Actual/ Planned(Phase11) 28% 57% 78% 72% 98% 99% 104% Planned Drice (fen/kWh) Appraisal Phase I 27 30 35 37 38 49 51 Appraisal Phase I1 52 52 52 52 52 52 52 Actual Average revenue Approved kWh (excl. VAT) 24 23 24 21 27 27 27 Actual/ Planned(Phase 11) 46% 44% 46% 40% 52% 52% 52% Total Revenues (mn Y, excl. VAT) 150 483 778 770 1,370 1,367 1,436 Source: Yellow River Water & HydropowerDevelopmentCorporation 2006 * =estimated Efficiency 2.101 Overall efficiency o fthe project i s assessed as substantial. The EIRR o f 13 percent as presentedinthe ICR i s probably underestimated since it undervalues the economic benefits from power and urbanwater supply and does not include any benefits at all from the improvedecological conditions downstream and inthe delta. OUTCOME 2.102 The outcome o fthis two-phase project i s rated as highly satisfactory for both phases. For both projects relevance and efficacy are ratedas highand efficiency i s rated as substantial. Table 7 summarizes the results by objective and outcome. 2.103 The Corporationi s the first instance inChina o f a financially autonomous water resource entity, runon `semi-commercial' lines (para 2.115) that i s responsible for covering its full operating costs and repaying loans contracted for construction. At the same time, it is underthe control of Commissionregarding decisions relatingto flood control, urbanwater supplyand reservoir management policies. Equally for power generation, it is underthe operational control ofEPHfor real-time decisions on generation levels.  40 Table 7: FactorsDeterminingProject Outcome Objective Relevance Efficacy Efficiency Outcome 1. Introduceflood control inthe lower reaches of the Yellow RiverBasinto protectmajor High High infrastructureand 103 million people 2. Controlsiltation inthe 800-km downstream channelofthe river and preventfurther High High aggradation so that leveeheightsneednotbe raisedfurther during a period of20 years 3. Providewater for assured irrigationfor 2 High Substantial million hectares 4. Providemore stablewater supplies for on High High stream cities and industries 5. Generatehydropower for supplementing the base loadofthermal stations inHenanProvince High High andthe CentralChinaPower Grid Overall High High Substantial Satisfactory 2.104 Despite the commercialnature o f its operations, the Corporationhas very little autonomy since it has no control over the key parameters o f its operations - decisions on water storage and releases as well as the price andquantity o fthe electricity it generates are all made by other bodies. This i s understandable since the expertise on water resource management andthe analytical skills for optimizing the use o fthe Xiaolangdi reservoir as well as inpower operations andpricing are all located inother entities. 2.105 Duringthe decade prior to the completion o fthe Xiaolangdi project in2002, the Bankwas heavily involved inassisting the Commission and inbuildingupits analytical capabilities. Today it i s evidentthat the Commission's technocrats have the full capabilities to analyze and manage the complex issues relating to the water resources o f the Yellow River basin. With their access to the latest analyticaltools as well as the possibility to seek advice from any international experts that they mightwishto consult, it i s clear that the Bank's institutional development and knowledge management role has beensuccessfully completed. RISKTODEVELOPMENT OUTCOME 2.106 Overall risk i s rated as significant primarily because o f unresolved water allocation issues and under-priced water. Continued subsidies for capital investmentrisks misallocationo f resources. While there are potentially highseismic risks, these have been mitigated inproject design and operation. The sustainability o fthe Xiaolangdi project i s assured from a technical and financial point o fview becausethe project infrastructure i s being well maintained and the Corporationi s able to raise enough funds from its sales of electricity to meet its full 0 & Mrequirements.However, from an economic perspective, long-term sustainability i s less likely. Continuing water shortages, the modest progress in addressing issues o f inter-sectoral and regional competition for water use and the political reluctance to use water pricing as a tool for demand management all tendto undermine economic sustainability.  41 2.107 Inter-regional equity and conflicts over water. As pointedout as long ago as 1992 bythe Investment PlanningStudy: 2.108 "...the remote, heavily erodedLoess Plateau and the arid upper reaches of the Yellow River ...are the least eflcient water users. Delivery andfield losses are extremely high and many areas must be served by energy-intensivepumping schemes. With increasingfrequency, Yellow River water will be in short supply, and the water used in these regions will have an ever higher opportunity cost downstream.Sooner rather than later, Yellow River planners w ill have to squarelyface equity vs. eflciency issues. '' 2.109 The differences inthe needfor Yellow River water by seasonandby reach create conflicts betweenupstream and downstream provinces, since their priorities over water use do not coincide with a basin-wide water allocation formula based on maximizing economic benefits. 67 The lower reachesare more productive interms o f the annual output o f grains and their efficiency o fwater use, but farmers have to make do with what i s left after the upperandmiddlereacheshave drawnofftheir `requirements'. Agricultural output couldbe raised further ifmore irrigation water was provided to farmers inHenan and Shandong provinces. The highereconomic value o fwater inthe lower reaches should also translate into a higherwillingnessto pay for additional irrigation water. 2.110 At present, provinces along the Yellow River are entitledto withdraw annual quotas o f surface and groundwater that have remained unchanged since a State Council ruling in 1987, despite major shifts inpopulation, urbanizationand industrialization that has taken place inthe past 20 years. While the Regulation Stipulation o f 1998 allowed the Commissiontopro rata the quotas to take account o f annual variations inwater availability, redistribution o fthe quotas i s not allowed. Evenso, upstreamprovinces tendto withdraw more than their full quotas every year, regardless o f annual rainfall, which puts a disproportionate burdeno fthe shortfall due to poor hydrology on the downstream provinces. Altering these provincial quotas and puttinginplace an enforcement mechanism i s a political `hot potato' for central government that needs to be addressed. And as noted earlier, downstream provinces are able to bypass official regulationmaking nonsense o fthe quota allocation system (paras 2.88). 2.111 Taking the Yellow RiverBasinas a whole, there seems to be scope for a rational reapportioning o fprovincial water offtake quotas insuch as way as to raise overall net economic benefits. Ina capitalist market system it would be possible to arrive at this end result by trading water rights. This i s currently not feasible inChina. But it i s possible to envisage a government administered `trading' scheme to value water, reapportion provincial quotas and financially compensate those upper and middlereach provinces that surrender aportiono ftheir quotas to lower reachprovinces. 2.112 An integral element o f such a scheme would be the introductiono f higher irrigation water charges and volumetric pricing inthe lower reaches inreturnfor guarantees o f additionalwater supply inthe growing seasons. The upstreamprovinces would receive an annual payment from the central government to invest inrehabilitatiodupgrading o f their 67. In 1997 water usewas 809m3/muinthe UpperReach, but only 304 m3/muinthe Lower Reach(WB rpt # 22040, May 2002).  42 irrigation networks to cut water losses and cushion the impact o f reduced water availability. A panoply o f accompanying measureswould also berequiredto improve irrigation and cultivationtechniques to raise the value added from irrigation water inthe upper and middlereaches.Unlikelower reaches, water chargesto farmers could be left unchanged untilirrigationsystemshadbeenupgraded. Fromthenon, highercharges would beneeded to further encourage the economic use o f water inagriculture. 2.113 Without such direct interventionby the central government more efficient inter- provincial water use allocations cannot emerge inthe Yellow River Basin. Presently,the upstream provinces have no incentive to surrender any o ftheir actual water quotas. The downstream provinces have to make do with what water remains inthe lower reacheso f the river and evenifthere was a readiness ontheir partto pay for highervolumes, no inter- provincial trading mechanisms exists to enable this to occur. 2.114 Underpricing ofwater reduces return on investmentand lowerswater use efficiency. The Xiaolangdi.project has not lived upto the expectations o fthe Bank as regards the commercialization o fwater supply. Contrary to expectations, the Corporation does not charge for water or receive any portion o fthe water charges levied by Commission. Furthermore, water tariffs, particularly for irrigation, remain far lower than the economic cost o f supply, although at the time o f appraisal by the Bank o fthe first phase, the Commissionwas supposedly already committed to full-cost recovery pricing. 2.115 Yet inthe Project Agreements with the Bank for both Loans, the Corporation committed itselfto setting water rates sufficient to cover the full O&M costs right after commissioning the dam and to achieve full capital cost recovery after 12 years from commissioning. These undertakingsseemed feasible at that time, and were entirely consistent with the 1988 Financial Directive o fthe Ministry o f Water Resources that endorsedmarginal cost pricing for water and complete cost recovery pricing by 1997. But a decade later it i s apparent that the target was overly ambitious and i s unlikely to be met for quite some more years to come. 2.116 The Bank raised the bar evenhigher inits loan for Phase 11.The PAD (section 14) indicated that the Corporation"would initiate chargesfor flood and sediment control services, which would be appliedas a surcharge on the billsfor end users of Commission - supplied water." However no formal undertaking to this effect was included inthe legal documents. The lack o f action since thento introduce such charges i s unsurprising, given that many view such services as public goods to be providedby the state. 2.117 Onthe other hand, the slow progress inraising user charges for irrigation water and the lack o f any financial transfers to the Corporation for water from the Xiaolangdi reservoir i s more difficult to defend. The newwater law o f 2002, inaddition to authorizing full cost recovery, (including for flood and sedimentcontrol), also makes explicit provision for a resource cost element inwater charges to reflect scarcity. Translating the law into 68. This would reassurethose provincessurrenderingthe water quotas that compensationwould be paid. Initially the centralgovernmentwould probably needto allocate additionalbudgetaryresourcesto cover the cost of the scheme, but it couldequally oblige the downstreamprovinces to transfer some o f the revenues from higher water charges to cover these payments.  43 highertariffs is politically difficult. Although Corporation is an enterprise under MWR, approval o fwater tariffs i s a provincial matter and apart from the political unpopularity it would face ifit steeply raised water charges, the provincial government i s also unenthusiastic about a measure that would transfer funds to a central government body. 2.118 In2004 MWR introduced the regulations relating to water tariff settingthat emanated from the State Council decision designedto promote water conservation. InJuly 2006, water prices inthe lower reaches o fthe Yellow Riverwere raised, but they still do not contain any elemento f cost recovery for the 0 & Mor capital cost o fthe Xiaolangdi project. Current bulk water charges leviedby the Commission on supplies deliveredto the provincial river affairs bureaux are s h o in ~the Table 8. Subsequentlythese bureaux retail the water to water user associations with anadded management surcharge. 2.119 These charges, which produce revenues o f about Y100 millionannually for the Commission, Tab,e 8: Bulk water prices do not evencover the full capital cost o f YRCC bulk water supply charges (fedm3) investmentsinwater distributiondownstream o f Xiaolangdi. Evenallowing for retailers' mark-up it Price April-June July-March i s apparent that the gap between the price and economic value o f water i s enormous, particularly Irrigation 1.2 1.o for irrigation. The estimated economic value of water inthe Yellow Riverbasin in2000 was inthe M& Iwater 9.2 8.5 range o fY1.4-1.7/m3. Farmers currentlypay only about 1-2 percent o fthis figure.69A similar disparity was seen inthe Tarim Basin. 2.120 Continued subsidies for capital investmentgive the wrong signals on fiscal responsibility.Despite Xiaolangdi's lower than anticipated electricity tariffs andthe absence o f any revenues from water sales, the existing level o ftariffs have proved sufficient since 2005 cover its O&M expenses, as well as pay the debt service on the loans contracted for the project (including the IDA credit for the resettlement program). The financial covenants inthe loan agreement with the Bank have beenmet, with the exception o fthe rate o f return on the Corporation's total revalued fixed assets, which with hindsight appearsto have beenoverly ambitious, giventhe difficulty o f imposing charges for flood control and sediment reduction. 2.121 Prior to 2005 the central government was obliged to cover part o fthe debt service on the IBRD loans as electricity sales were still rather low and central government only exemptedthe Corporationfrom VAT in2006. Since 2005 the increase inelectricity sales has allowed the total debt service relating to Xiaolangdi to decline steadily and in2007 this i s about Y 640 million (US$78 million), a third less than two years previously. The windfall from VAT exemption i s worth about Y 70 million (US$8.5 million) a year. As a result, the Corporation will start to accumulate cash reserves that could be used inpart to finance new investmentsand/or acceleraterepayment o f outstanding debts. Alternatively, as a wholly state-owned entity, the government could require the Corporation to pay it an annual dividendonthe equity initially investedinthe project. At present there i s no 69. Source: Table 4.9, Agenda for Water Sector Strategy for North China (Rpt # 22040, May 2002)  44 government policy regarding the return on investmentsmade by it inmajor infrastructure projects and these are treated as 100 percent capital grants. This seems an overly generous approach inthe case o f revenue-earning enterprises like the Corporation. 2.122 The obligation o fthe Corporationto provide a post-resettlement fund for the 184,040 people displaced by the project was taken over by the MWR due to the inadequate cash flows o fthe Corporation inthe first few years o f operation. Subsequently, central government has instituteda national subsidy for resettlers -about Y 600 per capita per year for 20 years -to assist themto enter the economic main stream. Now that the Corporation i s financially healthy becauseo fthe large capital grant from central government there i s a danger that Xiaolangdi could set a precedent that large public revenue-earning enterprises are not financially responsible either directly or through taxation for the longer-term negative social impacts they create. BANKPERFORMANCE 2.123 Overall Bank Performance on Phase Iand Phase I1i s rated as highly satisfactory. Quality at entry: project designand appraisal were very thorough and o f a highstandard. Hence quality at entry i s rated as high. Quality o f supervision: the project benefitedfrom frequent and regular supervision, and long continuity o f Bank staffworking on the project, particularly through preparatiodappraisal and Phase I. The postingto the Bank's Beijing office o fthe project team leader during crucial years o f implementationwas also valuable inprovidingcloser andeasiercontact withthe project. Supervision quality istherefore rated as high. 2.124 Inaddition to direct project supervision, the Bank also providedthe government with significant analyticaland advisory services inthe water sector. For example, the Bank carried out a comprehensive andwide-ranging study o f water sector issues inNorth China in2001-02. While itappearsto havebeeninfluentialinshiftingthe government's thinking at the highest levels, translating it into action seems to be proceeding rather slowly, given the urgency to addressproblems for which there are no quick fixes. BORROWERPERFORMANCE 2.125 Borrower performance i s highly satisfactory for the Phase Ithat focused on construction management, financing and application o f safeguards. Phase I1performance i s ratedas satisfactory basedon the relative contributions o fthe project and governmental entities involved inthe project -greater weight being givento slow progress on institutionalreform andwater pricing. 2.126 The main implementingagency successfully implementeda large and complex project with many engineeringand management challenges. Inaddition, the project was completed under budgetand ahead o f schedule. Since completion the Corporation andthe Commissionhave also demonstrated their capabilities to manage the Xiaolangdi project facilities to highstandards. For these reasons implementingagency performance i s rated as highly satisfactory. 2.127 Government Performance under the project i s rated as satisfactory, albeit with some reservations. Despite the evident in-house expertise inthe Commission and the  45 dissemination by them o f advice and recommendations on addressing the tradeoffs inthe use o fYellow River water and water pricing reform, it would seem that political will i s lacking, particularly at the provincial level, to face upto the hardchoices. Similarly for many o fthe same reasons electricity tariffs have beenmaintainedat a low level. In addition, at the national level, it i s unclear yet whether there i s any appetite to face upto the difficult issue o f intra-provincial water allocation quotas which have remainedfixed since 1988. The problems are well understood and the solutions are clear. SUMMARY OF ISSUES XIAOLANGDI PROJECT RAISED BY THE 2.128 The Xiaolangdi experience raised five issues: 0 The importance o fwell designedM&E systems that clearly define the counterfactual and the means to monitor agricultural and socio-economic impacts in additionto the physical impacts. 0 The correct pricing o fwater inits various uses inorder not to distort its allocation 0 The unwillingnesso fprovincial governments to levy increasedwater andpower tariffs according to national policies and approved by the central government 0 The central government's unwillingnessto recover capital investmentfrom water and hydropower utilities that could be commercially viable 0 The unwillingnessonthe Yellow River Conservancy Commission to use economic criteria for water allocation, raise water tariffs for agriculture and pay the Hydropower Corporation for the water it utilizes downstream. TARIM BASINI1PROJECT OBJECTIVES 2.129 The Tarim Basin I1Project (Tarim 11) aimed to build on and expand the achievements o fthe first Tarim IrrigationProject (Tarim I) that was completed in 1997. The Tarim Iproject markeda new direction inwater management for the Bank inChina in that it targeted the majority indigenous peasant populations and the irrigation systems which sustain the oasis inwhich they live. Its primary objective was to introduce some advanced technologies to detect and correct for alkalization inarid zone irrigation, improve soil and water management to preventdesert expansion, and control and mitigate hail damage to economically vital fruit and vegetables crops. At completion in 1997 the outcome o f Tarim Iwas ratedas Highly Satisfactory inthe Implementation Completion Report (ICR). SubsequentlyIEGrated outcome as Satisfactory after a desk assessment in 2 0 0 0 . ~ ~ 2.130 Objectivesof Tarim 11. Its three objectives were to increase incomes o fpoor farmers through irrigated agriculture development; establish systems to ensure sustainable use, development and management o fwater resources and land inthe Tarim Basin; and 70. IEG.PerformanceAudit Report.CHINA: NorthernIrrigationProject, Shaanxi AgriculturalDevelopment Projectandthe TarimBasinProject.ReportNo. 20683. June 29,2000.  46 partially restore and preserve the natural riverine forests along the lower reaches o fthe Tarim River. 2.13 1 Tarim I1effectively replicated the objectives o f Tarim Iinfive new sub-basins thus extendingmodernwater andagriculturalmanagement techniques to the whole basin.71In addition Tarim I1objectives put far greater emphasis on institutional development. This included support to the Tarim Basin Water Resources Commission established in 1997, prefecture water resources bureaus, self-financing irrigation and drainage districts (SIDDs) and water user associations. InTarim Iless than one percent o fproject investmentwas devoted to institutions for integrated river basin management; inTarim I1this increased to 14 percent. 2.132 The project's three objectives were to increase incomes o fpoor farmers through irrigated agriculture development; establish systems to ensure sustainable use, development and management o f water resources and land inthe Tarim Basin; and partially restore and preserve the natural riverine forests along the lower reaches o fthe Tarim River.72These objectives were to be achieved through 16 components covering physical works and engineering,land development, voluntary resettlement, agricultural support services and institutional development. Table 9 summarizes objectives, components and costs. 2.133 The first objective -to increase incomes o fpoor farmers through irrigated agriculture development -would be the outcome o f success inmeetingthe other two objectives derivedfrom the synergy generatedby the project components. For example, increased fresh water inthe TarimRiver would not only regenerate riverineforests but also forest products and grazing. The third objective -the restoration and preservation o fthe natural riverine forests inthe lower reaches o fthe Tarim River-would be the result o f reduced water use inthe basin. This would be the outcome o f more efficient agricultural water use and better water resources management brought about by improved institutions, planningandtechnology. DESIGN 2.134 While the objective to increase farmers' incomes could be fairly easily achieved through a traditional combination o f engineering upgrades to irrigation infrastructure, better operation and improved agricultural support services and market access, the other two objectives were far more challenging. Not only was the project to increase irrigated area by about 75,000 ha it also hadto save sufficient water to rejuvenate the lower reachesofthe Tarim River. 2.135 Ensuringsustainable use o f land and water resources was addressedthrough buildingwater planningknowledge bases and institutions for its improvedmanagement 71. Tarim Icoveredthree sub-basins:the Yerqiang BasininKashgarPrefecture,the WeiganBasininAkesu Prefectureandthe TarimBasin,which falls inthe Akesu andBayingolengPrefectures.TarimI1coveredfive sub-basins:the HotanBasininHotanPrefecture,the KashgarBasininKizilsuPrefecture,the AksuBasinin AksuPrefecture, andthe Kaidu-Konquebasinandthe Tarim River mainstreaminthe BayingolPrefecture. 72. These are the objectivesstated inthe DevelopmentCredit Agreement and, with very minor changes in wording, those inthe SAR.  47 underthe environmental protectionandmanagement component. Firstthe hydrologic system-the dynamic relationships betweenannual river and groundwater inflows and usage, agricultural practices, water storage, systemlosses and the relationship to waterlogging, soil salinization andwater quality variations -was investigated and understood. Second, this knowledge was incorporated into a fully calibrated mathematical model that identifiedthe most critical system improvements neededto conserve water and augment the lower reaches o fthe Tarim River.Third, the appraisal team recognized that prefecture administrations neededan incentive to buy into water conservation, particularly as most prefectures wanted an increase inthe area irrigated and more predictable irrigation supplies. To this purpose a TarimBasin Water Resources ProtectionFundwas established to promote investmentsinsub-basin water resources management that would be implementedbyprefectures. 2.136 The major institutional component -improvement o fthe current water allocation andlicensing system andcost recovery -was to ensure that water saved could be reallocated elsewhere. At the apex level this involved buildingthe planning, management and regulatory capacity o fthe Tarim BasinRiver Basin Commission (Commission) and the Tarim Management Bureau (Bureau). High-level study tours for senior Commission staff to river basininstitutions inAustralia and related technical assistance aimed at capacity- buildingwere financed by AusAid.At prefecture and lower levels capacity-buildingwas addressedthrough creation o f Water Supply Corporations and pilot water user association (WAS),attention to management of service delivery and betterO&M. 2.137 Project designmade explicit the relationship betweenimproved service delivery, water conservation and mechanisms for water fee settingand collectionthrough pilots in eachprefecture. And these efforts were aided by extensive training and technical assistance and support inwater resourcesmanagement, irrigation andwater conservancy, and agronomy. Itwas expectedthat all these efforts would produce a Tarim BasinMaster Plan and a workable TBRBC with the participation and acceptancethe prefectures by 2001 2.138 The project was not subject to a Quality Assurance Reviewby the Bank's Quality Assurance Group.  48 Table 9: Tarim I1-Objectives, Componentsand Costs Objectives Components ProjectCosts (US$Million) Planned Actual 1. Increase incomes of Projectcomponents listed below were to benefit existing poorfarmers through farmers inthe project area and inaddition this included a Included Included irrigated agriculture programfor the voluntary emigration and resettlementof insums insums development about 10,000 peoplewithin the project area. below below Institutional Strengthenthe capabilities of the Tarim Basin Water ResourcesCommissionand RegionalManagement 12.3 6.5 institutions inplanning and overall management ofwater and relatednaturalresourcesincluding MIS and M&E. InstitutionalDevelopment and Supportto Improve prefectureand county level water resourcesmanagement to ensure sustainability and self-financing of irrigated 15.2 17.5 agriculture, improve irrigation system performanceand effective decentralizedirrigation services, cost recovery and farmer participation. This component also included appliedresearchand developmentand anumberof 2. Establishsystems to specialstudies. ensuresustainableuse, Strengthen agriculturalsupport services to support 16.8 21.1 developmentand agricultural developmentofreclaimedland improved managementof water yields on low-yieldupgraded land. resources and land Improve EnvironmentalProtectionand Monitoring 17.7 16.9 through developmentof abasin-widewater and salt and balance modelto test alternativemanagement options for surface and groundwater. 3. Partially restore and preserve the natural Phvsical riverineforests along Water Conservancy investments to includeriver 120.3 116.8 the lower reaches of the diversion headworks, trunk and main canals, reservoirs, TarimRiver well fields, pumping stations and 2 hydropowerstations andrelated distribution networks. Land Reclamationcovering 75,400 hectaresofnon- 60.7 35.8 irrigated wasteland locatedwithin or nextto existing oases. Land Improvementcovering 105,400 hectaresof low- yield landthroughthe constructionofcanals, drains, 29.1 64.0 bridges, culverts, gates and farm roads, land leveling and planting oftrees. River Engineeringalongthe Tarim River to improve water managementand delivery ofwater to the lower 7.4 9.2 reachesofthis river andreplenishLake. Total Cost 272.6 287.8 IMPLEMENTATION 2.139 The project was coordinated by a multidisciplinary LeadingGroup under the leadership of the Region's FirstVice Chairman and the Vice Chairman responsible for agriculture and water resources. Similar groups were operational ineach of the 5  49 prefectures and 22 counties comprisingthe project. The MWR's Water Resources Bureau was responsible for project design, implementation, procurement, river basinwater management, and development o f self-financing irrigation and drainage districts. The TarimManagement Bureauunderthe Commissionwas responsible for regionalwater resourcesplanning and management and civil works on the Tarim River.The Regional PlanningCommissionwas responsible for reviewingo fthe project work plans, project monitoring and supervision, and coordination, mobilization and allocationo f counterpart funding. The Regional Project Management Office (PMO) and PMOs ineach prefecture were responsible for overseeing project designand implementation. Within the PMOs the Finance Bureauwas responsible for overall project management and financial management. 2.140 At the outset there were some serious problems and misunderstandingsbetween the regional and prefecture PMOs about the objectives o fthe project, andpoor coordination and cooperation. Misunderstanding stemmed from the common view that the project was to reduce water use inthe upstream areas andtransfer the saving to downstream. Thus many prefecture chiefs, particularly those inupstreamareas, felt that the basin-level management gave insufficient weight to their opinions andthe interests. More focused outreach and publicity efforts by the TMB started a year into the project and by 2001 a commonvision o f water conservation gained credibility among basin stakeholders at all levels. 2.141 Inadequate staffing, poor coordinationand cooperation at senior levels plaguedthe first halfo fproject and heldup attempts to apply unifiedmanagement and regulation to the basin. The incomplete complement o f senior staff, allied with insufficient and uneven fundingfor management activities, contributed to a lack goodcommunicationamong the TMB, the RegionalWater Bureauandthe RegionalPMO on integrated water and land resourcesmanagement. It was only inmid-2001 whenthe Regional Government's leaders -andtheregionalFinanceBureau(thelocalarmoftheMinistryofFinance)-tooka greater interest inthe project that a full-time Standing Vice-Director was appointed to manage the project. This with other senior appointments, and transfer o fhighly motivated staff to key positions, greatly accelerated progress andthe investmentcompletion rate increased from 42 percent to 90 percent over the period 2002-03. 2.142 Despite managerial improvements counterpart fundingwas problematic, slowing implementation untilmid-2002. Counterpart fund allocation, timingand amounts were not well harmonized with the project's requirements,particularly at county and prefecture levels. The multiplicity o f funding sources - seven regionalAoca1 and one central agency73- exacerbatedthe issue and annual shortfalls. Central funding was not immune to shortfalls either in2000, for example, there was no financial support from MWR for the Regional - Water Bureau whose activities ground to a halt as a result. The Bank-proposed "Water ResourcesProtection Fund"was inbreach o f central government rules and policies and this hinderedfinancial support for Commission. Ingeneral the shortage o fcounterpart funding 73. For example, in 1999 counterpartfunds o f Y272 million(US$34 million)were due from: CountiesY118 million(43%); PrefecturesY42 million (15%); Comprehensive AgriculturalDevelopmentFundY34 million (13%); RegionalCottonFundY30 million (1~`XO);Reliefkaborfund Y20 million (7%); Water Resources BureauY12.3 million (4.5%); andRegionalFinanceY5.2 million (2%) and local farmers.  50 was a serious problem for progress onthe capacity-building and institutional aspectso fthe project until2002 -irrigation and drainage engineeringwas less badly affected as it was drivenby local demand. Accordingly farmers' labor contributionwas generally more than planned. Local funding ceasedto be a problem after 2002 whenthe independentand centrally-funded Y10.6 billion(US$l.3 billion) Tarim Basin Water Resources RehabilitationProgram topped-up local contributions. 2.143 Procurement, despite beinggivendetailed by the Bank attention during appraisal, exacerbatedthe counterpart fundingproblem causing delays inproject works, technical assistance, Self-financing IrrigationDistricts (SIDD) pilots andresearch and implementationo fManagement Information Systems (MIS).Major problems were unfamiliarity with internationally accepted norms, insufficient staff and financial support at provincial and county levels, poor coordinationinthe regional PMO, and failure to follow Bank norms and procedures for procurement. While proactive supervision and additional training provided corrective action inmany cases this was not always sufficient. An ex-post field reviewo fcontracts in2004, for example, found that almost 60 percent o fcontracts examined hadprocurement issues, the most common being inadequate attentionto advertising bidpackages, inconsistent biddingdocuments andprequalification. Evenso, none were sufficient to invalidate disbursement. 2.144 Giventhe various delays the project closing was extendedby one year after mid- term review (MTR, June 2002) and some o fthe project components were adjusted. Objectives remained unchanged. The main changes introducedat MTR following pressure form the Bank were a 55 percent reductioninland reclamationto about 41,000 ha driven andanincrease inlow-yield landimprovement by 22 percent or 23,200 ha. These changes curtailed irrigation expansion and enabled water to be usedmore efficiently inline with basin development objectives. Monitoring and Evaluation 2.145 Design. The project gave great importance to Monitoring and Evaluation (M&E) and the establishment o f baselines, but providedonly modest guidelinesfor reporting inputs, outputs, outcomes and impacts. The PlanningDivisions o feach level ofPMO managedthe M&Eprocess focusing on inputs/outputs and environmental indicators. Each year there was independentM&Eo fkey socio-economic performance indicators. Halfthe project counties (11) were covered by the Regional Agricultural Department's Rural Economic Stations, the balance by Xinjiang's Agricultural University School o fEconomics and Management. The impact o fthe project was captured by M&E o f 1,100 households that includes about 6,000 people.74A survey selected 50 households from 3 townships and 3 villages stratified according to average income (high, medium, low) ineach o fthe 22 counties. Inaddition, sample selection included one non-project todvillage for each two withinthe project area. While the selectedhouseholds are described as "random" this is not 74. The survey included 6,491 peoplein 1998 and by2003 this had declinedto 5,962. Thus the sample ratio was slightly over 1:1,000.  51 the case. The Chinese designteam told IEGthat at the village level the households were purposely selected by the local heads o f admini~tration.~~ 2.146 Implementation. The baseline was satisfactorily established inNovember- December 1998. Although annual longitudinal surveys hadbeenplannedthis was only achieved for macroeconomic indicators gathered by state statistical agencies and the household surveys contacted to independentnon-project agencies. Project agencies (with some exceptions) only startedto take M&Eresponsibilities seriously with the 2002 change inmanagement andlocalfunding(paras 2.141). Procurement o fessential MISequipment was only completed in2001 and installation at prefecture level in2002. Thus comprehensive monitoring data for all indicators are only available for 2003 and 2004 - eventhenthe M&Ebudgetwas reduced by a thirdinthe last two years o fthe project.76 2.147 At the completion o fthe project funding for M&E ceased. While detailed monitoring o f inputs and outputs i s continued routinely at all levels, the cessation o fthe Tarim I1M&Eledto a lack o f focus on evaluating date collected and loss o fthe independentassessmento f outcomes and impacts. This is a major failing as the project benefit stream was still growing (e.g. orchards and the cumulative effects o fbetter water management.) Given the modest cost o fthe exercise - US$37,000 a year for Xinjiang University's survey o f 11 counties -andthe US$1.3 billionallocated to the follow-on Tarim Basin Water ResourcesProgram -this demonstrates little official concern for demonstrating, rather than asserting, impacts. 2.148 Utilization.Until2002 many o fthe requiredperformance management indicators were inadequately monitored. This adversely affected planning for implementation o fwater and environmental conservation measures and O&M, particularly at provincial level, and was heavily criticized duringthe Bank's first six supervision missions. Subsequently, attention to M&E improvedand this enabled improvements to project designand implementationafter mid-termreview. 2.149 Monitored data were extensivelyreported by the project and provide a wealth o f indicators and other information from which to infer project outcome and impact.77There were a numbero f shortcomings. First, a comprehensive evaluation o fall the data was hinderedby differing data protocols: Xinjaing Universitymaintaineda computerized data base; the RuralEconomic Stations kept only manual records. Second, although there were without project controls, the quality o fthese was eroded by the expansion o f the project area and spillover effects. Thus by the end o fthe project instead o f one control for two project samples, there was only one control for 4.7 project sample t~wn/villages.~* 75. The administrative heads selected specific households by income group following the guide 30% high, 40% medium, and 30% low income. The initial survey instrument contained 43 questions put by the Bank and the final questionnaire grew to 60 questions after local consultation and field testing. 76. Y300,OOO a year was the amount allocated to Xinjian Universityfor 11counties; in 2003-04 this was reduced to Y200,OOO. 77. Monitoringand EvaluationReport o f Years 2003-2004. Xinjiang Agricultural UniversitySchool o f Economics and Management. March2005. 77 pp. 78. The ICRnotesthat it used data only from Aksu prefecture to determine project impact on farmers' incomes (ICR, page 23.)  52 Although IEGinterviews o f XinjianUniversity staff confirmed the existence o f basin-wide none-project controls none o fthese data are presentedinthe official evaluation reports or were made available to IEG.Thus project impact was determinedfrom a single difference "before and after" approach including all the sample households. This may give a slight downwardbias to estimated project impact as it includes an unspecifiednumbero f "non- project" households. IEGwas also informed that there was no formal attempt to triangulate household findings with other independent surveys.79 2.150 Overall rating o fM&E i s modest. SafeguardCompliance 2.151 Five safeguard policies were invoked by this project. 2.152 Environmental assessment (OD 4.01) at appraisal categorized this project as "B." However, becausethe project includedthe new large dam at Xinir ina seismically active area and 75,350 ha o fnew irrigation on reclaimed land it should have been classified as a category "A" project.'' The regionjustified its `B" categorization on the basis that: the dam would provide water from ecological purposes; there was no involuntary resettlement; land restoration would reverse negative environmental impacts from earlier poor water management; that China's own environmental screening and mitigation plans were a satisfactory substitutefor the Bank's policy; andthat this was normal practice for Bank projects inChina. The debate about the correct environmental classificationo f agricultural andwater projects inChina is not newas shownby the InspectionPanel findings onthe Western China Project." IEG's reviewo f the Bank's water policy pointedout that the dilemma o f safeguard policies i s that unless they are mainstreamed by borrowers, they primarily serve to protect the Bank from risk and criticism.82As the mitigation o f environmental impacts o f this dam was screenedby NEPA and Chinese experts, as well as an independentdam safety panel (OP 4.07) approved by the Bank, and there have beenno reported problems, IEGrates the Bank as satisfactory on application o f its environmental safeguards-this i s reinforcedby the due diligence to verify impacts.83 2.153 IndigenousPeoples(OD 4.20). While the project also significantly affected indigenous people who are the majority inthe area, they are greatest beneficiaries and fully participated inall project activities. Within Aksu prefecture poverty and miniscule land holdings (1.3 muor 0.9 ha per capita) lacking dependable water supplies induced almost 13,000 Uygur farmers to opt for voluntary relocation from the poorest areas o f Keping county to the Laiqilang irrigation subproject. After screening to target viable immigrants, 10,000 farmers were each offered 4 muo f irrigatedreclaimed land. IEGwere informed that 79. Forexample, the AgriculturalStatistics Bureau's RuralInventoryTeamresultsshowed"higher net incomesthan the project surveysbutthey followedthe sametrend." 80. The Xinir damwas 18mhighand 12.2kmlong. The reservoircovered 16.8km2. 81.The InspectionPanel.2000. The QinghaiProject - A Componentofthe China: WesternPoverty ReductionProject (Credit No. 3255-CHA and LoanNo. 4501-CHA). Report No. 20739. April 28,2000. 82. IEG.2002.BridgingTroubledWaters -Assessing the World Bank WaterResourcesStrategy,p 22-23. 83. See project supervisionreportsandthe ICRAnnex 9, page 35.  53 while volunteers hadbeengivena 3-year transitionto establishthemselves, most have retainedtheir original holdings and hadnot yet movedtheir families to the area. 2.154 Involuntary Resettlement(OP 4.30). No involuntary resettlement was foreseen at appraisal. However, the PAD cautioned that it could occur as alignment o f canals and drains was finalized as insurance theBank and the Regional Government agreed a - ResettlementPolicy Framework that fully complied with OD 4.30. 2.155 InJune 2000 the Bank's supervision team found about 30 percent o f sites inspected involved land acquisitiono fwhich one include involuntary resettlement. The Laodahe canal involved relocation inMarch2000 o f a primary school and 4 farmsteads, acquisition o f 1,388 mu(92.5 ha) o f land, and related assets o f 273 farmers.84 The supervisionteam found that compensation rates were generally lower and not incompliance withthe resettlement policy framework. 2.156 A secondresettlement missionmetwiththose responsible inFebruary2001, mitigationwas agreed andthe June 2001 supervision missionreported full compliance. The average income o f all 273 project-affected people was improved substantially as a result o f the project. IEGvisited 21Brigade, Ayikule TownofAksu, andmet withthose resettled who confirmed they hadbeenfully compensated. The increased focus o fresettlement predatedthe construction o fthe Xinir dam and ensured the region's attention to agreed compensation policies. 85 2.157 Internationalwaterways (OP7.50). The headwaters o fthe tributaries ofthe Tarim Riverrise inthe Kyrgyz Republic and Tajikistan. As there was no adverse effect onthe quantity or quality o fwater flows to other riparians the Bank's regionalvice president approved an exception inAugust, 1997. 2.158 Integrated Pest Management (OP 4.09). The value o f crop losses to pests inthe project area averaged US$86 million a year and control was primarily through application o f chemicals pesticides. ElsewhereinChina poor pest management on cotton had significantly reduced yields o fthis nationally important crop. Thus IPMwas justified for economic and environmental reasons. The project, supported by AusAidtechnical assistance, successfully carried out training o ftrainers to buildthe capacity o fthe Regional Plant Protection Department and prefectures. The initial proposalto train farmers directly and empower them to make decisions about timing and application o f IPMproved to be unworkable -too many farmers to train, official reluctance to delegate IPMdecisions to 84. The Bank's Resettlement Thematic Supervision Mission (June 8-15,2000) inspected 23 locations within16 water projects in4 prefectures- Kizilsu, Kashgar, Bayingol and Aksu. Most involved less than one ha and were on state land. The mission report noted: the RegionalAgricultural Division was aware of possible land acquisition inthe field, but no reports.. ..The Yes o fNo involuntary resettlement is checked by the RegionalPMO from the 5 prefecture PMOSthrough telephones.There is no efficient and accurate reporting system to check whether there is any land expropriated or house demolished for construction o fthe civil works inthe project area." Supervision Report. June 26,2000. 85. The ICR reports (page 34) that average income per capita was Y 995 per year; by 2003 this had increased almost three-fold to Y 2,584 per year. People interviewedby IEGat 21 Brigade Village indicatedthat their incomes had "risen a lot because o fthe project but this could not be verified. "  54 farmers and the low priority for training whenbudgets were constrained.86At project closure 84 percent o f farmers relied on unifiedmanagement and instructionfrom Township Chiefs and extension technicians. Farmers provided free labor for I P Mas almost two-thirds i s applied man~ally.'~ 2.159 The impact o f IPMon Figure3: IPMwas successfulin reducing dependenceon chemicalcontrol reducing overall pesticide - application andthe use o f chemical - -0. Chemical-Biological - Others pesticides i s substantial (Figure 3). 100% 1 Total pesticide use declined by an 0 . average o f 90 tons/year over the 0- 0. 0. period 1998-04. Although the M&E 0. 'cr data show a clear inverse '0 relationship betweenpesticide application and economic losses, the economically optimal level o f * 25% applications was not determined." Pest infestations o f cotton affected - 3.7 percent o fthe area in 1998 and 0% I 4.1 percent in2003 and less than 1 1996 1998 2000 2002 2004 2006 percent o f corn inboth years. Although the area affected by pests Data: M&EReport2004 remains fairly constant the average value o f economic damage approximately doubled between1998 and2003 suggesting that morepest management is required. Relevance 2.160 The objectives o fthe Tarim I1Project were clearly highly relevant to the Bank's landand water strategies and CASs interms o fwater conservation and environmental goals and poverty alleviation. The choice o f a river basinwhose institutions for management are unifiedwithinone province provided simplifiedconditions that enabled piloting o fnew approachesto management that were unfettered by inter-provincial differences inviews, ownership or environmental conditions. Efficacy Objective1.Irrigated agriculture development increasedincomesfor poorfarmers. 2.161 Incomes. Agriculture provides 87 percent o f all income inthe household longitudinal surveys conducted by the project's M&E team. While gross incomes from 86. Mangram, James. 1999. Memorandum on the evaluation o fIPM, revisedtraining plans and reform needs. August 28, 1999. 87. Farmers are expectedto provide up to 50 days free labor for Township-directed work programs. Spraying pesticides from aircraft accounted for 14percent o fapplications; large machinery24 percent. 88. Linear regression shows: Economic loss (000 Yuan) = -0.4982 (Pesticide total intons) + 1122.5; R2= 0.529.  55 agriculture declinedby about 9 percent between 1998 and 2003, net incomes rose by 28 percent because o f reduced productioncosts (Figure 4). At first sight this would appear to indicate that the impact of increasedirrigationon productionwas marginal. Iffact, the productivity responsewas substantial (see para 2.162) andthe decline intotal incomes was a result o f ma'or price declines incotton and grains that accounted for 90 percent of sown area in 1998. Mucho fthe decline infarm-gate prices was the result ofthe East Asian W financial crisis o fthe late 1990s and the adjustmentsto WTO accession.g0 2.162 Partly offsetting the Figure 4 :Farmers' net householdincomesfrom agriculture price sensitivity o f strategic increased crops to international markets and stocks, the project facilitated the diversification o f 5,000 agriculture into higher-value Gross Agricultural Income crops thus reducingrisks to 4,181 farmers' incomes from 4.000 4,342'", , l * ~ ~ ~ ~3,944 3 /// I ;''I exogenous factors (Table 10.). Household (4) Agricultural income also per capita income 3*000 benefitedfrom increased and forestry and growth of livestock expenditure that almost doubled invalue Wan) 2,000 1998-2003. At project completion it accounted for 20 1,000 percent of gross income. 074 Economic forestry doubled in area to 127,700 ha and 1997 1998 1999 2000 2001 2002 2003 2004 increased supplies o f feed (alfalfa), alliedwith improved grasslandmanagement, allowed Source: M&E Report, op cit. livestock offtake to grow from 298,000 to 720,000 head. 89. With one fourth o f global output, one fourth o f cotton stocks, and approximately 30% o fworld consumption, China plays a major role incotton, affectingthe movements inprices. From 1994 to 1998 Chinese cotton farmers produced 102.9 million bales o f cotton while Chinese mills consumed 97.1 million bales, adding to their domestic cotton stock. During that same period the Chinese government, through various agenciesthat control cotton reserves, imported 12.5 millionbales o f cotton increasing their annual carryover stocks from 10.8 million in 1994 to 26 million bales in 1998, and China held 52% o f the world's cotton stocks. This extraordinary turn o f events createdsome o fthe strongest cotton prices seen in 30 years. Then, as the Asian financial crisis plunged virtually all internationallytraded commodity prices to 40- year lows, the Chinese government beganto dump cotton from their stocks and pushedworld cotton prices even lower. Source:National Cotton Council of America. February II,2005 90. Chinese cotton export price fell from its historic peak o f $8,912/ton in 1996 to a low o f $1,009/ton in 1999, recovering to $5,122 in2001 and again sliding to $591 in2002. Maize export prices fell from $158 ton by almost 25 percent over the period 1998-2000 and remained fairly stable around $120/ton until2004 when prices doubled. Wheat export prices increasedsteadily 1996-2000 to $$943/ton but dropped steeply reaching a low o f $282/ton in2003- but rising to $476/ton in2004. Source:faostat.fao.org. 30 April 2007.  56 Productivity Increased 2.163 Overall cottonyields were raised by almost two-thirds by 2001 but fell substantially thereafter possibly inresponseto reduced prices and cutback o f inputs.Conversely wheat yields steadily increased by 31percent over Table 10: Crop diversificationoccurred -are the period 1998-2002. Generally, existing cultivated(000 ha) farmland produced higher yields than either - 2003 1998 Change the improvedlow-yield fields or reclaimed GrossCultivated 654.6 704.9 +8% marginal areas (Figure 5). Grain 354.4 326.2 -8% Cotton 234.1 196.3 -16% 2.164 ProductionCosts. The overall Fruittrees 47.5 120.4 +153% reductioninagriculturalproduction costs i s Vegetables 17.0 33.5 +97% largely the reason for increasednet incomes Melon/watermelon 5.6 20.1 +259?'0 from agriculture over baseline values. The Alfalfa 12.7 56.6 +346% cost o f growing cotton fell by 16percent per Source: ICR ha drivenby a 35 percent reduction inlabor costs. Conversely, while labor costs for wheat and corn increased by almost a tenthper ha, material input costs declined by 20 percent primarilybecause o freduced Figure 5: Yields of major cropsimproved 1998- costs o f chemical fertilizers and manure 2004 andplastic film. There is insufficient agronomic and soil monitoring data to 8,000 determine ifthe reduced fertilizer inputs Marginal0 LowYield 0 Existlng are a reflectiono fbettermanagement and 6,000 higherapplicationefficiency, better croP Crop yield 4,000 rotations or the mining o f soil fertility.9 kgha 2,000 Other evidence suggests that agricultural extension was effective as 97 percent o f farmers in2003 responded that the service was either "very helpful" or Cotton m e a t "rather helpful.'' Source: M&E Reportop. cit 2.165 Increasedproduction at lower cost greatly improved the input-output ratio. Thus the overall cotton output rose 48 percent to 18.6 kg/lOOpan and wheat output Table 11: Changesin Net HouseholdIncomes, Yuan rose by 41 percent to 119 kg1100yuan. Exceptionally, the output ratio o f Sample year Low Medium High cotton on existinglandfell by almost a 1998 4,500 14,000 23,000 tenth.Withinthe project areathis 2003 9,000 14,000 34,000 causeda contraction o f cotton planted Incremental 4,500 0 11,000 area by 16percent. But outputs for 91. The TarimI1projectincludedalarge set ofactivitiesfocusedon improvingagriculturalsupport services. These includedextensionthat improvedandreducedchemicaluse. At the mid-termreviewthe projectincluded50 yuanper mu for all landreclamationandlow-yield landimprovementareas for the incorporationof greenmanure(nitrogenfixing legumes) which greatly improvedsoils and fertility, increasedyields andreducedneeds for chemicals.  57 cotton on formerly low-yield landrose by 143percent and on reclaimedmarginal landby 90 percent demonstrating the efficacy o f drainage improvements and new irrigation. 2.166 Net income. This increasedfor the highincome and low incomefarmers but remained fairly constant for the mediumincome group according to the ICR (Table 11). The reasons for this are unclear andthe M&Edata made available to IEGprovide no answer. Despite these improvements there still appearsto be insufficient income for schooling for some o fthe poorest - over a fifth o f surveyedhouseholds have childrenwho have not attended school, and 80 percent o fthese households gave "lack o f money" as the reason.92 Evenso, there i s no evidence that living expendituresare beingboosted by the sale o f household assets. Quite to the contrary, asset accumulation has occurred. 2.167 First reduction instate taxes and itemshandedover to the state collective increased farmer's incentives to become more productive. Overall taxation was cut by more than half and it accounted for only 9 percent o fnet income in2003.93Household expenditureon durable goods increasedinall areas (Figure 6.) although Hetianprefecture, the poorest, showed only modest gains. Average household area was expandedby over 30 percent and there was a shift towards more durable quality construction -almost a tentho f homes replaced earth walls with either wood or b&k or concrete. Figure6: Ownership of durable goods rose 2.168 Insummary, the objective to increase farmers' incomes was fully achieved. How far this i s attributable 80 - to the project i s unclear due to the C o l o r N lack o f without-project controls. L Motorcycle / 40 - Telephone Objective2: Systemsto ensure Households Washing sustainableuse, development and wlth unk s ' ~Machlne (%1 managementof water resources and 20 - Refridgerator land wereestablished. 2.169 This objective was substantially achieved with few 0- 1997 1999 2001 2003 shortcomings. Major components includedestablishment o f effective Source: M&E Reportop. cit. institutionsto allocate andregulate water use to increase land productivity, development o f a knowledge base to inform decision-making and infrastructure to enable efficient management o f water. EffectiveRiver BasinManagementInstitutionswere Established 2.170 The Tarim BasinWater Resources Commission i s China's only river basin organization wholly within one province that functions as a decision-makingbody which includes all the key administrative or management interests o fthe basin. Established only 92. M&E Report o f Year 2003-2004, page 62. 93. M&E Report o f year 2003-2004. Table 2-38.  58 six months before the project was approved, the Commissionwas guidedby the "Regulations on the Management o fthe Water Resources o f the Tarim Basin'' approved by the RegionalPeoples' Congress.The primarypurpose was to ensure that water resources development and management would be under a systemo f unified administrationand management, according to five principles: 0 rational development, utilization, protection and management o fthe water resources o fthe Tarim Basin, 0 exercise control over water disasters, 0 filly derive the comprehensive social, economic and ecological/environmental benefits o fwater resources, 0 ensure the sustainable development o f the national economy inthe basin, and 0 improve the livelihood o fthe people and the environment" inthe basin. 2.171 The Commission i s composed o ftwo parts:the Board o f Commissioners (called a "Standing Committee" inPRC comprising 14members that meet twice a year); and the TarimBasinManagement Bureau. The Regulations combinedthe two entities into one organizational framework, gave them enhancedresponsibilities and accountability, and providedmechanisms to secure adequate funding. A small Executive Committee was appointed by the Boardto act on its behalfbetweensessions.94Prior to the formation o fthe Commissionthe Management Bureauoversaw all activities and attempted to reconcile the differences among the prefectures inorder to promote regulations for water licensing and use according the China Water Law. However, being confined to the mainstream o f the Tarim River they lackedthe authority and status to bringthe prefectures together, and were independento fthe Standing Committee. Similarly the Committee was unable to address the inter-jurisdictional issues adequately partly due to their structure andpartly due to their composition. Thus, at the beginningo fthe project the institutional framework for water planning and management was weak and ineffective and integrated water resources management was absent. 2.172 Most o fthese weaknesses were resolved by effective Bank supervision that helped clarify the details andrationale o fthe reorganization and facilitated agreement among stakeholders inthe first two years o fthe project. A five-year development plan for the basinwas approved by the Standing Committee inAugust 1998. More importantly, in 1999 the Boardagreed to give greater representation to the prefectures. With a 14 memberboard, the five prefectures felt they hadlittle chance to represent their interests duringthe two annual meetingo fthe Standing Committee. And duringthe interim,the Executive Committee (that hadno representation from the prefectures) could take action on matters delegated to it by the Standing Committee, which may not be inthe interest o fthe prefectures. A rule change was agreedthat would require concurrence o f at least three o f the five prefectures concurrence inthe Standing Committee voting, andthat one or more prefecture representatives would be members at large to the Executive Committee. 94. The ExecutiveCommitteehadthe Vice-chairmano fthe Regionalgovernmentresponsiblefor agriculture andwater resourcesas its Director, andfour Vice-Directorscomprisingthe SecretaryGeneralo fthe Regional Governmentandthe Directorsofthe Planning,Financeand Water ConservancyDepartmentofthe Regional Water ResourcesBureau.  59 2.173 Despite these rearrangements o fthe Commission it still provedto be very difficult to reach agreement on allocationandregulation o fwater use inthe basin- eventhough the first quotas had beenintroduced in 1988 under the Regional Water Resources De~artment.~~ The primaryreason givento IEGwas that the goal o fthe Commission was to rejuvenate the flow o fthe lower Tarim River was at odds with increased use o fwater that was neededto meet the goals o fthe 1Oth Five-Year Plan. Specifically key goals o f prefectures (inter alia, to double farmers' incomes and the prefectural contribution to regional GDP) would bejeopardized with reduced water availability. Although there was a two-thirds majority initially insupport o fwater quotas in 1999this broke down and eventually (2002) the Commissionworked with the regional government to impose quotas to bringclosure to the debate. 2.174 These quotas are imposed on the prefectures and fines are leviedifthe quotas are exceeded.96Traditionally most prefectures have taken more thentheir quota while at the same time continuing to extend irrigation infra~tructure.~~Fortunately, an effective dispute resolution system forms part o fthe Commission and this was satisfactorily resolved. Thus while most prefectures agreeto reduced water quotas, the date for reaching targetvalues have recently beendelayed from 5 to 15 years. Indications are that average usage inexcess o f quotas i s about 20 percent and that most prefectures have agreed to reduce consumption by about 10percent per year.98 2.175 Local institutionsfor water management significantly improved -but not as planned.Prefecture Water Resource Bureaus effectively manage water allocation to next level down, the counties, usingthe nationally-approved water withdrawal permit system. At this level six Water Supply Corporations (WSCs) were supposedto receive the bulk water supply and retail it via contracts to newly-created and self-managed Water User Association that, inturn, allocates it to the farmers. Inpractice it was found that the WSCs duplicated many o fthe functions o fthe exiting prefecture and county water bureaus and only one inAksuprefecture (serving 19 W A S )was still active at project completion. Also it was foundthat WSCs becameanextension ofthe township water bureaus andadopted a paternalistic approachto management that precludedW A Staking full responsibility for their ownaffairs. 95. The quotaswere developed by the RegionalWater ResourceBureau("top down") on the basis ofregional water balances, actualand proposed infrastructureand cropping system, and other water demands. Subsequentlythis was supplementedand adjustments madeon the basis of detailedwater balancesconducted throughthe project at country level ("bottom up"). Both could be conducted independentlybut as the knowledgebase is built there will be aconvergencetowards a commonly agreed quotavalue. Thus acounty could aim to save water through engineeringand management improvementsand put part ofthis saving towards reducingits use to meetthe Bureaus quotaand part towards servingnew irrigation or other water demandssuch as industry. 96. In2003 Aksu Prefecturewas fined 10,000 RMB for exceeding its quotabecause subdivisions 1and 2 exceededtheir sub-quota. They, inturn, were fined 6,000 RMB. 97. In2001, for example, the Kizilsuprefectureargued that althoughits historical usage was 0.107 billion m3, new irrigation constructionover 130,000 mu (8,700 ha) would pushtheir requirementto 0.196 billionm3. The quotaproposedby the Bureauwas 0.135 billion m3. 98. This information from IEG interviewswith Bureauofficials inUrumqi partially contradictsthe mostly "full compliance" notedbythe 11" SupervisionMission(para 11) in2003.  60 2.176 Upgraded or new irrigation and drainage infrastructure met all the project's physical targets andthis providedtimely andreliable water supplies that increased agricultural prod~ctivity.~~ This fully functioning system providedthe basis for the establishment o f viable WUAs whose numbergreatly exceeded appraisal expectations: 15 against the 8 planned. Inaddition another 208 were establishedbefore project closure and by the time if the IEGmissionthis hadreportedlyincreasedto over 1,500 including those outside the project areabecause o fthe Regional government's "Delivering water to Households" program that aimed to engage farmers ina more participatory approach.loo Onthe whole IEGfound from a series of field interviewsthat most WUAs are effective inmanaging and conserving water at the local level, but not without problems (Box). 2.177 Some WUAs are inname only -farmers do not actively participate intheir management. The share o f "paper" WUAs i s not known. Inmany o fthe poorer areas that IEGvisited, farmer's literacy is low (reportedly about 20 percent) andthe country officials take the initiative to organize the WUA and then runit as an extension o fthe township, water bureau or state enterprise. For example, at the tail end o fthe Dongfeng main canal in Aksuprefecture the TienFeng Seed Company (a state enterprise) manages one o fthe project's pilot WUAs that i s part o f a SIDD that includes 8 villages created on 11,000 mu (733 ha) o f reclaimed saline soils. Comprising 3 villages the farmers inthe WUA neither manage budgetnor water -everythingi s done by the state enterprise staff-but they do undertake O&M withinthe command area. Water tariffs are calculated from the cultivated area, not volumetrically. Despite beinga "paper" WUA water management followed the conservation principles inthe project designand the state enterprise staff were assiduous in collecting water fees and ensuring O&M was undertakenby beneficiary farmers. This confirms the findings o fresearch inNingxiaand Henanprovinces that, when implemented, water management reform leads to water savings and meets the primary goal o f water sector officials.'0' This study also noted increases infarmer's incomes appeared to more closely allied with manager's incentives thanwater reformper se -although reform may lower farmer's overhead costs (Box 2) and reduce water bills. 99. 123,373 ha of land were improved and 41,462 ha o f land were reclaimed. This included 1,009 km o f main canals; 1092 km o f main drains; 98 reservoirs and live storage; 29 water resource measuring stations; 2,340 kmo f lateral canals; 3,108 km o f field canals and 20,846 water control structures. 100. This program aims at improving water delivery efficiency to farm households through more accurate water measurementand accounting and application o f volumetric charging. 101. Wang, Jinxiz, Zhigang Xu, Jikun Huang and Scott Rozelle. 2003. Incentives to mangers and participation offarmers: which one matters for water managementreform in China? CCAP Working paper 03-El7.  61 Irrigation and Drainage Districts are not self-financing 2.178 The concept of SIDDS has taken a firm root inthe basin and IEGfound compelling evidence that accounting and billing procedures for water has been mainstreamed within the project area. Billings actually collected were generally better than 90 percent. But none ofthe SIDDs visited was fully self-financing. The primaryreasons are that water tariffs are too l o w and do not cover costs, and that the prefectural authorities prefer to reduce taxes and fees on poor farmers and subsidize water management activities. Inthe Qinbuluguqi WUA ofAksu prefecture, officials indicatedthat the present water fee of0.0594 RMB/m3 wouldhave to increase by about 30 percent to cover the SIDD's costs. Table 12 shows how water fees are collected and allocated inthis area. Objective3: The natural riverineforests along the lower reachesof the TarimRiver were partially restored andpreserved 2.179 Despite increased irrigated area and water use, the ICR and Table 12: How water revenuesare spent-Qinbuluguqi Chinese officials argue that better WUA, Aksu Prefecture and improved institutions, Sources Allocation Revenueusedfor: infrastructure and management, and shifts incropping patterns, 10% Administration and O&M o f canals and resulted inwater savings that were Revenue generated from drains supplying WUA more thanthe additional demand. water fees collected by WUA and given to water 35% Township Thus this objective was achieved resource bureau administration as a the result of the synergy 25% O&M o fwater source, generated by the reduction of canals and drains water losses incanal systems, land 15% Flood control leveling, improved water 25% Staff salaries o f water application efficiency to crops, resource bureau more effective use of basin storage W A Sown fee levied 100% Administration and O&M -most frequently Y -bothreservoirsandgroundwater On farmers - and better water scheduling at all .-- paid infree labor levels.`u' Source: IEG interviews of WUA and SIDDofficial and beneficiaries.November I.2006. 102. The Tarim river sub-component financed activities on the middle reaches o f the river to enhance conveyance o f water downstream This including 325 km o f river training works, flood control levees and 27 ecological and water control structures inthe levee system to rationalize diversion o f water from the mainstream.  62 Box 2: GrowingWUAs the experienceof Mousonma Irrigation District, Kashgar prefecture - The project started a pilot self-financing irrigation and drainage district (SIDD) from an initial WUA that included 5 villages served by a 14.3 km lateral, mostly unlined, canal covering 4,500 mu(300 ha). 150 local officials were trained inWUA management and improving water use efficiency and this was extended to 200 farmers from 17 townships and 22 villages. Overseas study tours were paid for by the prefecture. Over the period 1998-2000 about 13 km (81%) o f the branch canal feeding the lateral canal were lined and 63 water control structures installed (including 57 water measuring points and 4 water level controls). By 2003 the county included 21 W A Scovering 209 villages and 66,915 households over an area o f 589,570 mu (39,304 ha). The water supplyto the WUAs is managed by the Water Bureau on the basis o f cropping patterns, water demands and schedules compiled bythe individual W A S .The W A Smain functions are to: *** plan raise awareness o f water conservation techniques and apply them; and manage distribution o f water with the command area and provide O&M; arbitrate disputes, audit supplies and collect water tariffs. Main problems: *The portion o fthe water tariff retained by the WUAs covers only halfthe required O&M, in consequence there have been canal breaches and wastage o f water; ** The project neglected to upgrade on-farm irrigation distribution facilities; Water tariffs are too low. The prefecture sets the base rate and this is implemented by the county with upward adjustments for local circumstances. Eventhough the county has the final authority over the actual rate, prefectural concerns to reduce the burden on poorer farmers normally means that the county has no say inthe tariff actually charged to farmers; * About a third o fthe tariff revenue raised from the W A Sis used by the county for flood fighting and reliefworks; and * The pilot WUA's 3 technical staff are free inthat they are paid for bythe water resources bureau and the village committees from other income. Advantages (as reportedby a Group Headwho farms 11mu and manages supplies for 41 households): ****Before the project it took 40 hours to distribute the water, it now takes 3 hours; Earlier every household had to send people to ensure their fair share; now we trust the WUA Earlier we had many quarrels over water, now almost none; Earlier we paid 50 RMB per household for water, now we pay 30 RMB mainly because rules are obeyed, water losses are reduced and management is improved; *** We undertake O&M and the WUA members agree to pay extra for this - normally inkind; We have diversified cropping to vegetables and fruits ingreenhouses usinggrants from county Household income was 3,500 RMB, now it is 6,000 RMB. Source: IEG interviews, November 6,2006, Kashgar and inthe field 2.180 A notable accomplishment was the introduction o fthe conjunctive use o f surface and groundwater to minimize evaporative losses o fwater.Io3Among the many water conservation measures successfully promotedby the project were use o f plastic tube and greenhouse cultivation, mulching and hightechnology drip and spray irrigation. 2.18 1 An alternative explanation -or contributory one - i s that perhaps the supply of precipitation to the basin inthe form o frain and snow has slightly increased. Conversely, if precipitation was less this would augment the project's achievements. This was not 103. Lowering the groundwater table through improved drainage reduces the rise o f water to the surface through capillary action and thus evaporative losses and salt accumulation at the surface. Flood flows from the mountainous rim o fthe basin percolate into the piedmont slopes and were captured through the installation o fwell fields and supplemented surface water supplies downstream. Reducing the area o f water stored in shallow surface reservoirs duringthe hot dry summers by either not constructing them at low elevations (e.g. at Louhuluke) or replacing them with high elevation supplies all greatly reduced non- beneficial evaporation.  63 investigated duringthe project and as far as IEGcan determine was not even considered in the results chainof cause andeffect. This is a notableomission that should be investigated before it can befirmly concluded that man-made improvementsare solely responsiblefor the improved environmentalsituation. 2.182 Whatever the cause, water savings have allowed six separatewater deliveries totaling 1.7 billion m3to be made to the lower river "green corridor" since 2000. This has allowed expansion o fthe area o fterminal lake systems at Taitema from zero to 200 km2. The increased river flows andthe reinstatement o fthe lake systemhave resultedinthe water table rising between 3.2 m and 12.6 m inthe lower river reach and a halving o fthe mineral content o f groundwater from 4-5 g/1to 2-3 g/l. 2.183 The increased supply o f better quality water has rejuvenated trees, shrubs and grasslands on both sides o fthe river providing food, shelter andwater for wildlife and people. Twenty-fivespecies o f native birds, amphibians, reptiles, 11 species o f fish and other wildlife, such as reddeer, have also returned. Tree ring analysis o fPopulus Divesifoliu (called huyang inChinese) showed a dramatic increase ingrowth rates after 2000 andthe area under huyang inthe basin increasedover the period 1998-2003 from 158,941 ha to 175,016 ha. This notable environmental achievement receivednational and international press coverage.lo4 Efficiency 2.184 Overall efficiency is rated as substantial. The economic rate o freturn (ERR) for the overall project was recalculated inthe ICRat completion basedonbenefits generated from landreclamation component and from diversification to higher-value crops. The recalculated ERR i s 19percent compared to 14 percent estimated at appraisal. The Net Present Value (NPV) achieved was Y675 million at completion compared to the appraisal estimate o f Y411 million. The major difference between the economic analyses carried out at appraisal and that done at completion i s that the appraisal under-estimated the scale o f diversification into higher-value agriculturalproductionand also the assessment o f increases inmajor crop productivity. These higher benefits were offset to a degree by lower economic costs at ICR compared to appraisal because o f China's entry into the WTO. 2.185 Inspite o f a significant downturn incotton, wheat and other prices had a major impact on the project inthe first and second years, with a corresponding reduction in house-hold incomes, the Financial Rate o f Return(FRR) for the overall project at completion was 16percent. The appraisal estimate o f 17percent was made based on market prices prior to China enteringthe WTO and includedallowances for the distortions inthemarketatthe time andhigherfinancial prices. A morerecent and independent estimate o f the costs and benefits o f growing fruit inAksuprefecture gives very favorable FRRsfor apples andpearsunderpresent agronomic practices. Itnotes, however, that profits would be evenhigher iftrees were better prunedandmechanizationwere introduced 104.ChinaDaily.2004. Quenchingthirst inTarimBasin.April 13, 2204.  64 as labor accounts for almost a thirdo ftotal cost~."~ Thus it is likelyfuture benefitscould bemuchhigherthan anticipated at completion. 2.186 This assessmentbelieves the actual ex-post economic rates o f return usingthe same economic and financial models could be lower. This i s becausethe cost stream inthese calculations only captures costs directly attributable to the project and fully attributed 4 benefitsto these inputs alone. Yet we know that other internationalsources o ffunding supported development inputswhich directly or indirectly benefitedthe project, as did China's own regional investments.Inaddition, there were manyregional, prefectural and country inputs- such as China's own comprehensive agricultural development program. Also much o fthe O&M expenditurei s giveninfree labor and not included. Balancing these downside factors, none o f upside environmental or social benefits from the project i s included either. Itmay well be that these factors cancel out ifcosts and benefits canbe quantified. OUTCOME 2.187 The outcome o fthe project i s ratedas satisfactory. Thejustifications for this rating are summarized inTable 13 and elaborated below. Table 13: Evaluation of factors contributing to outcome Objectives Relevance Efficacy Efficiency Outcome I.Increaseincomesofpoorfarmers High Substantial through irrigated agriculture development 2. Establishsystems to ensuresustainable High Substantial use, developmentand managementof water resources and land 3. Partially restore andpreservethe High Substantial? natural riverineforests along the lower reaches of the TarimRiver Overall High Substantial Substantial Satisfactow 2.188 While it i s clear that the project hada major development impact, IEGcannot be certain that the benefits generated inthe project area and captured by the M&E systemare solely attributable to the Bank's financing and support. Despite beinga second phase project no evidence o f a counterfactual was presented. Inadditionto direct and very substantial Chinese development financing for water development withinthe basin, six other external agencies provided developmental support. The ADB assisted development o f local highways that facilitated development o f local markets and commerce. The UK's DFIDsupportedwater sector reform projects intwo prefectures and Germany sponsored a water-saving projects: bothwould have had spillover effects. Inaddition the Japan International Cooperation Bank assistedwith Y1.97 billion (US$240 million) for water 105. Machesini, Sergio,HuliyetiHsimuandMaurzionCanavari.2005. Production costs ofpears and apples in Xingiian (China). DEIAgraWorking paper 05-003. Universityo fBologna.  65 development in 15 prefectures. Kuwait and Saudi Arabia were also supportive. For these reasonsthe ICR rating o fhighly satisfactory was downgraded to satisfactory. RISKTO DEVELOPMENT OUTCOME 2.189 Overallrisk to development outcome is rated as moderate. The mainrisk factors contributingto this rating are economic, technical, governmental, financial and exposure to natural disasters. Local stakeholder, social, financial and environmentalrisks are low. 2.190 Economic and technical risks relate mainly to agriculture - cotton and grains in particular - and are moderate to high.The volatility of internationaltrade incotton and grains at the start o fthe project clearly demonstrated the major role these play sustaining in farmer's incomes, and the risk factors are still moderate to high(footnote 84.)'06 This risk i s compounded by technical ones causedby the widespread adoption o f genetically engineered cotton knownas biotechnologyor Bt-cotton. The first risk i s continued inadequate institutional support and extension guidance for Bt-cotton that causes over- application, misuse or mishandling o fpesticides essential for the continuedhealth o fBt- cotton andthat have potentially adversehealth and financial impacts on farmer^.''^ The second i s the risk o f second purchasing Bt-cotton seeds. 3Risksto pest infestation(stink bug) and the continued highcost o f grains are primarily changing government policy in China and the effect of internationalprices on local producer^.'^^ 2.191 Continuedgovernment subsidiesto SIDDs, ifsignificantly reduced or curtailed, pose a low to moderate riskto the physical maintenance o f infrastructure constructed under the project. And ifthe effectiveness o fthis infrastructure is put at risk, water conservation may bejeopardized and with it the "green river." There i s also a risk that continued subsidies for water costs may underminethe rational for water conservation, giventhe prevalence o f low water tariffs throughout the basin, andthe uncertain or marginal financial viability of WUAs. 2.192 The easternpart o fthe Tarim basinexperienced a major earthquake in2001that damaged infrastructure and slowed project implementationinthat area. Giventhe proximity to active plate boundaries the likelihood o fthis risk remains moderate but local impacts could be large. 106.MacDonald, Stephen. 2007.Progress andproblems estimating China's cottonsupply and demand. USDA OutlookForum. 2007. 107.Pemsl, D, H.Waibeland A.P. Gutierrez.2005. Institutional constraintsfor thesuccess of agricultural technology in developing countries: the case of Bt-cotton in ShandongProvince, China. Proceedingsofthe GermanDevelopmentEconomicsConference, Kiel. 108.Grain.2007. Bt-Cotton,the facts behindthe Hype.Seedling.January2007. http://www.grain.orgheedling~files/seed-07-01-4-en.pdf 109.Zhou, Zhang-Yueand Tian Wie Ming. 2006. Evolving trends in grainproduction in China. Australian Agribusiness Review. 14.2006.  66 BANKPERFORMANCE 2.193 Overall Bankperformancewas highly satisfactory.Designand appraisal i s rated as highly satisfactory. Innovative project designreflectedboth the Bank's Water ResourcesManagement strategies andthose o fthe China' central government and the region. The Bank's extensive global learning and experience was appliedto the project and partnership with AusAidleveraged world renownedtechnical expertise to assist the nascent river basin commissionto solve its complex water and environmentalproblems. 2.194 Supervision i s rated as highly satisfactory. Attention to fiduciary and safeguard issues and physical progress was highly satisfactory. Missions were intense and reporting was excellent. Aide memoirs clearly identifiedthe main implementation and development issues and advice to the regional government and implementingagencies was clear, direct andrelevant while being flexible andpragmatic.. Progress towards physicaldevelopment objectives and emergingrisks were clearly tracked and ratings accurately reflected reality inthe field.The supervision strategy employed anefficient combinationofthe TTL/Co- TTL (based inWashington and Beijing, respectively) with clear roles and responsibilities and the ability to interact with and be responsive to the daily needs o fthe client. The only downside i s that insufficient attention was givento utilization o f the M&E, particularly the omissiono fthe climatic variables neededto conclusively demonstrate project impacts on the regionalwater balance. BORROWER PERFORMANCE 2.195 Overall Borrower performancewas satisfactory. The technical aspects o fproject preparation were highly competent, responsive to the innovative ideas importedby the Bank and the views o f local stakeholders inthe basin. As a result this technically complex project was prepared, appraised and negotiated injust 17 months. Borrower preparation, however, overlookedthe lack o f managerial capacity on procurement and M&Ebelow the main national and regional agencies andthis later caused implementation delays. Evenso, the regional government recognized the problem and was assertive ingettingthe project back on track and ensuringgood leadership. 2.196 The performance o fthe implementationagency was mixed. Very poor initial performance causedby an inadequate complement o fkey staff, ineffective leadership and deficient counterpart financing was countered by outstanding performance inthe second halfonce these problems were remedied.Subsequently there was effective coordination across agency, technical, administrative and political boundaries. An earthquake impeded progress over the period 200-2001. As a result, implementation o fthis technically complex and geographically wide-spread project was completed with a one-year delay. 2.197 Finally, the provincial government's ambivalent approach to raisingwater charges indicates they still do not see the link betweenprice signals on scarce resources (e.g. water) and their stated conservation efforts. As this poses a substantial risk to the sustainability o f irrigation improvements and the adequacy o f water to feed the "green corridor" this assessmentrates overall Borrower performance as satisfactory.  67 SUMMARYOFKEYFINDINGSFROMTHE TARIMBASINPROJECT 2.198 There are six: Thisproject conclusively demonstratedthat higher crop yields andincomes are possible usingless water. Provincial officials are reluctant to raise irrigation water tariffs to levels approved by the central government and agreedwiththe Bank. The linkbetweenwater resource conservation and its economic pricing does not appear to have taken root inWestern China. Despite the ideas introducedby the Bankthe Chinese continue to rely onthe administration, notthe market, to allocate water. While manyWUAs and SIDDs were formed and the idea o f self-reliance and self- financing i s strongly emphasized by Chinese officials, inpractice many WUAs and SIDDsarejust extensions ("paper WAS")ofthe top-down state administration designedto meet imposedtargets. 0 Operation and maintenance o f irrigation infrastructure will mostly rely on state subsidies becauseo f anunwillingnessto charge for water. Consequently SIDDs do not appear to be viable. As withthe other projects, greater care is neededto the design, implementationand applicationo f M&E. Local capacity-building i s needed. The results chain for achievement o fproject objectives needs to be carefully thought through. Ifthis had beendone it is unlikelythat the climatic element would have beenoverlooked.  68 3. OVERALLFINDINGSAND CHALLENGESLOOKING FORWARD FINDINGS 3.1 Overview. The three projects provide a great range o fexperience covering more than ten years o fwater planning and development inChina. The objectives o fthese prominent projects were key to the sustainable development o f China's land and water resources and have providedinsight into what works well andthe remainingchallenges. The overall impression is that China is superlatively good at constructing highquality infrastructure on time and generally within budgetbut i s less good at reforming its way o f doing business or developing new institutions. 3.2 Strongin-countrypresencematters.The mutualtrust and shared responsibility in the implementationand supervision ofthe projects was possible throughthe maintenance o f a consistent Bank task teams throughout implementation. The fact that the projects were task-managed from the field office facilitated Bank-Borrower cooperation andpartnership. 3.3 Key elements o f successrequire active Bankengagement over the long term and consistent promotion o f sustainable landpolicy and management practices. However, when projects are replicated indifferent areas care has to be taken that institutional assessment i s not basedon unfounded prior assumptions or higher level agreements. It i s notable below the central andprovincial level inChinathat locallevel institutionalcapacity and knowledge o f current best practice and i s very uneven. 3.4 Chinabenefitedsubstantiallyfrom the Bank's insistenceon applicationof its procurementand safeguard policiesand contractingprocedures. Staff ofthe central government, its agencies and provincial governments were unanimous inthe view that this greatly improvedChinese institutions, buildingsignificant human capital and making them internationally more competitive. Inthe Tarim Basinand at Xiaolangdi projects, in particular, improvedprocurement significantly lowered costs, as did introduction o f standard dispute resolution systems. 3.5 China's resource mobilizationwas problematic.At the provincial levelresource mobilization was problematic, widespread and created delays inorderly implementation. While the principle o fdevolving financial responsibilityto the lowest levels o f government are laudable they pose particularly difficult coordinationproblems for projects that cover severaljurisdictions. 3.6 Adjustment to the market-economyis slow -particularlyfor large infrastructure.Thereremains a willingnessto subsidize development evenwhenthis has strong private goods characteristics. The Bank expected beneficiaries to pay economic prices or at least the cost for the private goods generatedby public investmentinall four projects andthis was part o fthe project design. Despite this, central and provincial authorities were unwilling to raise water and power tariffs to agreed levels for major infrastructure. Conversely much smaller development was frustrated by the lack o f resolve  69 and/or ability to discriminate betweenpublic and private goods. Sometimes this created bottlenecks - for example inthe Loess Plateau many small dams that had significant down stream public benefits were cancelled because governorates only viewed the structure as providing local private benefits (inthis case newlycreated warping land and water storage.) 3.7 Integrated river basin planning has been piloted successfully but more needs to be done. The TarimBasinproject was the first successful attempt inChina at implementingintegrated river basindevelopment andmanagement. Itis notable for the inclusiono f environmental improvement withinits primary objectives. And unlike other riverbasin organizations inChina it is the first to promote andnurture strong links between the riverbasinmanagement agency, localpolitical stakeholders at prefectural and county level and farmer beneficiaries. As inseveral other Bank-supported projects inChina it has emphasized creation o f financially self-sufficient water user associations. But as with those projects, the government sends mixedsignals on cost-recovery and at the prefectural and county levels welfare consideration continue to outweigh financial prudence. 3.8 Thereis a lack of political resolveto reducewater consumptionthrough new institutions. The absence of strict enforcement o fthe water quota systemor realistic water pricing i s a major issue that could, inthe mediumto long-term, jeopardize project achievements. Undervaluedwater resources are may be misallocated. The larger issue o f maximizingthe value-added from water development has not beenaddressed. The current quota system by administrative allocation takes no account o fthe economic value o fwater or its opportunity cost incurrent uses. New mechanisms needto be agreed to develop criteria to allocate water efficiently. 3.9 Monitoring and Evaluation requires capacity building and reorientation to measureoutcomes and impacts.M&Ewas problematic for all four projects evenwhenit appearedto be carefully planned at appraisal. While all projects excelled inmeasuring process and physical indicators, most had difficult i s establishing robust systems to measure socio-economic impacts and interpret the findings. The major difficulty was that most o fthe implementingagencieswere engineering-focused and delegated non- engineeringM&Eto external agencies o fvarying capability and experience. CHALLENGES LOOKINGFORWARD 3.10 The overall conclusion from this project performance assessmentis that while the Government o f China has promotedwater use efficiency through individual projects there i s very little political will inthe provinces and below to make the hard institutional choices thiswould involve. Respondingto these constraints the government has continued to facilitate high level thinkingand policy debate on national water policy and modify the governance framework from time-to-time. 3.11 To provide a firm base for comprehensive management o f China's land and water resourcesthe government initiateda long series o f laws, rules and regulations startingin the 1980s.Initially these laws were reactive and aimed at addressing specific problems in water pollution, soil conservation, flood control, water management, environmental protectionand land administration. Later laws and regulations have tendedto strengthen  70 the increasingly holistic approach to water management although the pathhas beendifficult to navigate. The major difficulty has beenmodifying China's supply-drivenwater system, a legacy o fthe earlier planned economy, and harnessing the more recent demand-driven exploitation to create a law-based, economically efficient, and ecologically-sound water management regime.' lo Enabling environmentalmanagementthrough laws and regulations 3.12 Institutions to guide water planning, allocation, licensing, withdrawal permits, compensation, and construction, and consultation procedures for water disputes, were created by the 1988 Water Law. This Law provided the first comprehensive set o f principles to manage water inChina and i s the fundamental "umbrella" law for water management. Water use for domestic consumers was givenpriority as was balancing utilization against potentially harmful externalities. Water and flood management were unifiedunderthe MinistryofWater Resources and multipurpose water for use was highlighted.l1 3.13 SubsequentlyMWR's responsibility for watershed and landmanagement was formalized through the 1991 Water and Soil Conservation Law. This focused onprotection o f soil and water loss as the first priority o f conservation work, giving emphasis to aforestation and ecologicall-sound forest management, prohibition o f cultivation on steep slopes, and restoration o fvegetative cover after completion o f infrastructure projects. Specific policies to cope with soil and water loss inmountainous and hilly areas include integrated regulationandmanagement, establishment o f a contract system for regulation and management o f small catchments insoil eroded areas, and establishment o fmarket- oriented mechanisms for soil and water conservation. Although this ledto the development o f plansand local level regulations for fee collection to cover the cost o f erosion control facilities, lack o f funds at the local level seriously constrained its implementation and enforcement. The Bank's Loess Plateau and earlier RedSoils projects alleviated these constraints locally. 3.14 Furtheremphasison ariver basin approach inthe preparation o fflood controlplans and coordinationwith land-use plans developed by provinces and municipalities was authorized through the 1997 Law o f Flood Control. Although local interests were supposed to take a subordinate role inthe planning, the responsibility for flood fighting and relief was placedwith them following policies and management rules developed incoordination withthe river basin authority. Conflict resolutionwas to be throughnegotiation, with higher-levelgovernment -the MWR having final authority. Overall environmental safeguards were formalized through the 1989 Environmental ProtectionLaw that gave the State Environment Protection Administration (SEPA) the authority to coordinate and manage environmental protection. While SEPA hadthe overall responsibility for water 110. Boxer, Baruch. 2001. Contradictions and Challenges inChina's Water Policy. Development Water International. Volume 26, Number 3, Pages335-341, September2001. 111. The 1988 Water Law built on the 1984 Water Pollution Control Law (the first law for water and pollution management in China) and covered the integrated use o f surface and groundwater, provision of water for environmental management and water conservation.  71 quality and pollution issues, the MWRretained the authority to supervise and manage protectiono fwater resources, thus creating overlap o fjurisdiction insome areas. 3.15 Inpractice these lawsproveddifficult to apply for anumberofreasons. First,water resources and pollutionmanagement are primarily the responsibility o fprovincial authorities. Secondis that, while the central water and environmental agencies retained technical control over river basins, there was no single administrative authority looking after the spatial accountability from one province to the next within a single river basin- the TarimBasinbeingthe single exception. Third, as already noted, some provinces withdrewmorethentheir allocation. Onthe Yellow Riverthe periods o fno flow inthe lower reaches reached crisis proportions inthe late 1990sand only then did government react. 3.16 The drying-up attracted national and international attention that ledto anappeal signed by 138 academicians o fChinaAcademy o f Science and ChinaAcademy o f Engineeringto plead for "Acting Together and Saving the Yellow River. " This group subsequentlyundertook its own investigationthat covered mucho fthe river basin. At the same time the China Central Television and Economic NewsDaily co-organized a media campaign named "LongMarch Exploration to the Cause of Yellow River Dry Up" that heightenedpublic awareness.As a result, following approval ofthe State Council, the State Development and Planning Committee and MWRjointly issued the Yellow River Water RegulationManagement Regulation Stipulation inDecember o f 1998.'l2The beneficial effects o fthis on operations at Xiaolangdi have been discussed earlier, as was the way provincial politicians were able to bypass the regulation when it did not serve their purposes. Overlap and ConflictingResponsibilitiesfor Water FrustrateReformProgress 3.17 The single biggestproblem is that a large number o forganizations and authorities have administrative, technical or defacto responsibility for parts or facets o f a unitary resource - water. Inmany instances the ultimate authority i s unclear because o f overlapping jurisdictions or lack o f clarity inthe enabling laws and regulations. Often key stakeholders are not representedor informed. Evenwhenjurisdictional authority i s clear at the provincial level and below, the ability to effectively manage i s fi-ustrated by inadequate information or the unwillingnesso f local governments to cooperate on "common good" issues becausethey may hinder local economic development. 3.18 Tencentral ministries and agencies have overlappingresponsibility for various aspects o fwater resources and service management and administration. The MWR i s the lead agency that delegates water resources management and planningto the nine River Basin Commissions that, inprinciple, helpresolve conflicts betweenjurisdictions and sectors and ensure that multipleuses are inaccord with establishedpriorities. They have two important constraints. First, the Commission's authority applies only to inter- provincial waters. Second-with the sole exception o fthe Tarim Basin-they are not 112.Fu, Guobin, and ShulinChen. 2006. Water Crisis inthe Yellow River: Facts, Reasons, Impacts, and Countermeasures.Water Practice& TechnologyVol 1No 2. IWA Publishing2006.  72 accountable to, nor have any representatives from, the provincial governments or lower levels o f administration. Centrally-funded projects are supervisedby the MWR's Provincial Water Bureaux implementand supervise below this level, their Water Affairs Bureaux manage local level water resources strategy, including water quantity and quality, provide integrated management for city and country water services and pollution management. The arrangementsfor continuity o f water and landpolicy from the centre to the local level are admirable but it has a serious flaw that creates local conflicts o f interest. The Bureaux report and depend upon fundingfrom provincial governments whose jurisdictions are unrelatedto hydrological boundaries. More critically, staff o fthese agencies are employees o fthe provincial or country government, not the MWR. 3.19 Control o fthe water resource base and degradation o f it by pollution i s critical to maintainutilizable supplies. Yet overlapping responsibility frequently fi-ustrates this objective. Pollutionreductiontargets are based on a basinpollution planprepared by SEPA who implementthem through their Provincial Environmental Protection Bureaux (EPBs) at their various EnvironmentalMonitoring Stations. However, their responsibilitiesfor quality issues and planningoverlap the mandate givento MWA's Water Affairs Bureaux particularly for urban areas. As with water quantity, management o fwater pollution also has conflicts o f interest -EPBs are subsidized with income from wastewater pollution charges and fees -as local governments may put a higher priority on unrestricted industrial and urban development following implementationo fthe regional and governor's responsibility system. Thus local enforcement i s weak. 3.20 Overall coordination at the apex i s also weak. While the governing laws intended close cooperation betweenMWR and SEPA both ministriestendto work separately and submit separate andoften conflicting water management plans to the State Council. Inpart this is becausetheir basic water planningcriteria, data and assessmentareas are neither harmonized nor compatible.' l3 Central Committee did institute a national leading The group for Water Resources and Soil Conservation in 1988 to improve inter-ministerial and departmental coordination and similar leading groups at lower government levels. However, this lost highlevel support inthe 1990s and disappeared inthe 1998 reform o f government's organizations. The only residual i s the State Flood and Drought Control ReliefHeadquarters. Currently, provinces feel squeezedbetweenMWRand SEPA and would preferto align themselves with a unifiedbasin organization to ensure consistent and equitable treatment and better integrationo f water quality and quantity management. l4 ' Government response to coordination and management issues is slow 3.21 Responding to these problems, the Water Law was revisedin2002 to provide a more comprehensive planning and management framework for China's river basins. It endorsedthe twin principles o ftotal quantity control and quota management; reinforcedthe concepts o fpermits for water users and fees for supply, and reemphasized the importance o f improvements inwater use efficiency. The legislation also set downthe principles o f 113. World Bank. 2006. China Water Quality Management -Policy and Institutional Issues. 114. McMahon, George F. 2005. The Yellow River Law: A Framework for Integrated RiverBasin Management. Proceedings o fthe Georgia Water Resources Conference April 25-27,2005. USA.  73 water pricing reform, including the concept o f reasonable return, equitable sharingo f costs andthe principles of compensation andprice relatedto water quality. For the first time water markets were included among economic instruments as was the principle that the "polluters pays." 3.22 However, the revisedlaw makes no provision inthe RiverBasin Commissions for the representatives o flocal stakeholders at the provincial and municipal levels andtakes no account on inter-provincial riverbasin agreements with concerned governments.' l5No attempt was made to mainstream the Bank-assisted experiments on broader participationin the Tarim RiverBasinbecauseit was too earlyto do so. Indeed, inhindsightitappearsthat the newresolve impartedbythe 2002 law may have beenthe reasonthat the government of Xinjiang Province revitalized implementationo f institutionalreform which had been problematic untilthen. But four years later the unclear relationship betweenriverbasin commissions and local governments still hinders dispute resolution.' l6More importantly, the revisedlaw didnot clarify or redefinethe roles o fthe national levelwater and environmental organizations that had overlapping responsibility for managingthis increasingly scarce resource. 3.23 The devastating floods onthe River Yangtze in2002 refocused the government on the consequences o fpoor coordinationamong the multitudeo f organizations that hadmade environmentally sound river basinmanagement elusive. Inresponse the government set up a Task Force on Integrated RiverBasinManagement (IRBM)in2003 to help it redress these problems.' l7The objectives o fthe IRBMTask Force were to (1) assess existing laws andregulations andmakerecommendations to state legislationauthorities; (2) review existingriver basin management practices, assessthe coordinationo f existingriver basin management, andmake recommendations to the State Government and river basin commissions at the national level, on the Yangtze RiverBasin inparticular; (3) promote relevant economic tools such as water rights, water pricing, subsidies, compensation, tradable permits, and greentaxation for integrated water resources management at the national level and inthe Yangtze River; (4) promote stakeholder participationand community involvement; (4) provide a platform for information sharing; and (5) establish and promote communicationtools including workshops and publications. 3.24 The 2004 report o fthe Task Force provided four far-reaching policy recommendations to China's State Council. It recommended establishment o f a national- 115. Wang, Yuhua. 2003. Water Disputeinthe Yellow River Basin: Challengesto aCentralizedSystem. Commentary.Woodrow Wilson Center. Washington.D.C. August 2003. 116. GanyanZhou. 2006. InstitutionalArrangements for River BasinManagement:A Case Study Comparisonof the United States and China.The World Bank China Water AAA. Study SummaryNote Series. Washington.D.C. 117. The Task Forcewas established bythe ChinaCouncil for International Cooperationon Environment and Developmenton March28,2003. CCICED is an advisory group supported by CIDA andthe Chinese Governmentand is chairedby a leader ofthe State Council of China.The Chinese Council Membersare ministers, vice ministers, and famous scholars inenvironmentand development; the international Council Members are of comparable stature.The Task Force comprised 12prominentexperts; six from China and six others drawn from Japan, Netherlands, UK, USA, the Ramsar ConventionBureauand WWF International. Source: http://www.ramsar.org/wn/w.n.china_yangtrce.htm.  74 level IBRMcommission involving the National DevelopmentReform Commission, MWR and SEPA that would coordinate policy andplanning, establish standards andtargets and resolve conflicts. 11* Practically it proposed that IRBMprinciples should be piloted intwo priority tributaries o fthe Yangtze RiverBasinto rationalize existing laws to reduce contradictions that hinder institutionalreform, and that subsequentlyMaster Plans for each major river basinshould be undertaken. The Task Force reaffirmedthe importance of the newdirection o fthe 2002 Water Law emphasizing mechanisms to enable open and free access to development planningproposals, economic and incentive measuresthat include social and economic costs, and including the value o f ecosystem services including `environmental pricing. Importantly, it called for examinationo fwater pricing as a perverse subsidy usingcase studies. 3.25 In2003 theNDRC andMWRjointly issuedregulations to increasewater prices for industryandwater supplies and, to a lesser extent, those for agriculture. The "Outline o f theNational Water Saving Plan" was endorsed by the Ministryof Constructionthat stated that all new andreconstructed buildingwould be requiredto install water-saving devices by 2010 and that excessive consumers o fwater would be subject to higherwater tariffs."' The longer-term objective by 2010 i s to supply through water saving halfthe incremental industrial demand and most o fthe incremental demand inirrigation. The water-saving plan was officially released inFebruary 2O07.l2OHow far this plantook account o f the CCICED recommendations i s unclear but it clearly resonateswith them and i s inharmony on many issues. China Moves toward a Water-Saving Society, Slowly 3.26 As indicated, Chinahas increasingly emphasized the imperativeto move towards greater and more efficient water use. To this end the MWRstarted implementingwater saving pilots over the period 2001 to 2010 to determine best practice. The first phase focused on areas with serious water shortage and water pollution problems suitable for the demonstrative purpose. The secondphase will establish water conservation demonstrative areas inselectedprovinces and river basins. According to MWRbetween 1998 and 2004 US$2 billionwas allocated by central government for physicalmodernization and rehabilitationo f large-scale irrigation systems and since then annual investmenthas risen to about US$1.5 billion.121 3.27 Usingpricingto induce conservation has beenonly modestly effective. As this assessmentshows this i s mainly due to lack o f enforcement at provincial level. Even thoughthe MWR and Ministry of Finance issued guidelines in 1988 that requiredwater institutionsto become financially autonomous by raising water fees over the next 5-10 years to recover all costs, they never approached full cost recovery.122 The Xiaolangdi experience i s a prime example. Nationally average tariffs for water supply increased from 118. CCICED Task Force and WWF. PromotingIntegratedRiver Basin Management and Restoring China's LivingRivers. Beijing. October 2004. http://www.harbour.s~.cddlam/04riverbasin%2Orpt.htm 119. Xinhua News Agency. 2003. China to Speed up Urban Water Price Reform. August 19. 120. Worldwatch Institute.2007. China Sets water-Saving Goal to Tackle Looming Water Crisis. Feb 27. 121. Li,Yuanhua. 2006. Water Saving Irrigation inChina. Irrigation and Drainage. 55: 327-336. 122. World Bank. 1993. China -yellow RiverInvestmentPlanning Study. 2 vols. Report No. 11146-CHA.  75 0.028 Yuan per cubic meter in2000 to 0.06 Yuan per cubic meter in2005. A major constraint inthe agricultural sector i s that water charges are frequently included within local landand agricultural taxes and there i s no direct connectionbetweenwater fees, quality o f service andwater use. Thus there are few incentives for farmers to conserve water.'23, 124 Among the more positive signsthat things are changing there are now over 20,000 water user associations that are co-managing irrigation water at the local level. However, as this evaluation and other assessmentshave shown, the role and effectiveness o f WUAs i s very uneven125 3.28 Evenso, the researchby Wang and others showed that the mere fact o f shifting management from the collective to either a WUA systemor contracting didnot lead to water savings unless managers were givenpositive incentives. Whenthere were incentives water use per hectare was reduced by about 40 percent.'26 The incentives apparently also improvedthe efficiency of the irrigation systems: there was no change inthe output o f major crops, such as rice and maize, andrural incomes and poverty remained statistically unchanged. Conversely, water reform ledto a decline inwheat production because it dependedmainly on irrigation. Ona larger scale this would significantly challenge China's foodgrain self-sufficiency policy -a constraint that could be relievedby importing wheat.'27 An IncreasedFocus on Water Savings 3.29 Untilthe mid-1990sthe primaryobjective was to improvethe irrigationefficiency o fmajor physical infrastructure usingwell-triedmethods including canal lining and better water control and scheduling that now cover a third o f all irrigation. Water saved was typically usedto extendthe area or intensity o f irrigation. Subsequently, muchmore attention i s now paid to water saving technology. Nationally 10percent i s equi edwith low pressure pipes and 6 percent with sprinkler, drip or other micro-irrigation. Rf: 3.30 Improved agronomic and land management practices have also ledto water saving on over 3.5 millionha of rice fields.129Aerobic rice, for example, i s now grown on 140,000 123. IPIM. 2002. Sixth International Forum o f Participatory Irrigation Management, Heldby the Ministryo f Water Resourcesand the World Bank, Beijing, April 21-26,2002. Several papers at this conference review the Chinesereform experience: China Irrigation District Association. "Participatory IrrigationManagement: Management Pattern Reform o f State-owned Irrigation District; Management Authority o f Shaoshan " Irrigation District. "Positively Promoting Reform Based on Practices o f Irrigation District, Obtaining Achievement o f Both Management and Efficiency." 124. Ma, Z. "Deepening Reform o fFarmer Managed Irrigation System, Promoting Sustainable Development o f Irrigation District." Participatory Irrigation Management: Innovation and Development ofIrrigation System. L.Nian eds. Beijing, China: China Water Resourcesand Hydropower Publishing House, 2001. 125. Wang, Jinxia, Zhigang Xu, Jikun Huang And Scott Rozelle. 2005 op cit. 126. Typically the incentive structure allowed the system managerto keep the difference inprice between the standardquantity o fwater allocated and the quantity o fwater actually used. 127.http://www.fao.org/AG/aGL/aglw/cropwater/wheat.stm#methods.The figure given assumes that the overall irrigation efficiency is about 50%. 128. MWR.2006. 2005 Statistics Bulletinon Water Activities. 129. RegassaE. Namara, Intizar Hussain,Deborah Bossio and Shilp Verma. 2007. Innovative land and water managementapproaches inAsia: productivity impacts, adoption prospects and poverty outreach. Irrigation and Drainage Volume 56, Issue 2-3 ,Pages335 -348.  76 ha innorthernChina facilitated b the availability o f efficient herbicides, seed-coating technologies andmechanization''. Water use was about 60 percent less than that o f traditional lowlandrice grown instandingwater, total water productivity 1.6 to 1.9 times higher, andnet returns to water use 2 times higher.13'Use o fplastic tunnels, dry-tillage, and greenhouses has grown five-fold since 1995 to over 2.5 million ha in2004.132,133 3.3 1 Therecently completed Bank-assisted Water Conservation Project innorth-eastern China was highly successful indemonstrating that water consumptioncan be reduced while farmer's income increased.134 It also showed that workable institutional arrangements can be implementedto reduce the longterm depletiono fgroundwater resources. It successfully and explicitly mainstreamed the concept o f increasing value-added per unit o f evapotranspiration. This shiftedan earlier Bank focus away from promoting only crop diversification to promoting improved cropping, agronomy andthe volumetric measurement o f water use. Generally, higher incrementalincomes were achieved with the more advanced irrigation applicationtechnologies. 3.32 The lessonsonhow to conserve water utilized for agricultural productionare being scaled-up incentral government programs: the National Water Conservation Program financed by the Ministry o fFinance's State Office for Comprehensive Agriculture Development and the National Irrigated Agriculture Water Saving Program by the Ministry o f Water Resources. Local government programs for water conservation have been also initiatedfollowing the precedent set by the project. For example, the preparation o f county-level groundwater management plans, piloted infour counties underthe project, has beentaken upby a further 21 counties acrossNorthern China. Prosperous urbanareas (e.g. Beijing and Qingdao) are also planning to invest both inwater saving, with aview to urban water supply, and inthe "harmonious society" approach to balancingrural and urban development and incomes. 3.33 The more liberal Chinese approach has allowed provinces andtownships to experiment with innovative ways o fmanaging water particularly inthose provinces where internal demand exceeds the quota. Successful innovations are endorsed by the MWR (Box 1) butthey are not mainstreamed nationally innewpolicy initiatives. 3.34 The most notable successat water conservation has involved five water rights transfer inNingxiaand Inner Mongolia and was the first time the principles o f a water right 130.Also known as AWDI - alternatewetting and drying or intermittent irrigation. 131. Wang Huaqi, B.A.M.Bouman, DuleZhao, Wang Changgui, and P.F. Moya. 2002. Aerobic rice in northern China: opportunities and challenges.Proceedings o f a Thematic Workshop on Water-Wise Rice Production, 8-11 April 2002 at IRRIheadquartersinLos Bafios, Philippines. 132. Zhou Xingxiang, RenjieDong*, Shujun Li, Gaojun Peng, Lanfang Zhang Jicong Hou, Junhua Xiao, And Benhai Zhu. 2003. "Agricultural Engineering inChina". Agricultural EngineeringInternational: the CIGR Journal o f Scientific Researchand Development. August 2003. 133. Zhang, Z.B. 2005. ShadingNet Application inProtected Vegetable Production inChina. Acta Horticulturae 719. 134. China Water Conservation Project (IBRD-4589.) 2000-06. Report No. ICR0000191. March27,2007.  77 transfer were officially approved inChina.135 More importantly they recognizedthat these rightshad avalue and set up official procedures for approval andthe determination o fthe price of water rightstransfer.'36Precedingthe rights transfer inthe Yellow RiverBasin, the first water trade took place inZhenjiangProvince inthe southern part o f the Yangtze River delta.'37These trades also highlightthe fact that the highwater usage o fthese industrial facilities- a result of outdated technologies -offers a second opportunity to reduce water consumption once new standards are established and regulated. This i s true o f other sectors too (Table 14). Box 3: Inner MongoliaAutonomous Region's 2000 Water Trade A new coal powerplant couldnot obtainsufficient coolingwater becauseof its own highdemandexceeded the province's quotaoftheYellow River runoff.To resolvethis issuetheplant, with the help ofthe province andlocal government,investedabout US$11millionto increasewater useefficiencyinadjacentirrigation districts. Inreturnfor its investment,the governmentgrantedthe plantan annualwithdrawal right of 50 millioncubic metersofwater fromthe Yellow River-the amount ofwater saved.The plant foundthe trade considerablycheaperthanadoptingthe alternativeofusingair-coolingtechnology to reducedemand. Since then the two AutonomousRegionsofInnerMongoliaandNingxiahave sponsoredfour similar water trades. Thesewater trades were endorsedbythe MWR in2005. Source: Wang, Yahua.2000. Table 14: Water Use in China is Not Efficient Indicator Unit Developed China Country Agriculturalwater share oftotalwater use YO 9 64 69 Agriculturalwater useefficiency LL 70 80 -- 45 Proportionof industrialwater recycled LL 80 45 Water usefor steel products(m3/ton) LL 6 23-56 Unaccountedfor water -Urban supplies LL 12-25 >20 Shareof urbanwastewater treated LL 80 90 - 30 Productivityofwater as share of GDP Total Economy GDP%per m3 14-48 2 Agriculture Sector " 1.4-5.8 0.5 Industrialsector " 8-100 4.2 Service Sector " 27-175 12.6 Source: Wang, Y. 2005. Improvedlandmanagementhas been successful-but sustainable? 3.35 Since the 1970s the objective o f landmanagement has changed from a primarily productive focus to one inwhich it i s recognized that they also provide environmental services -particularly for water management and reduction o f desertification. According to MWR, bythe end o f2005 almost one millionha o ferosion mitigation hadbeen implementedincluding 0.4 millionha ofwatershed improvementand 0.6 millionha o f 135.In2004 MWRmadethe followingrules: "The DirectiveofWater RightTransfer Pilot Project in Mainstreamofthe Yellow River inNingxiaandInnerMongolia" and"Regulation on Yellow RiverWater RightTransfer (Tentative)." The five pilot projectsincluded3 irrigationschemes. 136.SuQin.2006. System& PracticeofYellow River Water RightsTransfer. http://www.mwr.gov.cn/eng1ish1/20060110/20060 110104434EVNCBZ.pdf 137.Wang, Yahua. 2005. op cit.,  78 ecological restoration. 138 But MWR i s now a small player because o f the large areas o f steeply sloping land afforested under Sloping Land ConversionProgramthat i s designedto convert 14.67 millionhectares o f cropland (including a target o f afforesting4.4 million ha o f land with gradients o f more than 25 degrees) in23 provinces to forest by 2010 and has an overall budget o fmore than US$40 billion. Inaddition, it had a soft goal o f afforesting an equivalent area o fwasteland. By 2004 SLCP had converted 7.2 million ha o f cropland and afforested 4.9 millionha o f~aste1and.l~'Ifsuccessful these programs are likely to have a very positive impact on China's landscape, natural environment and water resources. 3.36 Independentreviewso fthe implementationexperience, however, have raised a number ofissues affectingthe sustainability o fSLCP. There is also aneedto fine-tune the environmental objectives to the differing regional climates -plantingtrees rather then grass inLoess Plateauregionmay beefficient for carbon sequestrationbutmay not bethe best option considering water resource shortages.140Coordination betweenthe State Forest Administration and MWR on watershed management i s problematic. Although the subsidy i s directly financed by the central government, not all village, township and county governments are ina position to finance local costs (i.e. mediationbetweencentral agencies and farmers, inspections) particularly as productionbenefits on converted land are free o f all taxes -this could be an issue giventhat the overall survival rate o f trees i s about 75 percent, and perhaps on1 70 percent inthe Yellow RiverBasin. Fieldevidence indicates it may be lower.147 3.37 There are also concerns that short-term subsidiesmay not cover lost income in some areas while converted lands mature and yield returns. And about a third o f farmers have indicated that shouldthe subsidies cease they will take back the land into c~ltivation.'~~One way o f mitigating this risk would be to increase off-farm rural employment as it would reduce the pressure on fragile land resources. Additionally, there have beenreports that local governments are taking their costs from the subsidy, thus reducing incentives for farmers to participate. Farmersare reportedly also showing an increasing reluctance to "voluntarily" convert adjacent barrenlands (that are sometimes the worst eroded) because o f the highlabor requirement.Evenso, independentfield surveys found that highly sloping and low-quality landwas beingpredominantly targeted. Conversely, there i s a concern that disbursementpressures have ledto the inclusion o f land that should not be targeted, thus lowering program efficiency and increasing costs. 138. MWR. 2006. op cit.,. 139. Bennett,M.T. and J. Xu. 2005. China's Sloping lands Conversion program: Institutional Innovation or Businessas Usual? Workshop on "Payment for Environmental Services-Methods and Design inDeveloping Countries. Beijing. 140. Sherr, S.J., M.T.Bennett, M.Loughney and K.Canby. 2006. Developing Future Ecosystem Service Paymentsin China: Lessons Learned from International Experience. A report prepared for the China Council for the InternationalCooperation on Environment and Development Taskforce on Ecocompensation. 141. Xu, Z, R. Tao, Z. XU and M.Bennett.2005. China's Sloping Land Conversion: Does Expansion Equal Success? CCAP Working Paper. 142.Uchida, E., Z. Xu and J. Xu. 2005. Grain for Green: Cost effectiveness and Sustainability of China's Conservation Set-asideProgram. Land Economics May 2005. 81 (2): 247-264  79 3.38 Finally, there i s negligible monitoring and evaluation to measure and link the environmental benefits o fthe program to erosion control, sedimentation and water resources. This i s perhaps a casualty o fthe absence o fthe MWR from the SLCP program andthe multipledevelopment objectives. While China's State Forestry Administration see the programprimarily inecological terms, most o fthe provincial andcounty-level officials -andprojectbeneficiaries-seeitasacentrallysubsidizedpovertyalleviationprogram. Progresson essential reforms is beingoutpaced by growingwater demand 3.39 As a result ofthe rapidpace o furbanizationinChina, urbanand industrial water demandhas steadily grown at 5 percent since the 1980s and is projected to a~ce1erate.l~~ The driving force will be the more than doubling o furbanpopulations from 118 millionto about 260 millionbetween2002 and 2020. Income growth inboth rural and urban areas increasedper capita water consumption by more than a thirdover the period 1989-1997 and this i s likely to continue. Overall water demand inthe most water short areas o fNorth China inthe area covered by the Yellow (Huang), Hai and Huai river basins -the 3-H basins -will depend much uponthe growth o f GDP. As the recent GDP growth rates have exceededthe "high" scenario o f 8.5 percent envisaged by the investmentplanning studies and their 2002 updates (para 1.33), it i s projected that total water demandwill increase from an average o f 150billion cubic meters a year (bcm/year) in2000 to about 173 bcm/year in2020. Invery dry years this could exceed 210bcm/year. Against these demands the total water supplywill be 144bcm/year, o fwhich 73 percent will come from surface water. 3.40 The huge shortfall insupply is very significant as it will affect economic activity and the lives o f 324 millionpeople in 100major cities withintwelve Northern province^.'^^ Giventhat urban, industrial and domestic water supply are accorded ahigherpriority for water inthe nationalpolicy the agricultural sector will contract unless it can use water far more efficiently than at present.This will be difficult. Accordingly, it i s projectedthat because as much as 6 bcm/year o f water could be shiftedout o f agriculture, the area o f cropped land could contract by about 4.2 millionha. The overall impact o f water shortages inall sectors isprojected by 2020 to have anannual economic cost ofbetweenY 44 billion (US$5.4 billion) andY 70billion(US$ 8.5 billion.) These potentiallossescanbepartially alleviated by increasing water use efficiency, reuse o f and importing water. Inaddition to engineeringmeasures, water pricing would catalyze moves towards greater water use efficiency and reuse. 3.41 China has clearly chosen supply augmentation via the inter-basin South-North Water Transfer Scheme as the most obvious solution. Inpart this i s because it provides a guaranteed supply and the engineeringrequired, though expensive, i s well known, feasible and practical giventhe highquality o f China's organizational and engineeringskills demonstrated inthe assessedprojects. Intests o fthe various policy options it was found 143. Gunaratnam, D.2004. Chinawater ResourcesIssues and Strategy. Issuespaperpreparedfor the World Bank's Asia andPacific Region. 144. Hebai, Henan, Shandong, Anhui, Jiansu, Shanxi, Shaanxi, InnerMonglia, Gansu and a small part o f Sichuan andthe municipalitiesofBeijingandTianin.  80 that an overall increase inwater use efficiency o f 10percent would reduce economic losses in2020 byabout US$ 1billionayear. Includingreuse andthe fullimpact o fwater price increase would leadto overall reductions o f about US$3.5 billion a year. The reduction o f economic losses causedby the South- Northtransfer alone would be less, about US$3.1 billion a year. Obviously the water use option would not be cost free but it would be highly disaggregated and easy to implementas the intra-provincial water rightstransfer demonstrated. The only problem withthe institutional solutions i s the lack or resolve and uncertainty about how quickly it could make a difference. CONCLUSIONS AND LESSONS 3.42 Duringthe Xiaolangdi andTarimBasinProjectsthe Bankwas able to work with national-level organizations o f water policy issues giventhe cross-cutting nature o fthese enterprises. For the first and only time the Bank was inthe privilegedposition o f assisting MWR with the integrated planning ofthe Yellow RiverandTarimRiverBasins. This was importantbecausethe incremental impact o fthe projects the Bank supported on water resourcesmanagement could be modeledat the river basin scale andfully integrated with China's large portfolio o fpotentialprojects. Inboth cases, this approach indicatedthat the Bank- assistedprojects were better than project alternatives that were thus rejected. And in both cases the Bank-assisted projects put a large emphasis on saving water ina river basin context. 3.43 Reviewingthe post-1990 portfolio o fagriculturalwater projects supported bythe BankinNorthandWest Chinathere hasbeena clear trendfrom buildingnewirrigation that increasesbasin-wide water demand, to projects that seek to improve water use efficiency and conservation efforts. However, there have beensome strategic mistakes that should have beenavoided -the Guanghong IrrigationProject approved in 1999 effectively doubled water use for irrigation inthe mid-reaches o fthe Yellow River near Xian. At the same time the downstream Water Conservation Project approved in2001 was extremely effective inreducing overall water consumption and i s a prime example o fwhat the Bank should support. There are three lessons: The Bankneedsto elevateitsengagement in China's water resources policy discussions above the provinciallevelto emphasizethe perverseincentivesthat are created by extensive agricultural water subsidies. Subsidiesfor irrigation and landmanagementjeopardize water conservation effort, longer term sustainability o f irrigation and landreclamationprojects and water availability for towns, cities, industryand environmental management. Despite more than two decades o fBank partnership and several state-of-the art projects piloting new approachesthese tendto be enclave projects. While Chinese senior planners and politicians at the center acknowledge that pricing should play a role inensuring more efficient and rational allocation anduse o f water this i s still as at an early stage o f applicationand far from politically acceptable inthe provinces. Without such a change inthe near future water shortages and environmental consequences o fresource mismanagement will continue to grow to crisis proportions;  81 0Monitoring and Evaluation requires capacity building and reorientation to measureoutcomes and impacts.There is little virtue inimplementing development models for reclaiming degraded lands or for river basin management if impacts on welfare, incomes andthe economy cannot be accurately measured. While this is less o f a problem for physical achievements it is a major problem for measurement o f socio-economic outcomes and impacts. Muchmore attention has to be givento understandingthe appropriate counterfactual and ensuringunbiased sampling. There i s also an issue around selection o fthe most effective M&Emodel considering the outcomes expected and the cost o f the projects. Appropriate training to redressthe problems found inM&Ei s clearly indicated. Inaddition, more attention should be givento fully utilizing existing Chinese resources. The Ministry o f Finance has indicatedthat it has a national network o f monitoring that include socio-economic indicators and that consideration should be givento working with themto triangulate evaluative findings; and 0 Muchgreater care is required in dealingwith issues of attribution when estimatingex-post economic rates of return. Specifically this means clearer definition o fthe project counterfactual and a systematic accounting o fthe effects o f exogenous actors and investmenton project impact. Without such attention to these exogenous factors there i s a danger that estimated ERRSgive a false impression o f project impacts. This, inturn, may leadto incorrect lessons on development effectiveness and the efficacy o fpolicy, institutionaland engineeringmeasures utilized to achieve development objectives.   83 Annex A Annex A. BasicData Sheet LOESS PLATEAUWATERSHEDREHABILITATION PROJECT (LOAN 4477, CREDIT 3222) Key ProjectData (amounts in US$million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 252 Loan amount 150 Cofinancing 7.41 Cancellation CumulativeEstimatedand ActualDisbursements FYOO FYOl FY02 FY03 FY04 FY05 FY06 FY07 FY08 FYO9 Appraisal 37.5 75.0 112.5 142.0 150.0 150.0 150.0 150.0 150.0 150.0 estimate (US$M) Actual (US$M) 7.7 33.8 53.9 95.7 130.4 146.7 146.7 146.7 146.7 146.7 Actual as % of 20 45 48 67 87 98 98 98 98 98 appraisal Date of final disbursement: 12/1/05(DFID fund not completely disbursed) Project Dates Original Actual PCD 911011998 Appraisal 1/8/1999 Board approval 512511999 Effectiveness 911711999 911711999 MTR 9/20/2002 9/20/2002 Closing date 12/31/2004 613012005  Annex A 84 Staff InDuts Stage of project cycle Actual/Latest Estimate No. of staff weeks US$ (`000) Identification/preappraisal 55,008 Appraisalhegotiation 115,534 Supervision 363,214 ICR 13,516 Total 457,273 Mission Data Date No. of Specializations Implemen. Develop. (month/year) persons represented progress objective IdentificationlPreparation 0911997 7 A, 2AG, 2ES, WRS, F 1011998 6 AG, E, M, WRS, H, F AppraisallNegotiation 01I1999 6 AG, E, M, WRS, H, F Supervision 1111999 6 ES, OP, M, WRS, FM, AG S S 0912000 9 AG, 2E, 2M, ESD, FM, 2F S S 0912001 4 AG, E, WRS, F, PA S S 0912002 4 AG, E, WRS, F, PA S S 0912003 4 AG, E, F,OA S S 0812004 4 AG, E, F,OA S S ICR 0912005 3 E, OA, S HS HS Note: A = Agronomist; AG = Agriculturist, E = Economist; ES = Environmental Specialist, ESD = Engineer for the Safety of Dams, F = Forestry Specialist, FM = Financial Management Specialist; H = Horticultural Specialist; M = Monitoring Specialist, OA= OperationsAnalyst, OP = Operation Officer, PA = Program Assistant, P = Procurement Specialist, S = Social Specialist, WRS =Water Resources Specialist Other Project Data BorrowerlExecuting Agency: FOLLOW-ON OPERATIONS Operation Credit no. Amount Board date (US$ million) NONE  85 Annex A XIAOLANGDI MULTIPURPOSE I(LOAN PROJECT 3727-CHA) KeyProjectData (amounts in US$million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 2294.0 Loan amount 460.0 459.8 99 Cofinancing Cancellation 0.2 CumulativeEstimated and Actual Disbursements FY95 FY96 FY97 FY98 FY99 FYOO FYOI FY02 Appraisal estimate 115.0 265.0 400.0 457.0 460.0 460.0 460.0 460.0 (US$M) Actual (US$M) 124.4 221.8 360.7 458.9 459.6 459.7 459.7 459.7 Actual as % of 108 83 90 100 99 99 99 99 appraisal Date of final disbursement: 10/1812001 ProiectDates Original Actual PCD 7121I988 Appraisal 41811993 Board approval 4/14/1994 Effectiveness 912211994 912211994 MTR 6/30/1996 6/30/1996 Closing date 12/31/2000 12/31/2000 StaffInputs(staff weeks) Stage of project cycle ActuaMLatest Estimate No. of staff weeks US$ (`000) ldentificationlpreappraisal 130 650 Appraisal/negotiation 50 200 Supervision 64 460 ICR 8 25 Total 252 1305  Annex A 86 Mission Data Date No. of Specializations Implement. Develop. (month/year) qersons represented progress objective Identification/ Preparation 16 3 Economists,2 FMS, 1Water Resources Mission 1 Eng, 2 Environmental, 2 SocialScientists,2 GW and SW Specialists, 1 Procurement Spl, 1 Hydropower Eng, 2 lnst Specialist Mission 2 15 2 Economists,1FMS, 2 Water Resources Eng, 2 Environmental, 2 SocialScientists,2 GW and SW Specialists, 1 Procurement Spl, 1Hydropower Eng, 2 lnst Specialist Mission 3 13 1Economist, 1Cost Estimator,2 Water Resources,2 Environmental,2 Social Scientists,2 GW and SW Specialists,1 Procurement Spl, 1 Hydropower Eng, 2 lnst Specialist AppraisaIINegotiation Appraisal 16 2 Economists, 2 FMS, 1 Water ResourcesEng, 2 Environmental,2 Social Scientists, 2 GW and SW Specialists,1Procurement Spl, 1 Hydropower Eng, 2 lnst Specialist, 1 Legal Negotiations 3 1Water ResourceslDam,1 Legal, 1 Disbursements Supervision 1 2 1water resources/dam, 1geotechnical HS HS 3 1water resources/dam, 1geotechnical, HS HS training specialist 3 2 water resources/dam, 1dam HS S 4 1water resources, 1dam specialist, 1 HS S geotechnical,1 procurementlcostspecialist 3 1water resources, 1dam, 1 geotechnical HS S I 1water resources/ procurement HS S 5 1water resources, 1 dam, 1geotechnical, HS S lFMS, 1claimsspecialist 6 1 water resources, 1damkonstruction, 1 HS S dam, 1geotechnical, 1 FMS, 1 Environment 6 1water resources, 1damkonstruction, 1 HS S dam, 1geotechnical, 1 Financial SpecialistlAuditor,1 Environment 10 2 Water Resources/economist,1 HS S dam/constructionspecialist 11 10 1 water resources/economist, 1 undergrounc HS S works, 1geotechnical, 1dam, 1 dam safety, 1FMS, 1PrincipalFinancialSp1.,2 Environmental, 1Claims Spl. ICR ICR 5 1Water Resources/Economist, 1 HS S Finance/Economist, 1 Economist, 1 Environmental, 1 irrigation/water resources modeler 3W = Groundwater; SW = irface wa  87 Annex A Other ProjectData Borrower/Executing Agency: FOLLOW-ON OPERATIONS Operation Creditno. Amount Board date (US$ million) Xiaolangdi II L4200 430 6/24/1997 XIAOLANGDI MULTIPURPOSE I1(LOAN PROJECT 4200) KeyProject Data(amounts in US$million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 2855.8 Loan amount 430.0 309.5 72 Cofinancing Cancellation 80.5 Cumulative Estimatedand Actual Disbursements FY98 FY99 FYOO FYOl FY02 FY03 FY04 Appraisal estimate 80.0 210.0 340.0 410.0 430.0 430.0 430.0 (US$M) Actual (US$M) 7.2 110.0 195.4 234.9 306.5 349.4 351.5 Actual as % of 9 52 57 57 71 72 72 appraisal Date of final disbursement: 4/30/2004  Annex A 88 Proiect Dates ~ Original Actual PCD 2/28/1997 Appraisal 2/28/1997 Board approval 6/24/1997 Effectiveness 10/31/I997 1/12/1998 MTR 07/31/1999 611112002 Closing date 12/31/2003 12/3112003 Staff Inputs(staff weeks) Stage of project cycle Actual/Latest Estimate No. of staff weeks US$ (`000) Indentification/preappraisal 130 650 AppraisaVnegotiation 50 200 Supervision 64 460 ICR 8 25 Total 252 1305 Mission Data Date No. of Specializations Implemen. Develop. (month/year) persons represented progress objective IdentificationlPreparation 3/10/97 AppraisallNegotiation 03/11/1997 Mission Leaderwater Resources (I), Economist(5), Dam Specialist (3), ResettlementSpecialist(I), EnvironmentalSpecialist(3), FinancialSpecialist(2), Flood ForecastingSpecialist(I), SedimentSpecialist (I), Reservior Operations(1) Supervision 06/03/1998 8 Mission Leader(1); Dam Engineer(1); S HS FinancialAnalyst (1); Damflunnel Specialist(1); Geotech. Specialist(1); EnvironmentSpecialist (1); Hydraulics Specialist(1); Flood ForecastSpl. (1) 11/02/1998 6 Water ResourcesEng (1);Water S HS ResourcesSpec. (1); Damflunnel Specialist(1); Geotech.Specialist(1); EnvironmentSpecialist (1); Financial Analyst (1) 03/30/2000 10 Water ResourcesSpl (3); Claims S HS Specialist (1); FinancialAnalyst (2); Damflunnel Specialist (1); GeotechnicalSpl (1); Environmental SPl(2) 09/06/2001 5 Task Team Leader(1); Water S HS  89 Annex A Date No. of Specializations Implemen. Develop. (month/year) persons represented progress objective ResourcesSpcl. (1); Environmental Spcl. (1); FinancialSpecialist(1); Dam Specialist(1) 11/03/2002 6 TTL, Water Resources (1); Dam S HS Specilaist(1); Water Engineering(1); GeotechnicalEngineer (1); EnvironmentManagement (1); EconomicAnalysis (1) ICR 10/03/2003 6 Task Team Leader (1);Water S HS ResourcesSpcVEconomist.(1); EnvironmentalSpcl.(1); Financial Specialist(1); Dam SpecialistIModeller(1) And IrrigationSpecialist(1) Other Project Data Borrower/Executing Agency: FOLLOW-ON OPERATIONS Operation Credit no. Amount Board date (US$ million) NONE  Annex A 90 TARIMBASIN11PROJECT(LOAN AND CREDIT3093)4341 KeyProjectData(amounts in US$million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 272.6 Loanlcredit amount 150.0 146.7 97 Cofinancing Cancellation 2.6 Cumulative Estimatedand ActualDisbursements Appraisal estimate 20.1 57.3 98.8 127.5 143.8 150.0 (US$M) Actual (US$M) 12.9 23.9 49.5 74.1 105.9 139.5 Actual as % of 64 42 50 58 74 93 appraisal Date of final disbursement: 10/19/2003 ProjectDates Original Actual PCD 311911997 Appraisal 21911998 Board approval 61911998 Effectiveness 10/31/I998 12/18/1998 MTR 11/30/2001 Closing date 12/05/2004 12/31/2005 StaffInputs(staffweeks) Stage of project cycle Actual/Latest Estimate No. of staff weeks US$ (`000) Identification/preappraisal 495.38 AppraisaVnegotiation Supervision 696.53 ICR Total 1191.91  91 Annex A Mission Data ~ Date No. of Specializations Implemen. Develop. (monthlyear) persons represented progress objective IdentificationlPreparation 0911511996 3 IrrigationEconomist(1) Water ResourcesEngin(2) 11/25/1996 4 IrrigationEconomist(1);Water ResourcesEngin(1); Drainage Engineer(1); Anthropoligist(1) 03/25/1997 2 IrrigationEconomist(1);Water ResourcesEngin (1) 0511811997 I? Water Resour.Engin. (2); IrrigationEconomist(1); EnvironmentalSpec. (1); Economist(1); InstitutionallFinancial Spec. (1); Anthropologist;(1) RevierBasiin Mgt. Engineer (1) WaterAdministration/Sidd Spec. (1); FarmerOrg. Spec (1); DrainageEngin (1) AppraisallNegotiation 0211511998 8 Water Resour. Engineer(2); Water Resour. Management Spec. (1) IrrigationEngin. (1); Environ.Spec. (1); Operations Officer (1); DrainageEngineer(1); FinancialManagment Spec. (1) Supervision 11/I411998 7 Wat. Rs. Eng. (Co-Ttl) (2); Fin. Mgt. S S Spec. (1); ProcurementSpec. (2); Disbursement Spec. (1); Ipm Spec. (1) 05/07/1999 6 Co-TTLWat. Res. Eng. (2); S S Agri./Enviromentalist(1); Institutional Spec. (1);Wat. Res. Man. Spec. (1); Agri./lpm Spec. (1) 05/07/1999 4 Co-TTLWat. Res. Eng. (2); Economist S (1); InstitutionalSpec. (1) 05/07/1999 4 Co-TTLWat.Res.Eng. (2); Economist S (1); InstitutionalSpec. (1) 05/07/1999 3 Co-TTL (2); Procurement(1) S 10/27/2000 3 Co-TTL - Wat. Res. Eng (2); U Procurement Specialist(1) 06/08/2001 8 TTL (1); Economist(2); IrrigationSpec. S S (1);Water ResourcesSpec. (1); Social Spec. (1); Consultant(1); Interpreter (1) 11/30/2001 6 TTL (1);Water ResourcesSpec. (1); S S Economist(2); IrrigationSpec (1); SocialSpec (1) 06/22/2002 9 TTL -Wat. Res. Eng. (1);Wat. Res. S S Eng. (1); Economist(1); Fin. Man. Spec. (1); ProcurementSpec. (1); InstitutionalSpec. (1); Irrigation Engineer (1); RemoteSensing Spec. (2) 11/30/2002 5 Team Leader(1);Water Resources S S Engine(1); IrrigationEngineer(1); AgriculturalSpecialis (1); Procurement  Annex A 92 Date No. of Specializations Implemen. Develop. (month/year) persons represented progress objective Specialist (1) 11/14/2003 3 Co-'TTL Wr Engineer (1); Co-lTL HS HS Irrigation Spec (1); SiddNVua Spec (1) 1210412004 4 Water Resources Spec.(l); HS HS Irrigation Engin. (1); Social Devl. Spec. (1); Agri. Economist (1) ICR 0312005 4 Water Resour. Spec (1) Irrigation Spec. (2); HS HS Other Project Data Borrower/ExecutingAgency: FOLLOW-ON OPERATIONS Operation Credit no. Amount Board date (US$ million) NONE