Document of The World Bank

Report No: 18982-CHA

PROJECT APPRAISAL DOCUMENT

ON A

PROPOSED LOAN OF US$60 MILLION AND

A PROPOSED CREDIT OF SDR 73.8 MILLION (US$100 MILLION EQUIVALENT)

TO THE

PEOPLE'S REPUBLIC OF CHINA

FOR THE

WESTERN POVERTY REDUCTION PROJECT

June 1, 1999

Rural Development and Natural Resources Sector Unit East Asia and Pacific Region CURRENCY EQUIVALENTS

(Exchange Rate Effective April 1, 1999)

Currency Unit = Yuan (Y) Y1.00= US$0.12 US$1.00 = Y8.30

FISCAL YEAR

January 1 - December 31

GLOSSARY OF TERMS

Banner Mongolian equivalent of a County Sumu Mongolian equivalent of a Township Gacha Mongolian equivalent of a Village mu Measure of land: I mu=0.067 ha ha Hectare Mcm Million cubic meters Bcm Billion cubic meters

Vice President: Jean-Michel Severino Country Director: Yukon Huang Sector Manager: Geoffrey B. Fox Task Team Leader: Petros Aklilu ABBREVIATIONS AND ACRONYMS

ABC Agricultural Bank of China O&M Operation and Maintenance ACWF All China Women's Federation PAB Provincial Audit Bureau CPADO County Poverty Alleviation Office PADO Poverty Alleviation and Development Office CAS Country Assistance Strategy PEdC Provincial Education Committee CCICED China Council for International PFB Provincial Finance Bureau Cooperation on Environment and Development CIF Cost, Insurance and Freight PIC Public Information Center CSA Counterpart Special Account PIP Project Implementation Plan D&E Demonstration and Extension PLB Provincial Labor Bureau EIA Environmental Impact Assessment PLG Project Leading Group EPA Environmental Protection Agency PMO Project Management Office EPB Environmental Protection Bureau PP Project Province ERR Economic Rate of Return PPEO Provincial Project Execution Office (Qinghai) FAD Financial and Accounting Division PPMO Provincial Project Management Office FAO Food And Agriculture Organization Qinba Qinba Mountains Poverty Reduction Project FCPMC Foreign Capital Project Management RAP Resettlement Action Plan Center FMSM Financial Management System Manual REB Rural Enterprise Bureau FRR Financial Rate of Return SA Social Assessment FSR Farming System Research SAB State Audit Bureau GMS Grassland Monitoring Station SDR Special Drawing Right ICB International Competitive Bidding SOE Statement of Expenditure ICLS Integrated Crop/Livestock System SSB State Statistical Bureau ICR Implementation Completion Report STW Shallow Tube Well IDA International Development Association SU Sheep Unit IFAD International Fund for Agricultural SW Shallow Well Development IMRG Inner Mongolia Regional Government SWPRP Southwest Poverty Reduction Project LGPRD Leading Group for Poverty Reduction and TAG Technical Advisory Group Development of the State Council LMDC Labor Market Development Center TW Tube Well LMIS Labor Market Information System TWS Township Work Station MBD Model Bidding Document UNDP United Nations Development Program M&E Monitoring and Evaluation VSIP Voluntary Settlement Implementation Plan MED Monitoring, Evaluation and Dissemination VPIG Village Project Implementation Group MOE Ministry of Education WBO World Bank Office (Gansu) MOF Ministry of Finance WBLEC World Bank Loan Execution Committee (Gansu) MOU Memorandum of Understanding WCB Water Conservancy Bureau NCB National Competitive Bidding WFP World Food Program NPV Net Present Value XB Xiangride-Balong OD Operational Directive XRWCA Xiangride River Water Conservancy Administration (Qinghai) CHINA WESTERN POVERTY REDUCTION PROJECT

CONTENTS

Page A. Project Development Objective

1. Project development objective...... 3 2. Key performance indicators...... 3

B. Strategic Context

1. Sector-related Country Assistance Strategy (CAS) goal supported by the Project...... 3 2. Main sector issues and Government strategy ...... 4 3. Sector issues to be addressed by the Project and strategic choices ...... 4

C. Project Description Summary

1. Project components...... 6 2. Key policy and institutional reforms supported by the Project ...... 7 3. Benefits and target population ...... 8 4. Institutional and implementation arrangements...... 9

D. Project Rationale

1. Project alternatives considered and reasons for rejection ...... 17 2. Major related projects financed by the Bank and other development agencies ...... 18 3. Lessons learned and reflected in the Project design ...... 18 4. Indications of borrower commitment and ownership ...... 19 5 . Value added of Bank support in this Project ...... 19

E. Summary Project Analyses

1. Economic ...... 20 2. Financial ...... 20 3. Technical ...... 20 4. Institutional ...... 21 5. Social ...... 22 6. Environmental assessment ...... 28 7. Participatory approach...... 3-

F. Sustainability and Risks

1. Sustainability ...... 34 2. Critical risks ...... 35 G. Main Loan conditions

1. Effectiveness conditions ...... 38 2. Other ...... 38

H. Readiness for Implementation ...... 41

I. Compliance with Bank Policies...... 42

Annexes

Annex 1: Project Design Summary .43 Annex 2: Detailed Project Description .48 Annex 3: Estimated Project Costs .81 Annex 4: Cost Benefit Analysis Summary .84 Annex 5: Financing Plan .93 Annex 6: Procurement and Disbursement Arrangements .94 Annex 7: Project Processing Schedule .99 Annex 8: Documents in Project File .100 Annex 9: Social Aspects .101 Annex 10: Environmental Aspects .113 Annex 11: Statement of Loans and Credits .136 Annex 12: CountryataGlance .139

Maps

Qinghai Province No. IBRD 29993 Gansu Province No. IBRD 29994 Inner Mongolia Autonomous Region No. IBRD 29995 CHINA

Western Poverty Reduction Project

Project Appraisal Document

East Asia and Pacific Region

Date: June 1, 1999 Team Leader: Petros Aklilu Country Director: Yukon Huang Sector Manager: Geoffrey B. Fox Project ID: CN-PE-46564 Sectors: Al - Irrigation & Drainage, AL - Livestock, AM - Agro-Industry & Marketing, AQ - Annual Crops, and TR - Rural Roads Lending Instrument: Specific Investment Loan (SIL) Theme: Community-Based Development Poverty Targeted Intervention: Yes Project Financing Data Loan [X] Credit [X] Grant [] Guarantee [] Other (Specify) [] For Loans/Credits/Others: Amount (US$M/SDRM) Loan of $60/Credit of SDR 73.8 ($100 equivalent) Proposed Terms: []To be defined []Multicurrency [X] Single currency []Standard variable []Fixed [X] LIBOR-based Grace period (years): 5 Loan/10 Credit Years to maturity: 20 Loan/35 Credit Commitment fee: 0.75% Loan/0.5% Credit Service charge 0.75% Credit Front end fee on Bank loan 1.00% of Loan 4E^~ }=gi=) A-KWa""='VE'- MI In M 9§ 0Z Source Local Foreign Total Government 118.00 5.70 123.70 IBRD 0.00 60.00 60.00 IDA 69.40 30.60 100.00 Beneficiaries 28.00 0.00 28.00

Total: 215.40 96.30 311.70 Borrower: People's Republic of China

Responsible agency: Leading Group for Poverty Reduction and Development of the State Council and Provincial/Regional Governments of Gansu and Qinghai and Inner Mongolia Autonomous Region.

* See detailedfinancing plan in Annex5 2

Implementing agencies: Gansu Provincial Project Management Office Address: 326 Wudu Road, Lanzhou, Gansu Province, China, 730030 Contact Person: Mr. Yang Zixing Tel: (0931) 8853628 Fax: (0931) 8853558 Email: yangzx(lz.gs.cninfo.net

Inner Mongolia Project Management Office Address: East Wulanchabu Road, Hohot, Inner Mongolia, China, 010010 Contact Person: Mr. Xu Jianxin Tel: (0471) 4929396 Fax: (0471)4929396 Email: sjyhdk@,public.hh.nm.cn

Qinghai Provincial Project Management Office Address: 334 Qiyi Road, Xining, Qinghai, China, 810001 Contact Person: Mr. Gao Guien Tel: (0971) 8238426 Fax: (0971) 8256567 Email: fpwzxmb(q),public.xn.gh.cn

Estimated disbursements (Bank FY/US$M): 2000 2001 2002 2003 2004 2005 2006 Annual 16.60 21.50 38.30 31.00 25.80 16.40 10.40 Cumulative 16.60 38.10 76.40 107.40 133.20 149.60 160.00 Project implementation period: 1999-2005 Expected effectiveness date: 09/30/99 Expected closing date: 06/30/2006 OCS PAD Form:October 9, 1998 3

A: ProjectDevelopment Objective

1. Projectdevelopment objective: (see Annex 1) The objective of the proposed Western Poverty ReductionProject (the Project) is to reduce the incidenceof absolutepoverty in remote and inaccessiblevillages of Gansu and QinghaiProvinces and Inner MongoliaAutonomous Region (hereinafterreferred to as Project Provinces- PPs). The Project would empower poor households in the selected poor counties to raise their incomes, through increasedgrain and livestockproduction sufficientto meet the basic needs of food and clothing and, in many cases, also generate a marketable surplus to improve living standards. Establishmentof alternative income generation activities in poor rural areas, improvement of rural infrastructureand enhanced access to basic social services such as health and education would also increase the living standards of poor rural households.The Project's objective fully reflects the expectations and aspirations of the target population, as reflected in the Social Assessments(SAs).

2. Key performanceindicators: (see Annex 1) Key performanceindicators to monitor the overall CountryAssistance Strategy (CAS) objective, the Project's developmentobjective, project outputs, and components,are presented in Annex 1. Verifiable indicators,including the percentage of populationbelow the national poverty line, crop and livestock yield increases, grazing land management, feed balances, per capita grain production, school enrollment, visits to health centers, jobs created, etc., would be monitored during project implementation. A baseline survey has been prepared as part of project preparation.

B: StrategicContext

1. Sector-related Country Assistance Strategy (CAS) goal supported by the Project: (see Annex 1)

Documentnumber: 16321-CHA Date of latest CAS discussion:05/28/98 The Project is the fourth in a series of Bank-assisted,targeted poverty reduction operations in China, supportingthe Government's Eight-SevenPoverty Reduction Plan (the Plan) and similar to 31 other poverty orientedprojects in China. One of the five pillars of the CAS for China is the promotionof human resourcesdevelopment. Specifically, it calls for, inter alia, targeted poverty investments; support for the Government's universal primary education goal, with particular focus on access for girls and minority nationalities; improved maternal and child health; and disease prevention. The Project is fully consistent with the CAS poverty alleviation objectives, supporting selected investrnents in environmentally sustainable agricultural and livestock development,provision of basic rural infrastructureand enhancedaccess to basic educationand health services in the poorest regions in western China, where the incidence of poverty is the highest. Supportfor the Project is also in line with the Bank's strategy of shifting resourcesto China's interior as well as addressingthe issues of growing urban unemployment,resulting from reforms of state-ownedenterprise, and limited opportunitiesfor labor mobility to the developed coastalregions.

The Project would also promote the Bank's overall goal of poverty reductionas articulatedin the "Rural Developmentfrom Vision to Action" under which China has been identified as one of the focal countries. 4

2. Main sector issues and Government strategy: Main sector issues. The absolute poor in rural China, comprising disproportionately of minority nationalities, reside mainly in remote and inaccessible upland areas, where the endowment of natural resources is poor and population pressures high (China. Strategies for Reducing Poverty in the 1990s, World Bank, 1992). Government's estimates of the incidence of absolute poverty demonstrate a significant drop, from 270 million people in 1978 to about 60 million people in 1997. However, China's growth pattern has led to rising income disparities between the resource poor interior and western regions on the one hand, aLndthe more affluent and rapidly growing coastal areas on the other (China 2020 - Sharing Rising Incomes, World Bank, 1997). An unfavorable agro-ecological environment, degraded lands, low precipitation with recurrent droughts, fragmented and uneconomic land holdings combined with slow human resource and institutional development, constrain the prospects for increased agricultural productivity and income in the upland areas.

Government strategy. In 1994, China launched a far-reaching poverty reduction strategy under the 8-7 Plan, which targets 592 designated poverty counties located in inaccessible mountainous and semi-arid areas in the central and western parts of China. It's objective is to raise per capita incomes to Yuan 500 (in constant 1990 prices) for about 80 million poor living in abject poverty, within seven years. To achieve this objective, the Plan calls for close collaboration among the state, local governments, communities and charities, with assistance from external aid agencies, aimed at inter alia: i) creating suitable conditions for agricultural development and promoting related investments in off-farm income generating activities; ii) improving basic rural infrastructure such as roads, village electrification and rural water supply; and iii) investing in basic social services including education and health. Toward this end, a number of initiatives are being implemented with promising results, including Bank-assisted targeted rural development projects-Southwest Poverty Reduction Project (S'WPRP) and Qinba Mountains Poverty Reduction Project (Qinba), voluntary settlement on newly irrigated land (Gansu-Hexi Corridor Project), redeployment of surplus labor to non-agricultural activities (labor mobility), etc.

Impact of Government strategy. The impact of the 8-7 Plan has been significant. Since it's adoption, poverty has fallen markedly according to China's own estimates. In September 1996, a conference jointly organized by the Communist Party of China Central Committee and the State Council to review the implementation progress of ithe Plan and decided to: i) reconfirm the country's determination to meet the basic food and clothing requirements of every citizen by end of the 20th century; ii) establish the rural poor household as the focal point for poverty reduction interventions; iii) raise the annual funding of the Plan to Yuan 20 billion with obligatory local counterpart funds equal to 30-50 percent of the Poverty Fund allocation; iv) assign responsibility to the developed coastal and other cities and provinces to assist poor provinces; v) introduce policy incentives including tax exemption for enterprise investors in designated poverty counties, and reduction or exemption of agricultural tax and the special tax for poor households; and vi) adopt the Bank assisted integrated rural development approach to poverty reduction as a point of reference for formulating future similar operations in the future. These decisions mark another milestone in China's effort to reduce absolute poverty,

3. Sector issues to be addressed by the Project and strategic choices: Notwithstanding the progress to-date, sector specific interventions (agriculture, rural infrastructure, social sectors) have had a limited impact on the lives of the absolute poor in remote, mountainous, resource deficient areas. Improving their lot will require addressing a series of sectoral issues in an integrated fashion, as addressed below. 5

(i) Low agricultural productivity. For the target population, fragile agricultural and pastoral lands are the main source of livelihood. In Qinghai, crop yields are low due to poor soil quality, recurrent drought, and lack of productivity enhancing technology. Likewise, in Gansu and Inner Mongolia crop yields are low due to a comparatively harsh agro-ecological environment (poor soils, steep topography, recurrent drought, short growing period, significant grassland degradation), lack of access to productivity enhancing technologies and working capital for higher value seeds for crops and forages. Hence, continued cultivation and/or grazing without appropriate intervention is likely to further accelerate the degradation of the agro-ecological resource base in all three PPs.

The Project, by introducing productivity enhancing and environmentally sustainable crop and livestock development packages aims to progressively shift the focus from a food self-sufficiency system to one that stresses food security while maximizing cash income. The proposed technology packages are widely accepted and successfully adopted in China and elsewhere in previous Bank operations. Provision of these technology packages would be combined with selected basic rural infrastructure and social services, access to adequate and timely delivery of production inputs, supplemental irrigation, and agricultural support services. As productivity and incomes increase, steeper slope land and other areas subject to severe wind erosion and not suitable for sustainable crop production, would be gradually retired from cultivation and sown with forage legumes and grasses, or protective trees. The attendant improvements in soil water conservation and progressive reduction of soil erosion (wind, sheet, gully) would, over time, contribute to raising sustainable crop and forage production. In these development packages, improving the nutrition and productivity of existing livestock is a first priority. Increased livestock output would encourage cut-and-carry feeding of penned ruminant (cattle, sheep, goats) and improve non-ruminant (pigs, poultry, rabbits) integrated crop/livestock production systems.

(ii) Growing surplus labor. The surplus rural labor in China is estimated at about 30 percent. With limited access to off-farm employment opportunities and low skill levels, the poor face an enormous challenge to break the cycle of poverty. Employment opportunities in the urban sector are steadily declining as more workers are laid-off from state enterprises. The Project would help spur demand for labor through support for crop and livestock production and investments in small-scale rural enterprises and selected labor mobility programs. Cheap labor, a growing potential market, and rural infrastructure development combined with policy incentives is also likely to attract investors under the East-West Co-operation Program. Beijing and Tianjin Municipalities and Liaoning Province are designated to assist Inner Mongolia, Gansu and Qinghai, respectively.

(iii) Access to basic social services. A sample survey conducted in the PPs during project identification indicated relatively high enrollment rates (average 87 percent) in primary schools. However, pockets of low enrollment and high drop-out rates still exist due to the inability of households to pay school fees and other charges. Similarly, lack of access to basic health services-absence of clinics in close proximity and relatively high cost of services are resulting in poor maternal and child health. During preparation, delivery options for basic social services were carefuillyreviewed and in order to minimize complexity and duplication, the interventions to be supported under the Project were designed to supplement ongoing investments and programs in the pipeline. These ongoing activities include the proposed basic education project in Gansu, to be supported by the European Union; a special fund for student assistance, targeting Project villages established by the Gansu Goverrnent; and in Inner Mongolia, where student enrollment is comparatively high, a program for additional student assistance to encourage enrollment of out- of-school children. The social sector component of the Project would provide education and health services in and around Qinghai's new settlement area, and the 'move-out' counties and basic health services in Gansu and Inner Mongolia. 6

(iv) Targeting the poor. As noted, 592 counties were targeted for poverty reduction programs under the 8-7 Plan, based on per capita grain production of less than 200 kg and per capita income of less than Yuan 300 in 1986. While the poverty county programs have resulted in significant transfer of resources, China has recognized that integrated interventions being piloted under the SWPRP and replicated in Qinba, best serve the targeted population. The annual domestic resource transfer for poverty alleviation has reached Yuan 20 billion. By China's own account, however, the feeding and clothing targets of the Plan have yet to be met for the estimated 58 million living in absolute poverty. The Project, building on the lessons learned, from the SWPRP, Qinba and Gansu Hexi Corridor operations, recognizes that the presence of strong national, provincial and local institutions, and active participation of target beneficiaries, will be critical for successfully targeting the poor.

C: ProjectDescription Summary

1. Project components: (see Annex 2 for a detailed description and Annex 3 for a detailed cost breakdown) To achieve its development objective, the Project would support:

(i) Land and Household Development. Provision of improved agricultural and livestock technology packages-seeds, fertilizer, insecticides, plastic mulches, breeding/fattening stock; forestry development; and upgrading agricultural and livestock support services.

(ii) Irrigation and Land Improvement. Construction of a new dam1, renovation of an existing dam and construction of irrigation and drainage systems on 19,200 hectares in Qinghai; and construction of small scale irrigation works on 34,050Dhectares, terracing of 26,500 hectares of sloping land and soil and water conservation works in Gansu and Inner Mongolia.

(iii) Rural Infrastructure. Construction of 216 kilometers of Class III and IV rural roads to provide access to 289 remote administrative villages; construction of drinking water supply facilities to serve 370,000 people in 760 administrative villages; and extension of electric power lines to 410 administrative villages. (iv) Rural Enterprises. Provision of credit to establish non-state owned and household-based rural enterprises for construction of food and fiber processing facilities based on agreed criteria. (v) Labor Mobility (Gansu). Organized and voluntaiy placement of about 65,700 surplus rural laborers in off-farm employment in and outside Gansu and support to Gansu's Labor Bureau to improve the management and referral services. Assurances were obtained during negotiations that the Borrower would ensure that all regional governments employing surplus labor under the Gansu Labor Mobility Program, implement the program in a manner designed to ensure safe and equitable working and living conditions for the laborers (page38(i)). (vi) Voluntary Settlement (Qinghai). Voluntary seltlement of an estimated 57,800 living in absolute poverty in degraded mountainous areas of eastern Qinghai and of 4,000 local herders and farmers living around the settlement area. Start-up costs of settlement, including grants for transport, housing, fuel, and initial subsistence expenses, would be financed out of counterpart funds; rehabilitation of 85 percent and terracing of 15 percent of the land to be vacated would be financed by the Project counties. (vii) Social Sectors - Education and Health. Construction and upgrading of basic education and health facilities in Qinghai's new settlement area and in the 'move-out' counties; and upgrading services of selected basic health clinics in Gansu and Inner Mongolia. Assurances were obtained

1The dam is an off-riverreservoir which will inundateapproximately 128 hectaresof unuseddesert land to an averagedepth of about 15 meters. 7

during negotiations that PPs would develop and implement their medical financing systems not later than September 30, 2000 in accordance with policies and procedures acceptable to the Bank (page 38(ii)). (viii) Institutional Building and Project Management. Establishing project management offices at provincial, prefecture, and county levels, and Township Work Stations (TWSs) and village project implementation groups (VPIGs); launching a pilot integrated poverty reduction management program in one Banner in Inner Mongolia to coordinate poverty reduction activities and entities; implementing these under a single management structure and establishing systems to monitor, and evaluate the impact of the Project on poverty reduction, voluntary settlement, labor mobility, and environment and to disseminate the results.

The Project cost estimate, including contingencies, is presented in the table below. The front-end fee of US$0.6 million will be financed under the Bank loan.

1.Land andHousehold AA-Agiutr 141.00 85.2 S005. Development

2. Irrigationand Land Al - Irrigationand 86.30 27.7 42.20 26.4 Improvement Drainage

3. Rural Infrastructure TR -Rural Roads 28.70 9.2 11.90 7.4

4. Rural Enterprises AM - Agro-Industry 13.60 4.4 6.80 4.2 and Marketing

5. Labor Mobility(Gansu) SE - Employment 15.70 5.0 7.70 4.8

6. VoluntarySettlement VR - Resettlement 6.50 2.1 0.10 0.1 (Qinghai)

7. SocialSectors RR - HumanResources 8.70 2.8 0.00 0.0

8. InstitutionalDevelopment BI - Institutional 10.60 3.4 5.70 3.6 and Project Management Development TotalProject Costs 311.10 99.8 159.40 99.6 Front-endfee 0.60 0.2 0.60 0.4 Total Financing Required 311.70 100.0 160.00 100.0

2. Key policy and institutional reforms supported by the Project: A clearly articulated Government policy for poverty reduction, including an established institutional arrangement for its implementation underpin the Project. This policy has been reviewed by the Bank, and has been assessed to be very satisfactory. Hence, no major reforms are envisaged in this regard. Lessons learned from ongoing Bank-assisted poverty reduction operations, however, attest to the need for fine-tuning to make production credit programs more sustainable, increase flexibility in procurement arrangements to reflect the dispersed nature of project activities, improve targeting, and enhance, rationalize, and co-ordinate line agency responsibilities. These lessons have been taken into account in the design of the Project. Households would be introduced to formal credit channels and established input credit lines. To foster coordination, a Memorandum of Understanding (MOU) to be signed between Project Management Offices (PMOs) and executing line agencies, specifying time-bound delivery of 8 services, has been introduced. An integrated pilot poverty reduction management program would be implemented in Inner Mongolia. The county resources for poverty reduction activities would be merged, and implementation arrangements harmonized, including sub-project financing and selection criteria. Under the pilot activity, no distinction would be made between locally and externally financed activities. This pilot is anticipaited to demonstrate and evaluate the value added of integrated management.

3. Benefits and target population: The Project, by creating employment and income generating opportunities, would benefit an estimated 1.7 million people living in absolute poverty (674,000 persons in 167,000 families in Inner Mongolia; 930,000 persons in 197,000 families in Gansu; and 61,800 persons in 13,600 families in Qinghai) including minority nationalities residing in 3,700 remote and inaccessible villages of Qinghai, Gansu and Inner Mongolia. Food grains, cash crops, fruits, wood, forages, meat and wool constitute the primary project outputs.

A survey of selected project counties indicates that an average household produces adequate food grain supplies for about six months. For the remaining six months, most households borrow from relatives or better-off neighbors to supplement the limited welfare ration provided by local governments. In Qinghai, benefits would accrue tc, the migrants in the 'move-in' area, the existing households in the 'move-in' area, and to famiilies who remain behind in the 'move-out' counties and villages. In the case of the latter, out migration would reduce population pressures and contribute to arresting ecological degradation. St.ep uplands previously cultivated would be returned to grass and forest land, thereby reducing soil erosion and increasing water conservation. Rehabilitation of 85 percent and terracing of 15 percent of the land to be vacated would be financed by the project counties. Some poor households who remain in the 'move-out' villages would be able to increase their per capita grain production through cultivating larger areas of the terraced land. County governments would face less demand on their limited resources for welfare assistance in times of food shortage. When fully implemented, the project beneficiaries would meet basic needs of food and clothing and, in many cases, earn cash income from a marketable surplus and labor mobility.

The Project would also improve the overall education and health status of the target population by increasing primary school enrollment and enhancing completion rates, and reducing infant and maternal mortality and morbidity. Provision of safe rural water would also contribute to reduction of water borne diseases. Selected rural roads would provide access to social services and markets and promote investments in rural enterprises.

The target households from poor villages in the nationally designated poverty counties would be selected based on multiple criteria including per capita income and grain production. Often, the income variation among households in a selected poor village is insignificant. 9

4. Institutional and implementation arrangements: Institutional arrangements

China is intensifying the reforms envisaged under its decentralized development strategy; delegating responsibilities to the lowest levels of quasi-government structure-Village Committees, while maintaining the policy, institutional and financial responsibilities under central, provincial and local governments. The Project would utilize existing institutions, strengthening them as appropriate, to cater to the special conditions in the PPs. A five-tier implementation arrangement is proposed:

(i) At the national level. The State Council's Leading Group for Poverty Reduction"'and Development (LGPRD) and its Secretariat, the Foreign Capital Project Management -Center (FCPMC), are the supreme bodies entrusted to coordinate the 8-7 Plan. FCPMC successfully directed the preparation of the SWPRP and Qinba and continues to play a key role in their implementation. For the Project, the FCPMC would make available its implementation and monitoring expertise to the PPs and provide quality control service, which include FCPMC reviews of all draft project documents prior to submission to the Bank. It would also be responsible for project supervision and reporting, supervising the poverty impact consultants and organizing overseas study tours. Details of specific tasks to be carried out and estimated service costs will be reflected in the MOUs between PPs and FCPMC. Assurances were obtained during negotiations that the Borrower would continue to maintain FCPMC to be responsible for project oversight, overseas training coordination, and poverty impact monitoring and that the PPs would carry out all training, studies and research under the Project in accordance with a program acceptable to the Bank (page 38 (iii)).

(ii) At the provincial level. Project Leading Groups (PLG), headed by Vice Governors, have the overall responsibility for poverty reduction. Comprising provincial line agencies, the PLGs approve annual work and economic plans and programs of the poverty counties and coordinate the activities of line agencies. Each PP has a Poverty Alleviation and Development Office (PADO), serving as the Secretariat of the provincial PLG. With assistance from line agencies, PADOs are responsible for the execution of targeted provincial poverty reduction operations, monitoring the program and liaison with the LGPRD. The three Project PADOs have established dedicated provincial PMOs with technical staff seconded from the line agencies to better co- ordinate implementation of the Project. Provincial PMOs have signed MOUs with the line agencies (water conservancy, agriculture, education, health, etc.) for specific project related tasks. Procurement Agents would be appointed for the procurement of large works and goods under International Competitive Bidding (ICB). Provincial PMOs would maintain consolidated county project accounts for annual audits. Technical Advisory Groups would also be appointed in each PP to advise on technical issues. Assurances were obtained during negotiations that PPs would maintain provincial PMOs, county PMOs, TAGs and environmental experts (page 38 (iv)).

(iii) At the prefecture/league/municipal level. Project Leading Groups, headed by commissioners and comprising project related technical bureaus, have been set up in all PPs, as have the PMOs in all Project prefectures, in 'move-in' and 'move-out' areas in Qinghai.

(iv) At the county level. Project Leading Groups, comprising line agencies and chaired by a county magistrate, have been established to oversee the poverty reduction programs. Executive powers are vested in the County Project Management Office, responsible for preparing the county's poverty reduction plan, implementing projects, coordinating the activities of the line agencies, providing technical support to Township Work Stations and Village Project Implementation Groups (VPIGs), monitoring physical and financial progress, and reporting to PADOs. The county PMOs would sign local contracts for works, carry out designated works on force account and maintain separate project accounts. In Qinghai, each 'move-out' county has 10

established a Settlement Office to implement the Voluntary Settlement Implementation Plan (VSIP) and manage the selection and transfer of the migrants.

(v) At the village level. Project beneficiaries would be organized by VPIGs to participate in the Project. The VPIGs, with technical support from TWS on new agricultural practices and technologies, would be responsible for implementation at the village level. Except for fertilizers, which would be directly purchased by farmers, the VPIGs would procure and distribute the agricultural input packages, based on approved village development work plans. The VPIGs would also collaborate with the Village Committees on issues relating to crop and agricultural land leasing; monitoring and enforcement of grazing; and preparation, updating, and approval of annual feed and forage balances. The All China Women's Federation (ACWF), as members of the VPIGs, would assist in promoting the participation of women in the Project. Villagers would also be mobilized for local infrastructure works.

Implementation arrangements

Project Implementation Plans (PIP) and Annual Implementation Plans. The PMOs have prepared PIPs describing the Project objectives, components, costs, financing plans, implementation arrangements, reporting arrangements, procurement plans, financial management systems, and project and poverty monitoring and evaluation arrangements. The PMOs would adopt the PIPs as key implementation guides. In add ition, detailed annual implementation plans would be prepared and furnished to the Bank. The Pro ject would be implemented over a period of six years, provision has been made for retroactive financing to cover cost of agreed Project activities to be carried out prior to effectiveness. The cost of reclaiming about 300 hectares in Qinghai as a pilot for voluntary settlement, and of social services and rural infrastructure would be financed retroactively. Selected sub-projects establishing the institutional and project management arrangements in Gansu and Inner Mongolia will also be financed retroactively. Assurances were obtained during negotiations that starting in 1999, each PP wouldfurnish to the Bank its annual implementation plans no later than December I of each year and that the annual implementation plans would include all activities to be implemented in the coming year, including training, studies and research, under its part of the Project and to be accompanied by a training plan for the activities (page 38 (v)).

On-lending arrangements. The Ministry of Finance would allocate the International Development Association (IDA) credit to the PPs on the same terms and conditions of IDA, except for the grace and maturity period which would be five and 17 years respectively. The International Bank for Reconstruction and Development (IBRD) loan would be allocated to the PPs on the same terms and conditions as the Bank, with grace period of five years, and 20 years maturity. The PPs would bear the foreign exchange risk. The Agricultural Bank of China (ABC), on behalf of the three Provincial Finance Bureaus (PFB) would extend fertilizer loans to households and would be reimbursed from the Speciali Account. In Gansu, the ABC, on behalf of the PFB would appraise loan applications for rural enterprises. In Qinghai, ABC would use its own resources for rural enterprise loans. Lending rates to households for fertilizer and to investors for rural enterprises would be at the ABC rates applicable to similar investments. Payment terns and conditions on loans to households made in-kind (other agricultural inputs and farm machinery), would have the applicable ABC rates on loans for similar investments. Assurances were obtained during negotiations that sub-projects would be appraised in accordance with policies and procedures acceptable to the Bank and sub-loans would be made on conditions acceptable to the Bank. Lending terms for sub-loans to households and rural enterprises would be the applicable ABC lending terms. The PPs would engage the ABC to assist in appraising and processing fertilizer and rural enterprise loan applications: Sub-projects would be eligible for financing under the Credit/Loan only if the PP has determined that the following conditions have been met, and, in the case of a rural enterprise subproject in Gansu, 11 has been approved by the Bank. (a) the beneficiary is creditworthy and has the required technical skills and resources; (b) the sub-project is economically and financially viable, and commercially and technically feasible; (c) if the sub-project is for livestock, the beneficiary has adequate sources offorage or grain feed for the livestock, and the subproject is to be carried out in a village which has developed a livestock development and grassland management plan; (d) in the case of a rural enterprise subproject in Gansu, less than 12 of the beneficiary's proprietary interest is controlled by the government, and the beneficiary has undertaken to contribute at least 30% of the total estimated cost of the subproject in the form of equity, and the subproject is designed to alleviate poverty through job creation, provision of affordable inputs or Technical Assistance to poor households, or provision of profitable markets for these households; and (e) the subproject is in compliance with environmental standards acceptable to the Bank, as well as Chinese environmental regulations, and in the case of a rural enterprise subproject in Gansu, has been developed on the basis of an ELA approved by the local EPB. Rural enterprise subprojects in Qinghai would be appraised, financed and supervised by ABC in accordance with its standard policies and procedures for poverty alleviation loans (page 39 (vi)).

Procurement arrangements

The Bank's Guidelines. "Guidelines: Procurement under IBRD Loans and IDA Credits" (January 1995 revised in January, August 1996, September 1997 and January 1999) and "Guidelines: Selection and Employment of Consultants by World Bank Borrowers" (January 1997 revised September 1997 and January 1999) will govern all Bank-financed procurement. Bank-approved Chinese Model Bidding Documents will be adopted for all ICB and NCB for goods and works and the Standard Bid Evaluation Form will be used. The Bank's Standard Documents will be used where no relevant model document exists. Due to the scattered location and different nature of project activities, each PP would undertake its own procurement. Procurement agents will be selected by each province to provide assistance for ICB procurement. The procurement profile is shown in Annex 6, Table A.

A procurement assessment was carried out in all PPs. The assessment concluded that the procurement implementation arrangements are satisfactory and recommended an action plan for the PPs to: i) participate in a procurement training workshop; ii) develop sample documents for small civil works contracts and national shopping for small value equipment with the assistance of the Bank; iii) revise the draft procurement procedures and management rules specifying the responsibilities of each PMO; iv) recruit adequate a procurement staff at each level following pre- appraisal; and v) set up procurement filing system.

(i) Works. Works worth US$142.5 million would be required for the Project to support crop and livestock production activities, irrigation and land improvement works, rural infrastructure construction, rural enterprise development, social sector facility improvements, and facilities for institutional development. Works would be scattered over 325 townships in 28 counties and carried out over a period of six years, and are not anticipated to be of interest to foreign construction firms. The largest contract is not expected to exceed US$7.4 million. Although not anticipated, any contract for works estimated to cost US$10 million equivalent or more, would be procured under ICB procedures. Contracts costing less than US$10 million equivalent but not less than US$200,000 equivalent will be awarded under NCB procedures acceptable to the Bank. These works, worth about US$37.7 million, comprise a Class III road in Inner Mongolia, major and medium-size irrigation works and rural electrification works in Qinghai. About US$29.6 million worth of minor works, costing less than US$200,000 equivalent per contract and consisting of many small undertakings, include works for well and pumped irrigation, bridges, piped water supply, rural electrification in Gansu and Inner Mongolia, minor office buildings in Qinghai's project area, and construction of processing units in Gansu. These works would be suitable for lump-sum and fixed-price contracts awarded on the basis of quotations obtained from 12 at least three qualified domestic contractors in response to a written invitation and advertised locally. About US$69.9 million worth of works in individual assignments not to exceed US$100,000 per assignment for crop and livestock establishment, water conservancy and rural infrastructure spread throughout project villages would consist of small scale and labor-intensive land leveling, animal shed and silage pits, fish pond construction, tree plantation, household- based grassland and fenced land establishment, irrigation and drainage canals, cisterns, shallow wells and terracing, minor soil and water conservation works, simple access roads, non-pipe water supply works, and construction of household processing in Inner Mongolia. With the prior agreement of the Bank, these works would be carried out under force account and beneficiary participation arrangements to best utilize the local know-how, available materials and labor- intensive technologies. Progress payments for works under force account and beneficiary participation arrangements would be based on unit prices and quantities agreed annually with the Bank. About US$5.2 million worth of works would be non-Bank funded including buildings for social sector services in all PPs, office buildings in Gansu and Qinghai, and rural enterprises in Qinghai.

(ii) Goods. Goods worth about US$122.8 million would be required for the Project. To the extent practical, contracts for goods would be grouped into bid packages estimated to cost US$250,000 equivalent or more whenever possible to attract cornpetition and permit bulk purchasing. All contracts for goods costing US$250,000 equivalent or more will be awarded under ICB. These contracts would include cross country four wheel drive wagons (US$0.9 million) and plastic film (US$2.8 million). A margin of preference equal to 15 percent of the CIF price of imported goods or the actual customs duties and taxes, whichever is less, would be allowed to qualified domestic manufacturers bidding under ICB procedures. Office equipment and furniture (US$0.5 million) in smaller packages by value and labor market information system (US$0.2 million) in Gansu, household enterprise equipment (US$1.6 million), irrigation pipes (US$2.9 million) and motorcycles (US$0.2 million) for Inner Mongolia, farm machinery (US$0.2 million) and irrigation auto-monitoring equipment (US$0.2 million) for Qinghai, and a small quantity of plastic film (US$0.7 million) in small packages for all PPs would be procured through NCB procedures acceptable to the Bank and in a total amount of about US$6.6 million. These NCB contracts for goods would cost US$100,000 equivalent or more but less than US$250,000 for each contract. The remainder, worth US$81.7 million, mostly farm tools, locally available pumps, sprinklers and diesel engines, small amount of office equipment, household processing equipment for Gansu (US$12.2 million), purchase of animals, fish fly, concentrate and additives, seedlings and small amount of steel pipes (US$69.4 million), pick-up trucks and specialized vehicles (US$0.1 million) would be procured with contracts under US$100,000 equivalent using national shopping procedures. They will be procured in small batches from local markets and suppliers. All national shopping contracts will require at least three price quotations. Delivery of chemical fertilizer in a timely manner over a 6-year implementation period to over 300,000 farm households for crop and livestock production is critical to the success of the Project. To ensure timely delivery, ABC, acting as a fiscal agent of the PFBs, would make loans (US$16.6 million-} available in cash or voucher to participating households for purchase of fertilizers. Fertilizers would be procured at a reasonable price, account being taken of other relevant factors, such as time of delivery and reliability of the fertilizer. Aboul: US$14.1 million worth of goods would be non-Bank financed including rural enterprise equipment (US$0.2 million) and vehicle (US$1.2 million) for Qinghai, social sector equipment (US$2.6 million), organic manure (US$8.8 million) and pesticides (US$1.3 million) for all PPs.

(iii) Consultant services and training. Consultant services and training worth US$8.6 million would be undertaken for the Project. Employment of consultants (US$2.6 million) would be in accordance with the Bank's "Guidelines for Selection and Employment of Consultants by World Bank Borrowers" published by the Bank in January 1997, revised in September 1997 and January 1999. Most consultants required for the Project are expected to be individuals who would be 13 employed in accordance with the procedures applicable to individual consultants. For other consultants to be hired through a firm, the "Quality-and Cost-Based Selection" procedure would be used. All consulting assignments over US$200,000 would be advertised in Development Business. Training and study tours (US$5.5 million) would be reimbursed based on programs agreed with the Bank. About US$0.5 million worth of domestic training for health staff would be funded by the respective provincial governments.

World Bank review. All ICB contracts for works and goods will be subject to prior review by the Bank. The first three NCB contracts for works and the first three NCB contracts for goods in each PP would be subject to Bank's prior review. All contracts for consultant services in excess of US$100,000 for firms and US$50,000 for individuals would be subject to prior review by the Bank. All other contracts would be subject to ex-post review by supervision missions. Details are provided in Annex 6, Table B.

Disbursement arrangements

Retroactive financing. To establish the implementation and coordination procedures for this complex Project and test the quality standards, retroactive financing, not exceeding the equivalent of SDR9.3 million (US$12.7 million), is recommended for expenditures incurred between January 10, 1999, date of appraisal, and date of Loan/Credit signing to: i) set up a pilot program for the voluntary settlement scheme in Qinghai, and ii) implement selected irrigation and land improvement activities, including establishment of crops, livestock and trees, training, study tours for PMO staff, consultant services and procurement of essential office equipment. A Bank mission will take place to Qinghai prior to the launch of the pilot program in order to ensure that the updating of the VSIP will take place in accordance with terms of reference acceptable to the Bank.

Statements of Expenditure (SOE). SOE would be used for disbursements against: i) contracts for all works except those subject to the Bank's prior review; ii) contracts for goods costing less than US$250,000 except those subject to the Bank's prior review; iii) training and study tour expenses; iv) consultant services costing less than US$100,000 for firms and less than US$50,000 for individuals; and v) labor placement expenses. Disbursements for all works carried out under force account and beneficiary participation arrangement would be made against statements of physical progress achieved at each site at unit prices agreed with the Bank at the start of the Project and subject to annual review. The supporting documents for SOEs would be retained by provincial PMOs and made available for review by Bank supervision missions. In the case of contracts for goods, works and services above these thresholds, disbursements would be made against full documentation of the contracts and other supporting documents.

Special Account. To facilitate disbursement, three Special Accounts in US dollars, one for each province to be operated by the PFB, would be established in banks and on terms and conditions acceptable to the Bank. The initial advances from the Bank would be US$6.0 million for Gansu, US$4.0 million for Inner Mongolia and US$3.5 million for Qinghai, the estimated average expenditures for a four month period. Applications for replenishment, supported by appropriate documentation, will be submitted monthly or when the amounts withdrawn equal 50 percent of the initial deposits, whichever comes first. Assurances were obtained during negotiations that the Borrower would open three Special Accounts in banks and on terms and conditions acceptable to the Bank (page 39 (vii)).

The Project is expected to be completed by June 30, 2005 and the Loan/Credit is expected to close on June 30, 2006. The allocation of loan proceeds is shown in Annex 6, Table C and an estimated schedule of disbursements is presented in Annex 6, Table D. 14

Financial management arrangements

Organization arrangement. Finance and Accounting Divisions (FAD) have been set up in the three provincial PMOs as well as in each county PMO. In each provincial PMO the FAD comprises a Division Chief in charge of finance, one disbursement officer, one accountant and one cashier. In the FAD of each county PMO, there is one accountant and one cashier, and the PMO director is responsible for the overall budget planning and authorization. Each county PMO would prepare disbursement applications with a completed set of supporting documentation to be cleared by the county finance bureau before submission to the provincial PMO. Provincial PMOs would consolidate and forward the applications to the Provincial Finance Bureau (PFB). PFBs have managed several projects assisted by the World Bank and the staff are familiar with the requirements of the MOF and Bank withdrawal application and disbursement procedures.

Accounting system and procedure. The proposed accounting and internal control system and procedure have been reviewed and are acceptable. All three provinces have drafted Financial Management System Manuals (FMSM), which describe the proposed accounting policies and procedures including a Chart of Accounts, accompanied by explanations of the items to be included in the various accounts, assignment of responsibilities and delegation of authority, explanations of documentation and approval requirements for various types of transactions and journal entries. Project accounting transactions woulcl be processed on the accrual basis, and the method of double entry bookkeeping would be adopted. Each county PMO would also maintain a project subsidiaries ledger, and cash disbursement journal, with account distribution headings corresponding to the Project's specific chart of accounts. Provincial PMOs would keep both a project general ledger and subsidiaries ledger. In addition, provincial PMOs would keep i) cash receipts journal for recording the multiple sources of funds; and ii) fixed asset register for identifying and controlling vehicles and equipment purchased under the Project.

Reporting requirements. The following reports would be prepared by each provincial PMO and furnished to the Bank semi-annually: i) Financial Statements (Balance Sheet, Summary of Sources and Uses of Funds by Project Component, Statement of Implementation of Credit Agreement, Statement of Implementation of Loan Agreement, Special Account Statement); ii) Project Progress (output monitoring report); and iii) Procurement Management (contract expenditure and procurement management for goods and works).

Audit arrangements. The State Audit Bureau (SAB) would authorize each Provincial Audit Bureau (PAB) to audit project accounts. PABs have audited accounts under World Bank-assisted projects and are considered suitable for such audits. Examination and sample tests will be made on reviewing the internal control system including accounting control, business course and division of obligations and responsibilities; checking records of accounts; verifying the fixed assets, confirmning accountants receivable and payable and other auditing procedures. The accounts to be audited would include the Special Accounts, withdrawals from the account made on the basis of SOEs, and the auditor's opinion as to whether such withdrawals were made against expenditures eligible for disbursement by the Bank Group. Assurances were obtained during negotiations that the PPs would furnish audited project accounts to the Bank within six months of the close of each fiscal year (page 39 (viii)).

Implementation progress reporting arrangement. The provincial PMOs have finalized the Project progress monitoring system in consultation with FCPMC, to serve as a management information system. The SWPRP and the Gansu Hexi Corridor Project monitoring formats would have been modified and adopted for the Project. Provincial PMOs with inputs from the county PMOs would furnish implementation progress reports to the provincial PLGs and the Bank semi- annually. Assurances were obtained during negotiations that the PPs would furnish progress reports semi-annually to the Bank not later than January 31 and July 31 of each year starting in 15

2000; a mid-term report not later than January 31, 2003, and an Implementation Completion Report (ICR) not more than six months after the Closing Date of June 30, 2006 (page 39 (ix)).

World Bank supervision plan. To supplement the PPs' and FCPMC's annual supervision of project implementation and exercise the Bank's fiduciary responsibility, the Bank would carry out semi-annual reviews in consultation with the provincial PMOs, starting with a Project Launch Workshop in September 1999. Bank supervision mission terms of reference would be prepared in consultation with FCPMC and provincial PMOs to ensure that emerging issues are reflected and the required skill mix is available. Great emphasis would be placed on social and environmental issues including regular participation of senior Bank specialists on the supervision team.

Project monitoring and evaluation arrangements

Poverty reduction and environmental monitoring and evaluation. The State Statistical Bureau (SSB) in collaboration with provincial Statistical Offices, would be responsible for developing and implementing the poverty impact monitoring, evaluation and dissemination system. SSB has gained valuable experience through similar assignments under the SWPRP and Qinba projects. FCPMC would, under its MOU with the PPs, take primary responsibility on behalf of PPs for supervision and quality assurance including oversight of SSB. The performance indicators presented in Annex 1 constitute the main elements for monitoring and evaluation. The indicators are elaborated in the baseline survey. Six annual sample surveys would be conducted during project implementation. Assurances were obtained during negotiations that PPs would furnish, not later than December 1 of each year, annual poverty reduction and environmental impact monitoring and evaluation reports to the Bank (page 39 fx)).

(i) Voluntary settlement monitoring. The Qinghai Plateau and Geographic Research Institute would be responsible for monitoring the implementation of the VSIP. The Institute has prepared the methodology and procedure for monitoring the indicators described in the VSIP. The Institute would also carry out the monitoring in the Pilot Village Settlement Area under the supervision of independent experts. A 5 percent population sample tracking procedure has been developed to track certain indicators (elaborated in Annex 9). Assurances were obtained during negotiations that Qinghai would take all measures necessary to ensure that the migration of all households to the move-in site will be implemented on a voluntary basis and in a manner designed to improve the living standards and productivity of all households in both the move-out and move-in areas. To this end, Qinghai would carry out the emigration in accordance with the VSIP. In particular, it would allocate irrigated farm land in the move-in area to each migrant household and each household currently living in or using land in the move-in area in amounts and under arrangements acceptable to the Bank, including a written lease agreement for a minimum of 30 years on terms acceptable to the Bank and exemption from agricultural taxes for the first 3 years of the lease. Qinghai would also implement its Local Beneficiary Participation Action Plan in accordance with an implementation plan acceptable to the Bank2 Under this implementationl plan, Qinghai would take a number of measures to (a) ensure social stability in the move-in area, (b) ensure that pastoralists are able to maintain their preferred methods of livelihood, (c) promote culturally appropriate education for minority nationalities in both move-out and move-in areas, (d) ensure adequate, culturally appropriate access of people in both areas to health care services, and (e) integrate cultural preferences of all nationalities involved in the migration. In addition, prior to any migration to the move-in area, Qinghai would cany out a study to update its assessment of the social needs of both migrating households and households living in or using land in the move-in site and recommend measures to update the VSIP on the basis of this study. Once the VSIP is updated to take into account the Bank's views on the study and its recommendations, Qinghai would undertake a pilot voluntary settlement program to move up to

2This implementationplan is set out in a supplementalletter to the ProjectAgreement, which is includedin the packageof legalagreements made available to the Bank's ExecutiveDirectors. 16

200 households in one village into the move-in area. Prior to undertaking any further settlement in the move-in area, Qinghai would report to the Bank on the results of the pilot program and revise the VSIP to take into account the views of the Bank on the report and its recommendations. Qinghai would: (a) monitor and evaluate implementation of the VSIP in accordance with indicators acceptable to the Bank; (b) prepare and furnish an annual report on the implementation of the VSIP to the Borrower and the Bank not later than January 31 of each year for review; and (c) based on the report, revise the VSIP in agreement with the Bank. Qinghai would employ independent experts with terms of reference, qualifications, and experience acceptable to the Bank, not later than December 31, 1999, to assist in monitoring and evaluating the VSIP. Finally, within 3 years after commencement of VSIP implementation, Qinghai will: (a) carry out a study to evaluate the social and environmental impact of the VSIP on the move-out counties and to recommend measures to enhance environmental sustainability and living conditions in these counties; and (b) implement the study's recommendations, taking into account the Bank's views on the matter (page 40 (xi)).

(ii) Labor mobility monitoring. Gansu would implement the Labor Mobility component strictly on a voluntary basis. During negotiations assurances were obtained that Gansu would take all measures required on its part to ensure that the labor mobility program is implemented on a voluntary basis and in a manner designed to ensure safe and equitable working and living conditions for the participating laborers. Gansu would monitor and evaluate this program in accordance with indicators acceptable to the Bank, and would prepare, with the assistance of independent experts employed not later than December 31, 1999, an annual report covering, inter alia, the employment selection process, the safety, treatment and living conditions of the placed workers, and the impact of the remittances on the workers' families, and furnish this report to the Bank not later than December 1 of each year for review. The labor mobility program would be revised, based on the report's recommendations, and taking into account the Bank's views on the matter (page 40 (xii)).

(iii) Environmental monitoring. Based on the PPs' environmental management plans, a set of indicators have been selected to be monitored during the life of the Project. Each PP has assigned the provincial Environmental Protection Bureau (EPB) to conduct periodic environmental evaluations of project investments and to review rural enterprise sub-projects prior to approval. Monitoring costs are included in the Project cost estimate to ensure timely reviews. The use of 1,400 hectares of natural grassland in Inner Mongolia and 8,000 hectares in Gansu would be monitored by the local Grassland Monitoring Bureaus. Village-area feed and forage balances would be prepared by the township PMOs, assisted by their local county PMOs. The determination of ruminant and non-ruminant livestock numbers in Project areas would be based solely on updated and cleared feed and grain balances. An independent Chinese institution would carry out spot checks on feed balance preparation, livestock numbers, and grassland integrity. Assurances were obtained during negotiations that the PPs would ensure that each village in their Project areas would prepare and implement annual livestock and grassland management- plans to ensure acceptable livestock feed balances and environmentally sustainable use of grassland. The PPs would ensure that all village grassland used under the Project is leased to the users for a minimum of 30 years under conditions designed to ensure the land's use in accordance with the grassland management plans. The PPs would monitor and report to the Bank annually on the livestock balances in these villages and adjust their plans to ensure continued acceptable livestock feed balances and environmentally sustainable grassland use, taking into account the views of the Bank on the matter (page 40 (xiii)). 17

D: Project Rationale

1. Projectalternatives considered and reasons for rejection: The scope and design of the Project was defined by the PPs based on their respective poverty alleviation strategies. In accordancewith the Plan guidelines,the PPs have targeted nationally designatedpoverty counties (33 counties in total). Developmentalternatives in the selected areas of the PPs are constrainedby the fragility of the environment.In Qinghai,the resourcebase of the six poor 'move-out' counties has been depleted as a result of the combination of population pressures, intensivecultivation, and over grazing. With steadily increasingpopulation pressures, more unsuitableland will be cultivated,thus accelerating environmentaldegradation. Even with the use of improved technologies and systems, the population pressure would prevent the achievement of even the most basic living standards. The only viable and environmentally sustainablealternative, therefore, is to voluntarilyrelocate parts of the population.The Australian Aid (AusAID), International Fund for Agricultural Development (IFAD) and World Food Program (WFP) have provided assistance to Qinghai in counties that have potential for in situ development.However, the areas selected for the Project do not offer developmentopportunities and the Bank has assessed that the only environmentally sound alternative is to contain cultivationof the degraded lands and assist the local populationto voluntarilyseek opportunities elsewhere.

Gansu is also tackling poverty by a combinationof voluntary settlement(e.g. the Bank assisted Gansu Hexi Corridor Project designed to relocate 200,000 people) and developmentof existing cultivated lands. The Project design in Gansu recognizesthat, with the adoption of recent best practice (the completed IDA assisted Gansu Agricultural Development Project) and environmentally friendly technology packages (combination of improved seeds, fertilizer, supplementary irrigation and agricultural advisory services) for crop and livestock systems, current productivitycould be increased.

Inner Mongolia poses a special challenge for poverty reduction.The Project counties suffer from arid climate, severe droughts and wind erosion. The interventionsproposed in Inner Mongolia combine resource conservation measures (cropland protection and controlled grazing of range lands) with irrigated agriculture and livestockdevelopment. In both Gansu and Inner Mongolia, fanrnerswould adopt rainwatercollection for storage in cisterns. These systems are cost effective and widely used in the PPs. In addition, a window of opportunitynow exists in the three PPs supportingsustainable arable and pasturelandmanagement to curtail further land degradationby crop and livestock production practices. By shifting from natural pasture-based livestock production to the feeding of cut-and-carrycrop by-products and planted forage crops, grazing pressure on overgrazedgrasslands would decrease.Furthermore, the government's establishment of long-term agricultural land leases up to 30 years will be adopted by all project beneficiaries utilizing crop or grasslandin order to provide the necessary incentivesfor more sustainableland use. 18

2. Major related projects financed by the Bank and/or other development agencies: (completed, ongoing and planned) ;~~~~~~~~~E

Bank-financed Implementationof an integratedand China - SouthwestPoverty Reduction S HS complexpoverty reduction project. Project- Ongoing Targetingassistance to absolutepoor and China - QinbaPoverty Reduction S HS predominantlyminority nationalities Project- Ongoing living in inaccessibleand resourcepoor areas. Communityparticipation. China - NationalRural Water Supply- S S Ongoing Povertyreduction through voluntary China - Gansu-HexiCorridor Project - S S settlement. Ongoing

Povertyreduction through voluntary China - NorthernIrrigation Project - S S settlement. Completed Other developmentagencies IFAD-PovertyAlleviation Ongoingin Qinghai AusAID-PovertyAlleviation Ongoingin Qinghai UNDP/WFP-PovertyAlleviation Ongoingin Qinghai EuropeanUnion - PovertyAlleviation Ongoingin Qinghai EuropeanUnion - BasicEducation Under preparationin Gansu

IP/DO Ratings:HS (HighlySatisfactory), S (Satisfactory), U (Unsatisfactory),HU (HighlyUnsatisfactory)

3. Lessons learned and reflected in the Project design: A recent Quality Assurance Group review of SWPRP and lessons learned from the Shanxi Poverty Alleviation Project concluded that strong commitment of the central and provincial governments to integrated poverty reduction is key to early success of the projects. The Review's recommendations for strengthening design of poverty projects include: i) flexibility in project design, and phasing of project components; ii) recognizing the long-term institutional implications of the project including the roles of the LGPRD, ministries and provincial agencies; iii) identifying a sustainable agricultural input financing source; iv) greater attention to sectoral and sub-sectoral policy and institutional challenges; End v) need for simplifying and introducing flexibility in the Bank's procurement procedures in view of the scattered nature of the poverty projects. The Project, while adopting the overall framework of SWPRP, has introduced new design features-selectivity in project components, scaling down investment in rural enterprises while promoting non-state sector joint ventures for commercially viable enterprises with sister cities and provinces in the east; delegating sub-project appraisal and approval to financial intermediaries; linking project beneficiaries with fi:nancial intermediaries for input credit for purchase of fertilizers instead of central procureiment and distribution; strengthening the 19

coordination of participating line agencies through signing of MOUs; and delegating primary project preparation and implementation responsibility to the PPs.

4. Indications of borrower commitment and ownership: The Borrower and the PPs have demonstrated strong commitment to and ownership of the Project. The Ministry of Finance, State Development and Planning Commission and the State Council's LGPRD have endorsed the Project. In January 1998, the Government wrote to the Bank, lauding the effective assistance the Bank is providing for poverty reduction in China and calling upon the Bank to continue its assistance. The PPs have the primary responsibility for project design, preparation and implementation, thus fostering stronger local ownership. The PPs have also devoted a considerable amount of staff and financial resources for preparation. The Vice Governors are actively overseeing the preparation of the Project and have mobilized technical institutes and line agencies to prepare and implement the Project.

5. Value added of Bank support in this Project: The Bank's support for the Project has the following value added:

(i) Experience in addressing poverty. The Bank has demonstrated effective formulation of multi-sectoral, integrated projects to effectively address poverty in the most difficult environments. (ii) Enhanced public sector institutions. Public sector institutions are often vertically organized to deliver services. The Bank's intervention in poverty reduction operations would continue to demonstrate to the Government that public sector institutions can also be effective when services are needed and delivered horizontally. The Project would enhance horizontal effectiveness of public sector institutions, formalized by MOUs between PMOs and line agencies. (iii) Leveraging resources. The Bank's assistance often leverages additional domestic resources for poor counties. The Bank's involvement would also promote the poor counties to potential investors under the East-West Cooperation Program. (iv) Transfer of knowledge. Technical knowledge transferred during project preparation such as developing a sound project design, introducing market tests to justify project outputs, developing transparent targeting mechanisms, application of social assessments, environmental procedures, training project staff in project cost and benefit estimates and analysis methodologies are the value added gained by the PPs. (v) Improved design features. The Project is introducing improved project design features: a) rigorous selection of project components; b) focusing on household level food and raw material processing facilities and scaling down township and village enterprises; c) inter-agency cooperation by adopting standards and implementation arrangements from ongoing Bank-assisted operations-roads, rural water supply, education and health; and d) piloting a program approach. for project management in one Banner in Inner Mongolia. Lessons from these new features and the pilot may form the basis for future investment planning and financing. 20

E. SummaryProject Analysis (Detailed assessments are in the project file, see Annex 8)

1. Economic (supported by Annex 4): [X] Cost-Benefit Analysis: NPV= US$1,282 million; ERR = 22 %

The detailed economic analysis of the individual project components (except Social Sector and Institutional Building) and the whole Project is presented in Annex 4. Economic rates of return (ERR) range from an average of 15 percent for rural infrastructure to about 50 percent for rural enterprises with the overall project ERR at 22 percent. The Project's ERR, lower than the ERRs estimated for SWPRP and Qinba projects, with 36 percent and 40 percent respectively, is explained by the significant investments made in water conservancy works and social sector activities, which account for about one third of the total Project investments.

2. Financial (see Annex 4): NPV=US$1,034 million; FRR = 20 % Financial analysis has been carried out for all Project components (except for Rural Road, Social Sector and Institutional Building) and the Project as a whole is also presented in Annex 4. The overall financial rates of return (FRR) are at 27 percent for Gansu and 19 percent for Inner Mongolia and 13 percent for Qinghai.

Fiscal Impact:

The financing of the Project is based on the govemrnent's policy of "who benefits who pays". Loans to provinces are on-lent to counties with guarantees issued by the Prefectures. Except for rural infrastructure and social sector investments, beneficiaries would either be liable for full cost recovery or take over the full responsibility for operation and maintenance. In Qinghai, irrigation charges combined with a land lease fee would cover the full operation and maintenance costs and a portion of the investment cost. The average annual Project investment cost represents less than 5 percent of the provincial annual expenditure on various poverty programs and would have minimal fiscal impact.

3. Technical: The technologies for the Project are not new. The crop and livestock development, the proposed production technologies and systems have proven track records, and are already widely accepted by technicians and farmers. Yield assumptions have been based on average yields obtained from progressive farmers in the project regions using the various technologies. In Inner Mongolia the average yield assumptions have been discounted by a "risk" factor to allow for effects of drought and occurrence of other adverse factors. The discount factor has been calculated by taking the average yield of grain and oil crops in the project leagues for the five years ending 1996 less the minimum yield divided by the average yield. No specific risk factor has been estimated for Gansu, but the projected yields have been based on conservative actual yields achieved by farmers using the project technologies. In both provinces, the "best practice" yields achieved by farmers are well above those used for project benefit estimates.

The Project's intervention would facilitate the poor households' access to financial resources and allow for expansion of the adoption of modern farm technologies. Low-cost labor intensive rural infrastructure construction has been most effective in mobilizing surplus rural labor for community self-help efforts; rural water supply and rural roads ranked high in order of priority by the very poor. The Project's technical design is consistent with that of the Bank assisted SWPRP and Qinba Project.

The Project cost estimates have been developed based on a careful analysis of unit prices for all project investments. Engineering costs for dams and major structures are based on feasibility 21 level designs, and for rural infrastructure schemes on representative designs specifically prepared for the purpose of arriving at reliable cost estimates. A composite physical and price contingency of 13 percent has been used. Since the prices of major construction materials (such as cement and steel) have dropped significantly in recent years and have stabilized, the allowance for contingencies is reasonable.

4. Institutional: Executing agencies

The project executing agencies are described in detail in Annex 2. Project execution requires the collaboration of line agencies in the provincial capitals, prefectures, counties, townships and village project implementation groups, ACWF and project beneficiaries. During project preparation, these agencies and groups have demonstrated effective teamwork. In Gansu and Inner Mongolia, the line agencies have been associated with Bank-assisted operations and are familiar with Bank procedures. Previous Bank activities in Qinghai have been limited to social sector operations and the line agencies had to develop familiarity with the Bank during project preparation. With the introduction of an integrated poverty reduction project, the line agencies would be required to function in a horizontal structure, synchronizing their activities with other related project interventions. Toward this end, PMOs will play a prominent role. Staff training and study tours to familiarize the staff with similar operations in China and abroad would be arranged. All the participating executing agencies are implementing provincial and central government financed sectoral programs and have adequate capacity to implement their respective project components. However, the added features of the Project including the participating role of beneficiaries, strict compliance with environmental standards, procurement procedures and project technical standards would require the agencies to adjust their current management practices. The involved agencies have demonstrated the willingness and eagerness to improve upon their current practices.

Project management

The three PPs have established PMOs at all levels (described in detail in Annex 2). Although the PMOs were recently created, their capacities have developed significantly during project preparation. Project staff have participated in a series of training seminars and benefited from technical assistance and study tours to similar Bank assisted projects in China. It has been emphasized that the PMOs role is to effectively coordinate implementation of the annual work plan by the technical line agencies, mobilize counterpart resources on schedule and ensure that the agreed procedures and technical standards are strictly applied. Toward this end, the PMOs, and the line agencies, have signed MOUs specifying the institutional support arrangements with specific delivery schedules. The Project's staff development component would arrange domestic and overseas training and study tours. Inner Mongolia has also volunteered to pilot a project management program in Wengniute Banner. It would strengthen the existing PADO, rather than create a county PMO and would apply the technical standards, sub-project selection criteria and implementation procedures adopted under the project to the Banner's use of poverty funds. 22

5. Social:

Social Assessment Review Process

Social Assessment (SA) documents were prepared for the three PPs by Qinghai Plateau Geography Research Institute, and the Agricultural aLndAnimal Husbandry Engineering Design Institute of the Agricultural and Animal Husbandry C'ollege, Inner Mongolia. The SAs for Gansu were prepared by local social scientists coordinated by the Western Poverty Reduction Project Gansu Project Office. In addition, a Voluntary Settlement Implementation Plan (VSIP), was prepared for Qinghai, which includes an Involuntary Resettlement Action Plan (RAP) as an annex, prepared in accordance with OD 4.30 Involuntary Resettlement. The Project also addresses the provisions of OD 4.20, Indigenous People by adoption of the VSIP, designed for predominantly minority nationalities in Qinghai, based on a detailed Social Assessment, and through a project design which reflects concerns of minority nationalities in Gansu and Inner Mongolia. The SAs and the VSIP include Local Beneficiaries Action Plans. The SA documents and the VSIP were furnished to the Bank prior to apipraisal and have been made available in the PPs to project affected groups, including women's organizations. The SA work was carried out in close coordination with the environmental assessment work, under the supervision of a Bank qualified, experienced social scientist, and in accordance with Bank practice.

Social Assessment Process

The SAs, which included the poverty assessments were prepared using a highly participatory approach in the form of surveys, interviews, and public meetings. Stakeholders and beneficiaries were consulted and participated in all stages of the process. In addition, extensive socio-economic surveys were conducted in both the 'move-out' and 'move-in' areas in Qinghai. Nationally designated poverty counties and townships were selected as project areas. Priority was given to townships with large numbers of ethnic minority nationalities. In Qinghai, six of the 14 national poverty counties were chosen as project areas for selection of voluntary migrants to form the 'move-out' population. Minority nationalities account for 51 percent of the total population of these counties.

The results of the SAs were used to ensure that the Project objectives and interventions are appropriate vis-a-vis the needs of the targeted beneficiaries and that the changes proposed in the agricultural, animal husbandry, and rural infrastructure components and Qinghai's voluntary settlement component, are acceptable to the affected population. The SA seek to ensure that changes brought about by the Project are equitable, irrespective of gender, ethnic and social differences.

In all PPs, project design reflects the recommendations of the SAs and the VSIP. Impacted stakeholders', including minorities', views have been fully analyzed and incorporated into the project design. The Bank will pay particular attention during project supervision ensuring that these recommendations be fully implemented.Assurences were obtained during negotiations that each PP would ensure that all persons affected by the Project fully participate in the Project's design and implementation and receive social and economic benefits from the Project that are compatible with their cultural preferences. To that end, the PPs would: (a) carry out the Project in accordance with their Local Beneficiary Participation Action Plan; (b) monitor the implementation of these plans in accordance with indicators acceptable to the Bank; (c) report to the Bank on the results of this monitoring; and (d) introduce revisions to these plans, as agreed with the Bank in order to achieve the plans' objectives (page 40 (xiv)). 23

Ethnic Characteristicsof Project Provinces.The population of PPs comprisesvarious ethnic groups, with Han Chinese in predominance.In Gansu, Han make up 88 percent while minorities (largelyHui, Tibetan, Mongol,Kazak, and Yugu) accountfor 12 percent.In Inner Mongolia,the main groups are Han (79 percent),Mongol (16 percent),Manchu (2 percent),Hui (1 percent) and other minorities (2 percent including Dawer, Korean, Ewenke etc.). In Qinghai, the total populationis made up of Han (57 percent),Tibetan (21 percent),Hui (14 percent),Tu (3 percent), Salar (2 percent),Mongol (2 percent)and other minorities(I percent).

Ethnic Characteristicsof Project Areas. The composition of beneficiaries in project areas varies across the PPs. In Gansu, Hui account for 7 percent and Han, 93 percent. In Inner Mongolia, about 14 percent of the beneficiariesare minorities and 86 percent Han. The ethnic compositionof project beneficiariesin Gansu and Inner Mongolia largely reflects the population compositionin the project counties/banners.In Qinghai, the migrants from the six 'move-out' counties are composed of Han (42 percent), Tibetan (6 percent), Hui (36 percent), Salar (7 percent) and Tu (9 percent). The 'move-in' county (Dulan) is comprised of Han (53 percent), Tibetan (23 percent), Mongol(14 percent),Hui (7 percent),and other minorities(3 percent). The project-affectedpopulation in 'move-in' townships is composedof Mongol(69 percent),Han (26 percent)and Hui (5 percent). Subsistencestrategies pursued by these communitiesfall into three broad classifications: the nomadic herders (60 percent), sedentary herders (9 percent), and farners (31 percent).

ProjectImpact and MitigationMeasures

In assessingthe Project's impact on local communities,a number of social issues were identified and analyzed in the SAs including: i) possibleadverse environmentimpacts; ii) land acquisition; iii) possible inequitabledistribution of benefits; and iv) issues related to ethnic minority rights. These issues are discussedbelow and described in more detail in Annex 9.

Possible Adverse EnvironmentImpacts. No major concerns regarding enviromnentalissues were identified in the SAs in Gansu and Inner Mongoliabecause the bulk of the Project activities are expectedto produce positive environmentalimpacts. In Qinghai, the main concern identified in the SA is related to the fragile ecology in the 'move-in' area. (See Annex 10).

Land Acquisition and Project Affected People. In Inner Mongolia, the construction of a Class III, 45 kilometer road would require acquisition of 40 hectares of Gobi grassland. The land is owned by three gachas and is not inhabitedor used for grazing. In Gansu, road constructionand rural enterpriseestablishment would require land acquisition from local communities.The total amount of land acquisitionfor roads is 85 hectares,of which 20 hectaresare dry farmlandand the rest wasteland. Land acquisition will take place in 29 project townships in 7 counties. It will affect 9,552 people in 1,931 farm householdswith lease agreements.No land acquisitionwould result in relocation of existinghouses. Land Acquisitionand CompensationPlans have been have been adopted.

In Qinghai, the voluntary resettlement component is going to affect groups of people in the proposed settlementor 'move-in'area. The primary affected group consistsof 63 households(352 herders) with leased, poor quality spring and winter pasture, stretching along the northern boundaryof the project area. Another group of 289 households (2,411 herders)use the project's south-northcorridor as a passageway.The third group of 248 households (1,237 members) farm amidst a dilapidated irrigation scheme. A Land Acquisition and CompensationFramework has been prepared for Qinghai in the VSIP. The Project will compensate all these affected people accordingto the Bank's OD 4.30, InvoluntaryResettlement and national and provincialrules and regulations. Three compensation packages have been agreed upon in consultation with the affectedfarmers and herders, village leaders and minorityreligious representatives: 24

(i) Herder Households with Leased Pasture. For the 63 herder households leasing spring and winter pasture, compensation includes trading in the poor quality grazing land for irrigated forage land. In addition, each herder household will be allocated land for housing and will enjoy the same preferential policies established for the voluntary settled migrants. (ii) Herder Households Requiring Passage. For the 289 households using the land as public passage way to grazing lands, three fenced-off transjit corridors will be established. In addition, the herders would acquire land for housing and farming on the same terms and conditions as the voluntary settled migrants. (iii) Households in Old Irrigation Area. 248 households living in the old irrigation area would be provided improved irrigation services, financed by the Project, and those farmers with less land than the standard set for the voluntary settled migrants would receive new land up to this standard.

Assurances were obtained during negotiations that the PPs would carry out their respective RAPs in a manner designed to improve the living standards and production levels of all persons suffering involuntary loss of shelter, productive assets, access to productive assets, income, or means of livelihood as a result of works under the Project that cause a change in land or water use. The PPs would: (a) monitor the implementation of their RAPs in accordance with indicators acceptable to the Bank; (b) report to the Bank on the results of this monitoring, and (c) introduce revisions to the RAPs, as agreed with the Bank in order to achieve the plans' objectives. If any sub-projects financed out of sub-loans would involve resettlement, the resettlement would be carried out in accordance with a resettlement plan which has been (a) prepared on the basis of a policy framework acceptable to the Bank, and (b) approved by the Bank (page 40 (xv)).

Ensuring Equitable Distribution of Benefits (including women). The SAs identified some concerns that the benefits of the proposed activities imay be inequitably distributed among ethnic nationalities and project affected households. Selection of households is going to be closely monitored to ensure fair and transparent selection. The SAs paid special attention to integrate women's' concerns, especially minority women who are generally not involved in public activities, into project design with specific project components benefiting women. Activities and programs designed to accommodate these concerns include: i) women's' education (literacy and technical skills); ii) improved health care (provision of midwives and training in women's health issues); iii) emphasis on attracting girls to schools; and iv) provision of special loans for women for cash-generating activities (sewing, noodle processing, embroidery, knitting, etc.).

Protection of Minority Rights. The SAs analyzed the characteristics of the major ethnic groups in the proposed project areas and established strategies (including cultural and educational programs and culturally appropriate health care access) in the project design. This was done in close cooperation with community leaders and religious elders. To retain "formal" protection of minority rights, Qinghai has represented to the Bank that the implementation of its part of theb project, including the VSIP, will not affect the autonomous status of Haixi Prefecture, the prefecture in which the 'move-in' area is located.

Specific Issues Related to Qinghai

Qinghai is the only PP where extensive voluntary settlement is proposed. Largely because of this, the project component has been the focus of scrutiny. The SA for Qinghai raises the possibility of conflicts between farmers and herders, and ethnic conflicts in the 'move-in' area, as potential risks. Public safety was also raised as a concern, based on previous experience with migrants. Additional concerns include: (i) claimed dilution of Tibetan presence in the 'move-in' area as part of a historical strategy of supplanting minority nationalities; (ii) perceived weakening of Tibetan culture through replacement of the herding lifestyle with that of farming, along with the 25

difficulties of ascertaining the "real" desires of the Tibetan herders in the project 'move-in' area regarding these changes; and (iii) possible use of prison labor in project construction. Each of these issues has been discussed and provided for in project design in general and the Local Beneficiary Participation Plan, Voluntary Settlement Implementation Plan, and Resettlement Action Plan specifically (see Annex 9).

Dilution of Tibetan Presence in the 'Move-In' Area. Of all the Tibetan autonomous districts or Tibetan-populated districts in China, the three project-affected prefectures (Haixi Mongol and Tibetan Autonomous Prefecture, Haidong Administrative District, and Xining Municipality District) are the "least" Tibetan. All have only small Tibetan population (11 percent, 10 percent, and 4 percent, respectively). Administratively, the three project-affected districts are the.only districts in Qinghai province not designated solely Tibetan Autonomous districts; all the other five districts are Tibetan Autonomous Prefectures.

Mitigation Measures in Qinghai. Under the implementation plan, Qinghai will settle minority nationalities in accordance with their expressed preferences, as intact communities from their villages of origin, and new villages will be designed with the objective of grouping culturally compatible nationality communities together. In response to concerns expressed by herders in the 'move-in' area, local police substations and people's courts will be established in the 'move-in' area to ensure overall public safety, and training will be provided to law enforcement officers. Grazing corridors will be constructed through the 'move-in' area reflecting herders' desires, and land will be allocated to herders in a manner to enable them to continue to use this land for herding. Primary and joint primary and junior secondary schools for minority nationalities will be constructed in the move-in area, and schools or programs for minorities will be constructed/developed in other areas of Dulan County as well as in the 'move-out' counties. Instruction in the language of the nationality served by these schools and nationality heritage curricula will be provided in each school. Similarly, health care facilities will be constructed in the 'move-in' area, and in other areas of Dulan County as well as in the 'move-out' counties. Special consideration will be given to integrating into the health care programs, providers and pharmacologies of the various health care and medical systems of nationalities served by these facilities. Consultations throughout project implementation will be held with community and religious representatives of each of the nationalities, to ensure that the VSIP and RAP are implemented in a manner compatible with their cultural preferences. In particular, cultural centers will be established and financial and technical assistance provided to develop cultural education and entertainment programs.

Special care will be taken to ensure that pastoralists will be able to maintain their preferred livelihood strategies. Provisions will be made to allow for participating families to adopt a range of strategies which would include full scale-farming, a mixture of farming and herding, and full- scale herding. As noted below, three separate passage corridors will be provided for herders whose access to traditional grazing areas would be disrupted by the proposed project which will enable their herds to pass through the settlement area.

Pilot Program. In order to ensure that no sudden negative impacts occur in the voluntary resettlement component, the implementation would be initiated by a pilot program. The program would be undertaken on 300 hectares for 200 households in advance of full-scale project implementation to test the proposed interventions at the field level with the participation of project beneficiaries. The pilot program would also include participatory evaluation techniques to solicit the views of a representative range of parties in the project area and to ensure that minority interests are taken fully into account and that Project components are appropriate to their cultures and ways of life. The pilot program would provide an opportunity to develop "lessons learned" which could be applied to the design and implementation of the full-scale phase of activities that would start only once the pilot program has been thoroughly evaluated by independent reviewers. 26

A 5 percent population sample tracking procedure has been developed. Indicators tracked to determine the success of the pilot program would include: (i) degree of inter-ethnic cooperation, and lack of serious conflicts; (ii) religious ceremonies and facilities are respected; (iii) both farmers and pastoralists continue to earn their livelihoods substantially undisturbed by the other; (iv) villages receive adequate water, electricity, and other infra-structural supports; (v) villages have access to adequate and culturally appropriate social services (schools, health care, credit, cultural enrichment); (vi) project beneficiaries frorn the move-out areas report the selection process ethnically equitable and efficient; and (vii) local governments and cadre from both the 'move-out' and 'move-in' areas collaborate to smooth the transition for host area population, migrants, and future migrants. A Bank mission will lake place to Qinghai, prior to the launch of the pilot program in order to ensure that the updating of the VSIP will take place in accordance with Terms of Reference acceptable to the Bank.

Monitoring of the Pilot Program. The pilot program is designed to avoid conflicts and enable adjustments in the project design before the bulk of the migrants arrive. The pilot settlement will be closely monitored to measure the effectiveness and adoption rates of the various program components and also to identify emerging issues, if any, in the relationship between migrants and receiving area population. Full scale settlement would only begin after the pilot program is completed and evaluated. In addition to the pilot itself, a 5 percent population sample tracking procedure has been developed for semi-annual progress reports to the PMOs and the Provincial Leading Group.

Indicators to determine the success of the pilot program are both economic and social. Economic Indicators include: (i) adaptation rates are high; (ii) incomes and yields increase; (iii) both farmers and pastoralists continue to earn their livelihoods wvithouthindering the other's effort; and (iv) villages receive adequate water, electricity, and other infra-structural supports. Social Indicators would include: (i) the project is carried out in a socioculturally appropriate manner; (ii) beneficiaries' satisfaction rates are high; (iii) there is not only an absence of conflict but mutual inter-ethnic respect in religious and other spheres; (iv) villages have access to adequate and culturally appropriate social services (schools, health care, credit, cultural enrichment); (v) project beneficiaries from the move-out areas report the selection process ethnically equitable and efficient; and (vi) local governments and cadre fi-om both the move-out and move-in areas collaborate to smooth the transition for host area population, migrants, and future migrants. The pilot program monitoring process will result in a report.

Project Provinces Capacity for Mitigation of Social Impacts and Protection of Minority Rights

Chinese law provides that minority nationalities practice regional autonomy in areas where they live in concentrated communities. Since the founding of the People's Republic of China, official administrative procedures have established a hierarchy of autonomous regions, prefectures, and counties. China's constitution delegates protected rights and powers of self-government to its 56 officially recognized nationalities. Nationwide, the non-Han 8 percent of the population listed as minorities enjoys special protections and status, including various degrees of administrative self- rule in a network of autonomous regions, prefectures, and counties. For each of these areas, special rights apply, such as the right to use their own language, have their own schools, control their own economic development and manage their own natural resources, among others. Once granted autonomous administrative status, there is no constitutionally explicit method for "decertifying" such status, and there is no automatic triggering mechanism for such a process even when the percentage of minorities in a minority autonomous area falls quite low.

On the ground, although minority policies have varied significantly over the decades of 20th century Chinese history, during past two reform decades, autonomous status has meant greater 27 legal protection for minority culture and languages and greater representation of minorities in local cadre/official posts. Such affirmative action policies apply to a wide range of social policies, not least of which is exemption from stringent Chinese family planning policies. Overall, there is clear and strong evidence that the Chinese commitment to poverty reduction is serious and long lasting.

In the Project, the PPs will ensure that SA recommendations are implemented and monitor the impact and effectiveness of actions is monitored. A number of mitigation measures for social impacts have been built in to the various project components. The Local Beneficiaries Action Plans have lists of preferential suggestions for women and minorities, including special credit lines, educational preferences, special training opportunities, and provisions for continuation of religious and cultural customs. For example, in Qinghai, schools and culturally appropriate health services will be established to preserve and promote minority cultures. In addition,families which volunteer to 'move-in' will have the option of returning to their original homes during the first two years, without financial or any other penalty, if they so decide.

Grievance Mechanism

Project beneficiaries and project affected people in all PPs will have access to a consultation and grievance mechanism to address their concerns in regard to the proposed project. The mechanism and procedures have been described to the project beneficiaries and stakeholders. It is believed that this will be particularly useful in the case of land acquisition if farmers feel unsatisfied with compensation implementation or if conflicts arise between farmers and herders. Complaints and grievances can also be raised and reported to the higher level PMOs, the land management authorities, and the external, independent monitoring groups. Beneficiaries and stakeholders also have rights to raise any questions directly to local government and provincial authorities. The right to make lawsuits has been made explicit to the farmers and herders. All complaints, both oral and written, will be recorded including resolution and the time of resolution. External monitoring reports will include a summary of grievances received, if any, and the manner in which they were resolved. 28

6. Environmental assessment: Environment Category: B

Environmental Category

Consistent with relevant laws and regulations of the People's Republic of China and Operational Directive (OD) 4.01, Environmental Assessment of the World Bank, Environmental Impact Assessments (EIA) have been prepared for activities in Gansu, Inner Mongolia and Qinghai. Review of the scope and proposed activities of the Project by the Chinese and the Bank environmental staff indicated that the potential impacts were not unusual or unprecedented in scope and had limited risk of adverse effects that could be addressed through design, mitigation and/or monitoring measures.

The land reclamation and improvement activities supported under the Project are of a kind that have been widely and successfully applied in other arid and semi-arid development projects in the Project Provinces/Regions and other Provinces/Regions in central and western China. Previous interventions of this type have been supported by the Bank under the Tarim Basin I and II Projects. In addition, the proposed activities would be implemented in a large number of dispersed sites with most of the anticipated negative environmental impacts being of local nature. The mitigation measures and methods to be utilized are well known and established in China. It should be noted that during preparation, the overall scale of the Project, particularly the scale of developments in Qinghai Province, was substantially reduced with a corresponding decrease in potential environmental impacts. No pesticides or herbicides will be procured under the Loan/Credit.

Environmental Investigations

The ElAs were prepared by the Environmental Protection Research Institute of Gansu Province, the Environmental Sciences Institute of Inner Mongolia Autonomous Region, and the Qinghai Institute of Environmental Science. They identify key environmental management issues, mitigation actions and environmental monitoring measures with their associated costs, and specify implementation responsibilities. The documents meet Chinese requirements for an EIA and World Bank provisions for activities placed in environmental category "B," which includes an Environmental Management Plan. The EIA work was carried out in close coordination with the Social Assessments, which facilitated broad based consultation on the EIAs with stakeholders during various stages of the preparation process. A summary of the findings and recommendations of EIAs is provided as Annex 10 and English language translations of these reports have been placed in the Public Information Center at the World Bank. Assurances were obtained during negotiations that the PPs would carry out their respective parts of the project in accordance with appropriate environmental management practices acceptable to the Bank (page40(xvi)).

Qinghai - Environmental Management

The key environmental management issues are: protection of soils developed for irrigated agriculture from wind erosion; proper development and management of sodic soils; and sustained availability of surface and ground water resources for use in irrigation and maintenance of aquatic ecosystems.

3 The Projectincludes the constructionof the KeriDam. T-hisdam is a regulatingreservoir located in extremelysparsely vegetated and unutilizeddesert area. As such,the environmentalimplications of its constructionwere determinedto be insufficientto justify assigningthe Projectto a CategoryA. 29

(i) Direct Impacts. The major direct environmental impacts arise from conversion of approximately 19,000 hectares of lower piedmont land to irrigated agricultural land. In their present state, these lands have very limited value as pasture or habitat with a sparse groundcover of less than 5 percent. In addition, the new Keri Dam, a regulating structure, will occupy an area of 128 ha of piedmont land that is at a slightly higher elevation than the irrigation areas and thus is much more gravelly and sparsely vegetated. This reservoir is located "off-river" and hence will not interrupt flow in the Xiangride/Qaidam river system. The environmental consequences of this development are anticipated to be limited. Nevertheless, a panel of experts reviewed the safety aspects of the design of the reservoir. Assurances were obtained during negotiations that Qinghai would recruit qualified engineering consultants by November 30, 1999, to supervise the construction of the Project Dams (page 41 (xvii)). It was also agreed that Qinghai would; (a) conduct periodic reviews of the adequacy of the design and construction of the Project Dams during construction and at the start of operations; (b) report to the Bank on these reviews and any recommended changes to be introduced into the design, construction or operation of the dams; and (c) introduce such changes, taking into account the views of the Bank on the matter. To assist it in these reviews, Qinghai would continue to employ the Dam Safety PaneL Qinghai would prepare and furnish an operation and maintenance plan and an emergency preparedness plan for the Project Dams to the Bank not later than December 31, 2001, and would adopt these plans, taking into account the Bank's views on the matter. Annual inspections of the dams would be undertaken after their completion to identify and rectify any deficiencies in their condition or operation (page 41 (xviii)). (ii) Indirect Impacts. The potential indirect environmental impacts arise from the application of approximately 300 Mcm/year of water to the new irrigation areas, much of which flows into the saline swampy areas within the Qaidam River system. Groundwater modeling undertaken as part of project design indicates that, in the long term (5 to 10 years) the average groundwater level in the swampy areas could be reduced by up to 0.5 meters, immediately north of the Balong irrigation district. It will be necessary to monitor groundwater levels and vegetation distribution to the north of the sub-project area to assess the degree to which these effects materialize and determine the need, if any, for corrective actions. Gansu and Inner Mongolia - Environmental Management

The primary environmental risk associated with the development strategy of activities in Gansu and Inner Mongolia arise from the proposed livestock programs. During the design-phase, livestock development models including feed balances were carefully prepared to ensure that the needed feed and fodder resources to support the proposed livestock production systems will actually be available. Analyses done so far suggest that the resources arising from project investments alone are likely to significantly exceed feed demands. Details of proposed safeguards are outlined in the Environmental Mitigation and Monitoring Plan provided in Annex 10.

Common Indirect Impacts. There are several other potential indirect environmental impacts that- are common to activities in all three PPs (and all other rural developments in China):

(i) Use of Plastic Mulch. Increased use of plastic mulch is becoming an aesthetic problem throughout China, notwithstanding its agricultural benefits. Measures will be taken to promote proper use, waste collection, and recycling of materials. (ii) Small Scale Agro-Enterprises. Some small-scale agro-enterprises could generate organic liquid wastes that require waste minimization and pollution control measures. Administrative steps will be included in sub-loan processing procedures to ensure that loans will not be made without locally mandated environmental clearances. 30

(iii) Agricultural Chemicals. On-farm productivity will be increased, at least in part by increased use of inputs, including fertilizers and pesticides. The farming practices supported by the Project will contribute to the reduction of nutrient losses from fields. World Bank funds will not be used for procurement of pesticides and herbicicdesand it is planned that outreach programs for proper use of pesticides and herbicides will be undlertakenby the Provincial and County Plant Protection Departments.

Environmental Mitigation and Monitoring Plan

The environmental mitigation and monitoring strategy is based on lessons learned from many years of applying environmental management criteria to a wide range of rural development projects not only in China, but also throughout the East Asian region. The approach taken acknowledges that an integral project objective is promotion and financing of environmentally sustainable and/or reduced environmental impact rural production methods.

Environmental Mitigation Measures. The projecl would support the following mitigation measures:

(i) Livestock Management. A variety of production systems already being used in or near the sub-project area will be made available. Township level feed balances must be satisfactorily demonstrated and a "five point" management plan czoncerninggrazing practices and livestock density must be followed. Animal wastes will be managed through environmental review of designs for larger scale production units and adoption of proven designs for small-scale units. (ii) Water Resources and Irrigation Management. Water availability has been evaluated on the basis of historic data, supplemented with mathematical modeling. Potential impacts on groundwater have been minimized by the selection of a conjunctive use option that provides a balance between development and environment criteria. For the Balong Soak, a management plan will be prepared. Standard construction phase environmental safeguards for arid, windy areas will be followed in the main irrigation development area (Qinghai). In the other PPs, irrigation developments will be at the mini and micro scale and no significant construction impacts are expected. (iii) Rural Enterprises. The focus will be on developing small scale and household level enterprises that generally have limited environmental impacts; ABC loan processing procedures provided for larger scale activities (exceeding in value Yuan 250,000 or US$30,000) include requirements for environmental clearances from the relevant Provincial EPBs. (iv) Land Reclamation. Attention would be given to the survey and management of saline and sodic soils. The scope of the Project has been reduced in areas with unfavorable soil types and, for residual areas, suitable procedures would be applied for soil leaching, land leveling and water management. Wind erosion would be limited through strip leveling, shelterbelts, and restriction of construction activities during windy periods. (v) Crop Production. The Project supports extension activities which include standard packages on agro-chemical management. The Project also includes support for adaptive research to optimize usage and promote manuring technologies and extensive soil conservation components to reduce non-point source emissions. (vi) Rural Infrastructure. Emphasis would be on small-scale gravel roads and upgrading of existing roads. Activities would focus on farm access roads, inter-village roads and connector roads that would not increase access to remote areas. Water supply support would focus on small- scale household systems and will include basic safeguards for public health. 31

(vii) Forestry. Species and planting techniques, already proven in project areas would be used. Plantings would include economic forests plus fuelwood areas based on quantitative demand estimates to ensure that participants have an economic stake in protecting trees.

Environmental Monitoring Program. The project includes a monitoring program with the following elements:

(i) Livestock Management. The area and status of grassland surrounding selected development areas in all PPs, will be monitored along with village feed balances. Animal waste management will be monitored by sampling ABC Loan approvals to ensure that the necessary EPB clearances are being obtained, and site inspections would be made of small-scale facilities. In addition, each village project area would prepare an annual livestock and grassland management plan to ensure acceptable feed balances and environmentally sustainable use of grassland. (ii) Water Resources and Irrigation Management (Qinghai). Monitoring activities will be undertaken for both surface and groundwater. This will include continuation of existing flow monitoring and recording of bulk water supply deliveries at Hatu Darn and commencement of recordings at Keri Dam. Surface water quality would be monitored at sample locations upstream, midstream and downstream of the swampy area north of Balong. Pumping and application rates in tubewell supplied areas, groundwater levels in tubewells and wells in the swampy area adjacent to Balong would all be measured. In addition, annual surveys will be conducted of vegetation and wildlife in the swampy area and random field inspections would be made of civil works to ensure that construction safeguards are being observed. (iii) Rural Enterprises. Loan approvals would be randomly checked to ensure that none were made for proposals that lack clearances from the EPBs. (iv) Land Reclamation. Monitoring of soils and drainage water in saline and sodic areas during leaching would be complemented by monitoring of drainage discharges (quantity and quality) from the development areas. Construction would also be supervised to ensure compliance with specifications and inspections would be made of the shelterbelt planting program. (v) Crop Production. The progress of extension and training programs under the Project would be monitored. The adaptive research and soil conservation programs would also be monitored. Input levels would be subject to random samples of plots to compare with relevant national standards (GB-4285-84). Procurement contracts would be reviewed to ensure that no pesticides or herbicides are purchased with Bank funds. 32

7. Participatory Approach: Overview

The SAs provided mechanisms to facilitate stakeholders' including community leaders, and religious elders, capacity to participate in project decision-making and implementation to promote long term sustainability of project interventions. Consultations and collaboration carried out during the SAs provided beneficiaries and stakeholders the opportunity to reflect their views and to participate in the project design at different stages of the planning. This process also supported the preparation of the environmental assessments. Participatory poverty assessments were carried out within the context of the SA work.

To promote stakeholder participation during project implementation, the SAs facilitated extensive interaction between all the stakeholders. The PMOs consulted with a wide range of stakeholders, from farmers and herders on the ground, to administrators at all levels, and exchanged views with other relevant groups. These consultations were launched at the stage of project identification with technical inputs from the Bank on the links between SA and participatory approach on one hand and project design and implementation on the other. Subsequent detailed fieldwork in the project areas (including the 'move-out' and 'move-in' areas in Qinghai) was undertaken during project preparation. Apart from the design institutes and universities assisting in project planning and design, various levels of local government were involved and consulted, including departments of planning, finance, agriculture, forestry, animal husbandry, water resources, education, and health. During the implementation phase of the Project, the affected people will also be consulted by the PMO and the external monitoring agencies.

Stakeholder Categories

Based on extensive surveys and analysis of local socio-cultural, and political conditions, the SA has identified the major factors that affect project impact and success. It classified four major parties as the project key stakeholders (presented in the matrix below): poor farmers/herders as the project intended beneficiaries, government authorities as the project sponsors, the technical line agencies and academic institutions as the project consultants and implementing agencies, and the Bank as the donor. It also analyzed all the stakeholders' participation and their roles in the project operation, so as to develop an appropriate projjectdesign.

Stakeholder Groups Identifiication/Preparation Implementation Primary beneficiaries/community groups IS/CON IS/COL Academic Institutions and experts IS/CON/COL IS/CON/COL Local government IS/CON/COL IS/COL Donor(s) IS/CON IS/CON

IS=Information Sharing, CON=Consultation, COL=Collaboration.

Consultation with Project Beneficiaries

To ensure that the aspirations of the poorest farmers, the primary beneficiaries, are fully reflected in the project design, the SA conducted comprehensive sample surveys in each project county and prepared case studies in a number of villages. These surveys included detailed investigations of the physical resource base, farming, herding, income sources, household composition, community organization, cultural preferences, and farmers/herclers' strategies for coping with poverty. In Gansu, the SA prepared full-scale draft village development plans for two sample villages in collaboration with the beneficiaries. 33

In carrying out baseline data collection and village diagnoses, the SA teams tested the villagers' awareness of the Project. They disseminated project information through various channels, such as radio broadcasts, bulletins, village conferences, focus group meetings, and household interviews. They also collected feedback from farmers/herders' on their perceived problems and priorities, and suggestions on the project design. On the basis of the investigation and consultations, a menu of village/household specific sub-projects was developed, ensuring consistency between project priorities and those of the target population. The detailed processes and results of the sample studies using participatory rural appraisal methodology is presented in the SA reports.

Women's Participation

Special emphasis has been given to women's participation in the project design, preparation and implementation. Women are targeted beneficiaries of the Project. A number of specific project activities have been designed taking into account women's specific needs as identified in the SAs. These activities were designed with direct input from women through interviews, workshops and through representation of the Local Women's Federation. The project management framework has been designed to ensure active women participation. The project implementation procedures require a minimum of 50 percent women attendance in the village conferences when discussing village development plans, and in addition to representatives from the local ACWF, there will be female members in the project management organizations at township and village levels.

Capacity Building for Beneficiary and Stakeholder Participation

In order to ensure beneficiary and stakeholder participation in project implementation and operation, the participatory mechanism has been built into the project management framework. A series of measures were adopted during project preparation so as to develop the institutional capacity to facilitate beneficiary participation. At the village level, VPIGs consisting of farmers' and herders' representatives have been established which, together with the village councils, will assist households in preparing and implementing their respective sub-projects in communication with the TWSs and higher level PMOs. The groups have full responsibility for design and implementation of the village development plans, which are jointly drafted by farmers/herders and technical experts.

The villagers' meetings, the local communication vehicle, are also expected to play an important role in facilitating the interaction between farmers/herders and the PMOs. At all stages in the Project, PMO personnel would participate in villagers' meetings to discuss with and obtain feedback from beneficiaries on sub-projects. Finally, farmers/herders will also be involved in the project monitoring and evaluation, not only as targets of annual sample surveys of Project beneficiaries but as evaluators of project implementation.

The Project is aimed empowering beneficiaries by promoting effective mechanisms for participation. Project design at the village and household level is flexible: farmers and herders are actively encouraged to select and modify the household development models. Project households would be trained in farming and herding techniques, finance, and marketing relevant to the models they have chosen. Moreover, the beneficiaries will be further encouraged to strengthen their participatory commitment and capacity not only by exercising their rights in the design and implementation, but also by carrying out their responsibilities, as prescribed in the Project. Each project household would be issued a Project registration card, which records the households' responsibilities, activities, and evaluation. 34

F: Sustainabilityand Risks

1. Sustainability: Sustainabilityof project benefits would be assured through: i) stakeholders' commitment;ii) appropriateoperation and maintenanceof the irrigation system; iii) appropriatemonitoring of livestockand densityand feed balances;and iv) accessto productioncredit.

Stakeholders'commitment. The central, provincial,prefecture and county governmentshave demonstratedstrong commitmentto poverty reduction. The central government continues to transfersignificant funds to the provincesfor povertyreduction programs which are matched by the PPs. Beneficiaryparticipation in the identification.,design, preparation, and implementationof the Project would also ensureproject sustainability.Following completion of rural infrastructure, the beneficiarieswould take overthe operationand maintenanceof the new assets.

Appropriateoperation and maintenance(O&M) of the irrigation system. In Qinghai, the O&M of the irrigationsystem would require technicalpersonnel. The new migrants,coming from rainfed marginalagriculture production systems, would initially have to rely on the management of the provincial/countywater resourcesteams and overtime assumefull responsibilityfor O&M. Full cost of O&M would be recoveredthrough a combinationof land rents and irrigationcharges after a three-year grace period. The unit irrigationdevelopment cost for Qinghai is the highest among the three PPs. To recover the full costs of: i) annual operationand maintenancecosts; ii) periodicalmajor repairs and renewal costs; and iii) depreciation.Based on the regulationin force governingwater fee collection,the annual irrigation fee for full cost recovery would be about Yuan 70 per mu. Assuranceswere obtained duringnegotiations that Qinghaiwould ensure that irrigation water charges are establishedupon completion of its project irrigation system and collectedat rates sufficientto cover operationand maintenancecosts of this system as they are incurredand a reasonableportion of its capital cosis over a period of not more than 30 years. The scheduleof watercharges and timetablefor its establishmentwould befurnished to the Bank not later than December 31, 1999 for the Bank's review The charges would be reviewed annuallyand adjusted as necessary. Qinghaiwould maintain a separate accountingfor water chargescollected and use these amountsfor the operation,maintenance and capital costs of the system (page41 (xix)).

The per capita investmentcost for piped water supplysystems would range from Yuan 170-300, which is of the same order of investmentcost as that fiorthe Bank assistedNational Water Supply Project. For full cost recovery,the annual water charge per capita would range from Yuan 14-24, representing3-5 percentof per capita income,and is consideredaffordable.

Appropriatemonitoring of livestock density and feed balances.PMO livestock staff at the county levelwill assist(and train, when feasible)the villageswith the preparationand updatingof village-area feed balances. County PMO staff responsible for this task will be trained b-y provincial PMO staff. Local Grassland MonitoringStation (GMS) will be closely involved in feed balance clearance and monitoring. Training for GMS staff, where necessary, will be conductedby the provincialPMO technical staff.

Access to productioncredit. Sustainableproject output requires timely provision of inputs beyond the project implementationperiod. Often,the high transactioncost of small-scalecredit deters formal financial institutions from catering to scattered and remote villages. However, experiencefrom lending to the poor have demonstraLtedthat poor usually are creditworthywith good repaymenttrack records.These farmers,with limitedprior access to credits in the project area, would be eligible to get funds from ABC, acting as a financial agent of the PFBs, for purchaseof fertilizer.This new window would establishpermanent access to future credit needs of such farmers,after the Projectcompletion. 35

2. Critical Risks (reflecting assumptions in the fourth column of Annex 1):

From Outputs to Objective In Qinghai, prospective migrants decide not to relocate N A successful pilot program convinces prospective to 'move-in' area. migrants of project benefits.

Delays in the adoption of environmentally friendly M Agriculture Bureaus provide timely demonstrations of agricultural inputs and technologies. Townships may irrigation technology, cropping, and husbandry practices fail to extend relevant adaptive research and extension on farmers' fields. services to villages.

Sustainability of crop and livestock development S Selection of proven technologies--inputs, irrigation and program is imperiled. Townships receive inadequate land terracing-are appropriate and supported by agricultural extension services. extension services with concentration on feed production to improve productivity of existing herd; ensure that livestock increase is supported by feed availability.

ABC may fail to extend branch services to project M ABC has joined the Provincial and County PLGs and townships. PMOs signed MOUs.

Completion of Qinghai's main electric grid line to M Completion date is planned for 1999, the first year of the project area may be delayed. WPRP. The village electrification component and the tubewells are scheduled to start in the second year of project implementation.

Prison labor is used directly or indirectly to work on N Heads of migrant households would be hired for project Qinghai portion of project. works. Government has provided written confirmation that no prison labor would be utilized, and this will be monitored on a regular basis.

Resettlement Action Plans (RAP) are poorly M RAP monitoring and evaluation process is carefully implemented; compensation is inadequate or delayed. implemented and monitored; with reliable reporting and feedback.

In Qinghai: i) the voluntary settlement program may S i) The responsible agencies in the 'move-out' and 'move- face problems in the implementation anrangements and in' counties have been established to carefully plan the frustrate the migrants; ii) conflicts may arise between details for implementation. A pilot settlement program the new migrants and the local population in the will be implemented to test the arrangements, and lessons 'move-in' area, as well as inter-ethnic conflicts among learned would be used to modify settlement the migrants may arise; iii) some nationalities may feel implementation plans; ii) The project would ensure that that benefits of the project are unfairly distributed the 4000 herders/farmers in the 'move-in' area would among the nationalities. benefit equally from the project and have equal access to all the services to be provided; their preference for herding or farming and cultural values are respected; the grievance mechanism is operational and effective; pubfiS safety is provided; and the settlement monitoring system provides timely feedback to project management; and iii) VPIGs will rely on respected community and culturalreligious leadership for suggestions; 36

RikRisk atin. Risk Miniiaton Measure Qinghaiherders suffer thefts of livestock,lack of water M Herdersguaranteed access to water,re-seeded access,further degradation of pastureland,and unequal pastureland,housing plot allocations,improved security, allocationof projectbenefits. and grazingcorridors through the projectarea.

FromComponents to Outputs Delaysin the approvaland releaseof counterpart S Dedicatethe PPs' shareof centralgovernment poverty funds. fundsto the proposedproject; open Counterpart Special Accounts(CSAs) in Qinghaiand InnerMongolia.

Weakco-ordination of provincialline agenciesby M PMOshave signedMOUs with line agencies.PMOs PMOs. would seek quarterlymeetings with the proyincialPLGs for briefingon implementationprogress, and decisions; FCPMCwould superviseproject and brief the provincial PLG meetings.

VPIGsmay be weakto promotebeneficiary M Provideadequate incentives-salary support and training- participationin all aspectsof the project.Village to membersof VPIGs. communitiesfail to mobilizefunds to meettheir share of projectfinancing plan.

OverallRisk Rating M/S RiskRating - H (HighRisk), S (SubstantialRisk), M (ModestRisk), N (Negligibleor LowRisk)

Oinghai. In Qinghai the complexity of the project calls for detailed planning and phasing of the multitude of activities. Coordination and cooperation among the implementing agencies in both the 'move-out' and the 'move-in' areas will be critical if success and timeliness are to be achieved. The first issue is whether the management skills and experience are available within the line agencies to successfully implement the program. The second issue is whether satisfactory coordination between the relevant authorities in both the 'move-out' and 'move-in' areas can be achieved thus mitigating any potential etlnic conflicts or conflicts between farmers and herders. The third issue is whether the required level of cooperation and coordination between the technical organizations can be achieved. The fourth issue concerns the ability to implement the project so that its benefits are spread equally among the participating nationalities and among both herders and farmers.

These issues would be addressed through the implementation of a pilot settlement program, which would test and identify weaknesses and bottlenecks in the planning, coordination, implementation and management systems. The pilot settlement scheme would involve one village of about 200 households. By using groundwater and diesel engine driven pumps, the irrigation, land and forage development can proceed wvithoutthe major irrigation system having first been put in place. The pilot program would also include participatory evaluation techniques ensuring that feedback from all nationalities and from both herders and farmers to be used for program modification. During appraisal, the pilot program was approved for retroactive financing. The pilot program would provide a valuable learning experience for all those involved and is hoped to result in acceleration of the implementation of the whole program which would start only after thorough evaluation of the pilot program by independent reviewers.

Gansu and Inner Mongolia. The key issues in these two PPs are first, the sustainability of the technical packages proposed and second, whether the risk factor of recurring drought on the yield and income assumptions has been adequately considered in the technical packages. All of the proposed technical packages are already being adopted by farmers. This suggests that farmers find the technologies profitable and the risk largely acceptable. The effect of the Project would be to accelerate the rate of adoption by providing poorer households with access to the 37 necessary resources. Risk aversion strategies adopted by the great majority of farmers in rain-fed cropping systems, ensure that, in most years, they do not apply all of the recommended fertilizer rates. While wheat is likely to remain the dominant crop for some time to come, most farmers grow a variety of crops, which offsets the risk of drought or other adverse weather effects on crop production. The drought factor cannot be entirely eliminated by the project activities. To the extent possible, however, expansion of the irrigated area will remain the priority throughout both PPs. Moreover, the effect of risk on resource poor farmers and how it affects the proposed technical packages will be a priority training area for all technicians. Crop budgets and cost- benefit analysis and a wider range of sensitivity analyses applied to farm models will also assist in defining more precisely the level of risk involved in the proposed technical packages. The adaptive research program being proposed under the project is designed to refine existing technologies so as to ensure they are both economically and environmentally sustainable. Similarly, forage and crop byproduct production and storage will determine how much livestock can be sustained in the project areas.

Counterpart Funds. A common risk in the Bank's China portfolio is that counterpart funds are not released on time, resulting in implementation delays. To address this recurrent issue, Inner Mongolia and Qinghai plan to establish in each project county, CSAs similar to the Special Accounts to be opened in the PPs. Initial deposits, equivalent to three months of expenditures would be deposited into these CSAs. Gansu has an existing mechanism to allocate poverty funds directly to the counties. 38

G: Main LoanConditions 1. Effectiveness condition: Standard

2. Other: During negotiations, assurances were obtained on the following:

(i) The Borrower would ensure that all regional governments employing surplus labor under Gansu Labor Mobility program implement the program in a manner designed to ensure safe and equitable working and living conditions for the laborer (page 6)

(ii) PPs would develop and implement their medical financing systems not later than September 30, 2000 in accordance with policies and procedures acceptable to the Bank (page 7).

(iii) The Borrower would continue to maintain FCPMC to be responsible for Project oversight, overseas training coordination, and poverty impact monitoring, and the PPs would carry out all training, studies and research under the Project in accordance with a program acceptable to the Bank. (page 9).

(iv) PPs would maintain provincial PMOs, county PMOs, TAGs, and environmental experts (page 9).

(v) Starting in 1999, each PP would annually, not later than December 1,furnish to the Bank its annual implementation plans for the coming year and that the annual implementation plans would include all activities to be implemented in the coming year, including training, studies and research, under its part of the Project and be accompanied by a financing plan for the activities (page 10).

(vi) Assurances were obtained during negotiations that sub-projects would be appraised in accordance with policies and procedures acceptable to the Bank and sub-loans would be made on conditions acceptable to the Bank. Lending terms for sub-loans to households and rural enterprises would be the applicable ABC lending terms. The PPs would engage the ABC to assist in appraising and processing fertilizer and rural enterprise loan applications. Sub-projects would be eligible jfor financing under the Credit/Loan only if the PP has determined that the following condlitionshave been met, and, in the case of a rural enterprise subproject in Gansu, has been approved by the Bank: (a) the beneficiary is creditworthy and has the required technicai'skills and resources; (b) the sub-project is economically andfinancially viable, and commercially and technically feasible; (c) if the sub-project is for livestock, the beneficiary has adequate sources offorage or grain feed for the livestock, and the subproject is to be carried out in a village which has developed a livestock development and grassland management plan; (d) in the case of a rural enterprise subproject in Gansu, less than Y/2 of the beneficiary's proprietary interest is controlled by the government, and the beneficiary has undertaken to contribute at least 30% of the total estimated cost of the subproject in the form of equity, and the subproject is designed to alleviate poverty through job creation, provision of affordable inputs or Technical Assistance to poor households, or provision of profitable markets for these households; (e) the subproject is in compliance with environmental standards acceptable to the Bank, as well as Chinese environmenjlal regulations, and in the case of a rural enterprise subproject in Gansu, has been developed on the basis of an EIA approved by the local EPB. Rural enterprise subprojects in Qinghai would be appraised, financed 39

and supervised by ABC in accordance with its standard policies and procedures for poverty alleviation loans (page 11).

(vii) The Borrower would open three Special Accounts in banks and on terms and conditions acceptable to the Bank (page 13).

(viii) PPs wouldfurnish audited project accounts to the Bank within six months of the close of each fiscal year (page 14).

(ix) The PPs wouldfurnish progress reports semi-annually to the Bank not later than January 31 and July 31 of each year starting in 2000, and a mid-term Report not later than January 31, 2003, and an Implementation Completion Report not later than six months after the Closing Date of June 30, 2006 (page 15).

(x) The PPs would furnish, not later than December 1 of each year, annual poverty reduction, and environmental impact monitoring and evaluation reports to the Bank. (page 15).

(xi) Assurances were obtained during negotiations that Qinghai would take all measures necessary to ensure that the migration of all households to the move-in area will be implemented on a voluntary basis and in a manner designed to improve the living standards andproductivity of all households in both the move-out and move-in areas. To this end, Qinghai would carry out the emigration in accordance with the VSIP. In particular, it would allocate irrigated farm land in the move-in area to each migrant household and each household currently living in or using land in the move-in area in amounts and under arrangements acceptable to the Bank, including a written lease agreement for a minimum of 30 years on terms acceptable to the Bank and exemption from agricultural taxes for the first 3 years of the lease. Qinghai would also implement its Local Beneficiary Participation Action Plan in accordance with an implementation plan acceptable to the Bank4. Under this implementation plan, Qinghai would take a number of measures to: (a) ensure social stability in the move-in are; (b) ensure that pastoralists are able to maintain their preferred methods of livelihood; (c) promote culturally appropriate educationfor minority nationalities in both move-out and move-in areas; (d) ensure adequate, culturally appropriate access of people in both areas to health care services; and (e) integrate cultural preferences of all nationalities involved in the migration. In addition, prior to any migration to the move-in area, Qinghai would carry out a study to update its assessment of the social needs of both migrating households and households living in or using land in the move-in area and recommend measures to update the VSIP on the basis of this study. Once the VSIP is updated to take into account the Bank's views on the study and its recommendations, Qinghai would undertake a pilot voluntary settlement program to move up to 200 households in one- village into the move-in area Prior to undertaking any further settlement in the move-in area, Qinghai would report to the Bank on the results of the pilot program and revise the VSIP to take into account the views of the Bank on the report and its recommendations. Qinghai would: (a) monitor and evaluate implementation of the VSIPin accordance with indicators acceptable to the Bank; (b) prepare and furnish an annual report on the implementation of the VSIP to the Borrower and the Bank not later than January 31 of each year for review; and (c) based on the report, revise the VSIP in agreement with the Bank. Qinghai would employ independent experts with terms of reference, qualifications, and experience acceptable to the Bank, not later than December 31, 1999, to assist in monitoring and evaluating the VSIP. Finally, within 3 years after commencement of VSIP

4The implementationplan is set out in a supplementalletter to the ProjectAgreement, which is includedin the packageof the legalagreements made available to the Bank's ExecutiveDirectors. 40

implementation, Qinghai will: (a) carry out a study to evaluate the social and environmental impact of the VSIP on the move-out counties and to recommend measures to enhance environmental sustainability and living conditions in these counties; and (b) implement the study's recommendations, taking into account the Bank's views on the matter (page 16).

(xii) Gansu would take all measures required on its part to ensure that the labor mobility program is implemented on a voluntary basis and in a manner designed to ensure safe and equitable working and living conditions fior the participating laborers. Gansu would monitor and evaluate this program in accordance with indicators acceptable to the Bank, and would prepare, with the assistance of independent experts employed not later than December 31, 1999, an annual report cover^ing, inter alia the employment selection process, the safety, treatment and living conditions of the placed workers and the impact of the remittances on the workers 'families, ald furnish this report to the Bank not later than December I of each year for review. The labor mobility program would be revised, based on the report's recommendations, and taking into account the Bank's views on the matter (page 16).

(xiii) The PPs would ensure that each village in their Project areas would prepare and implement annual livestock and grassland management plans to ensure acceptable livestockfeed balances and environmentally sustainable use of grassland The PPs would ensure that all village grassland used under the Project is leased to the users for a minimum of 30 year under conditions designed to ensure the land's use in accordance with the grassland management plans. The PPs would monitor and report to the Bank annually on the livestock balances in these villages and adjust their plans to ensure continued acceptable livestock feed balances and environmentally sustainable grassland use, taking into account the views of the Bank on the matter (page 16).

(xiv) Each PP would ensure that all persons affected by the Project fully participate in the Project's design and implementation and receive social and economic benefits from the Project that are compatible with their cultural preferences. To that end, the PPs would: (a) carry out the Project in accordance with their Local Beneficiary Participation Action Plan; (b) monitor the implementation of these plans in accordance with indicators acceptable to the Bank; (c) report to the Bank on the results of this monitoring; and (d) introduce revisions to these plans, as agreed with the Bank in order to achieve the plans' objectives (page 22).

(xv) The PPs would carry out their respective RIAPs in a manner designed to improve the living standards and production levels of all persons suffering involuntary loss of shelter, productive assets, access to productive assets, income, or means of livelihood as a result of works under the Project that cause a change in land or water use. The PPs would: (a) monitor the implementation of their RAPs in accordance with indicators acceptable to the Bank; (b) report to the Bank on the results of this monitoring; and (c) introduce revisions to the RAPs, as agreed with the Bank in order to achieve the plans' objectives. If any sub-projects financed out of sub-loans would involve resettlement, the resettlement would be carried out in accordance with a resettlement plan which has been (a) prepared on the basis of a policy framework acceptable to the Bank and (b) approved by the Bank (page 24).

(xvi) The PPs would carry out their respective joarts of the Project in accordance with appropriate environmental management practices acceptable to the Bank (page 28). 41

(xvii) Qinghai would recruit qualified engineering consultants by November 30, 1999, to supervise the construction of the dams (page 29).

(xviii) Qinghai would: (a) conduct periodic reviews of the adequacy of the design and construction of the Project Dams during construction and at the start of operations; (b) report to the Bank on these reviews and any recommended changes to be introduced into design, construction or operation of the dams; and (c) introduce such changes, taking into account the views of the Bank on the matter. To assist it in these reviews, Qinghai would continue to employ the Dam Safety Panel. Qinghai would prepare and furnish an operation and maintenance plan and an emergency preparedness plan for the Project Dams to the Bank not later than December 31, 2001, and would adopt these plans, taking into account the Bank's views on the matter. Annual inspections of the dams would be undertaken after their completion to identify and rectify any deficiencies in their condition or operation (page 29).

(xix) Qinghai would ensure that irrigation water charges are established upon completion of its project irrigation system and collected at rates sufficient to cover operation and maintenance costs of this system as they are incurred and a reasonable portion of its capital costs over a period of not more than 30 years. The schedule of water charges and timetable for its establishment would be furnished to the Bank not later than December 31, 1999 for the Bank's review. The charges would be reviewed annually and adjusted as necessary. Qinghai would maintain a separate accounting for water charges collected and use these amounts for the operation, maintenance and capital costs of the system (page 34).

H. Readiness for Implementation [X] 1. The engineering design documents for the first year's activities are complete and ready for the start of project implementation. [X] 2. The procurement documents for the first year's activities are complete and ready for the start of project implementation. [X] 3. The Project Implementation Plan has been appraised and found to be realistic and of satisfactory quality. 42

1.Compliance with Bank Policies [X] This project complieswith all applicableBank policies.

The project is in compliancewith: i) OD4.01, Environmental Assessment, with adoption of an EnvironmentalMitigation and Monitoring Plan; ii) OD 4.20, IndigenousPeople with a project design that reflects the concerns of minority nationalities as reported in the Social Assessments;iii) OD 4.30, InvoluntaryResettlement involving land acquisitionin the three PPs; iv) OD 4.37, Dam Safety with the appointmentof a Dam Safety Review Panel in Qinghai and implementationof recommendationsmade by the Planel;v) BP 10.02, Financial Management, with the adoption of a FinancialManagement System; and vi) OP 11, Procurement,with the completionof a ProcurementAssessment. No pesticidesor herbicideswould be procuredunder the Loan/Credit.Documentation concerning environment and resettlementaspects of the Project was not providedto the Public Infornation Center (PIC) priorto appraisalas mandatedunder the Bank public disclosureguideline. These documents are nowavailable at the PIC.

2~~~~~~~~~~~~~

eros Ak lu Geo B.Fox ukon Huang - t Team Leader Secto& anager Country Director 43

Annex 1: Project Design Summary

CHINA: Western Poverty Reduction Project

Povertyalleviation through . Provincial...... population below .Annual Reportissued by SSB Chinais ableto avoid any major economically,environmentally povertyline reduced. basedon rural household economicand financial crises and socially sustainable rural surveys. and maintain its poverty development. Illiteracy, infant and maternal reduction priority in the western rates reduced. Project environmental and regions. groundwater monitoring reports. Crop lands and water quality in Government extends the 8-7 Pfan project areas meet standards set (1994-2000) beyond 2000 and by the Provincial EPBs. maintains the current level of funding in real terms. Conversion of degraded lands into artificial pastures and other China continues to pursue its perennial crops. education and health targets to be achieved by 2000.

Provincial EPBs have sufficient capacityto monitorand enforce theenvironmental standards and _ _ _ _ ~~~~~~~guidelines. inChronicproject povertyareasalleviated. of absolute poor cropIncreasedand livestock. marketable surplus of Annual basedon MEsample report issued Agricultural prices provide controlgroupsurveys.project and adequate productionincentives. employmentAltemative off-farmopportunities Financial aid for education and created. reports.Six-monthlyproject progress households.healthservices provided topoor Primaryschoolenrollment andBank review mission reports. Gansu andInner Mongolia emphasiscompletiononratesfemale increasedand with Mid-termn Review. educationimplementprojectstheirpipelinesuccessfully. minority students. Project Implementation Completion Report. Access to affordable health services increased.

Farners'/herders' access to markets and other social services enhanced. 44

NarrativeSummary KeyPe rman i and CriticalAssumnptions Indicators Evaluation Outputs: . .Outputs ( to Objectives) 1. Land andHousehold Development 1.I Cropactivities enhanced. Annualproduction of 269,000 AnnualProject Implementation Farmersremain eager to moveto tons of grain,150,000 tons of Plan. newlyirrigated lands. cash crop products,70,000 tons of fruitat full development. Annualaudited project accounts. Farmersadopt packages of environmentallyfriendly 1.2 Fisheryand livestock Annualproduction of 40 tons of Bi-annualprogress report. agriculturalinputs and farm productionimproved. fresh wateraquatic products, 3.2 technologies. millionheads of animalsand 3 Annualpoverty reduction and milliontons of other animal environmentalimpact Livestockproduction undertaken productsat full development. monJitoringand evaluation on the basis of agreedactions. report. 1.3Tree plantation completed 1.5 millioncubic metersof GMSsmonitor village feed timberat finalcut. Annuallivestock and grassland balances. managementplan by each village. Townshipsprovide relevant agriculturaladaptive research Banl;mission review reports. and extensionservices to villages. Mid-.termreview report. ABC extendsbranch services in ImplementationCompletion projecttownships. Report. Qinghaicompletes the main electricgrid line to Xiangride town.

Sustainabilityof Gansucrop and livestockdevelopment program. 2. Irriyationand Land Improvement Qing,hai'sannual water charge 2.1 Storagereservoirs constructed. One new reservoirof 40 m report. heightand 26 millioncubic meterstorage constructed and an QinghaiDam SafetyPanel's existingreservoir of 10 m height reportand plans. renovatedby 2003.

2.2 Wastelandsreclaimed and 19,200ha of new landsand rainfedlands irrigated producing 34,050ha of rainfedlands grain,fruits, vegetables and fodder irrigatedfor agricultural for livestock. productionby 2005. 45

3. Rural Infrastructure 3.1 New rural roadsconstructed. 216 km of classIII/IV roads ProjectProgress Reports. Implementationof the constructedby 2003. resettlementaction plans is effectiveand satisfactoryto all 3.2 New drinkingwells 5,060of shallowwells, 821 deep parties. constructed. tube wellsand 5641shallow tube wellsconstructed by 2003.

3.3 New electricdistribution lines 410 villagesconnected and Qinghaicompletes the main extendedto projectvillages. receivingreliable electricity by electricgrid line to Xiangride 2004. town. 4. RuralEnterprises

4.1 Smallscale and household- Numberof facilities ABC reportsto PMOs. Non-statesector enterprises basedrural enterprisesestablished constructed/upgradedby 2005. applyfor agro-processingloans.. and or upgraded. 5. LaborMobility

5.1 Laborexport agencies 65,700 surplus rural workers Gansu'sannual labor mobility Demandfor labor in Gansuand strengthened. voluntarily placed in off-farm report. otherprovinces identified. employmentby 2005. 6. VoluntarySettlement

6.1 Settlementplan established and 57,800people voluntarily settled Qinghai'sVSIP implementation All nationalitieshave "buy-in" to operational. in Balong,Qinghai and 4,000 report. settlementplan. herders/farmersvoluntarily participatein the Projectby Potentialpublic security and 2006. inter-ethnicconflicts are anticipatedand mitigatedin Twonew townshipsand 21 'move-in' area. villagesestablished. HerdersweTcome opportunity to pursuemixed herding/farming strategy. 7. SocialSector Services

7.1 Primaryschools and basic 21 primary/secondaryschools ProvincialEducation and Health QinghaiProvincial Government healthcenters established and and 23 healthcenters established BureausReports' to PMOs. releasesannual budget on time. upgraded. by 2005. 46

8. InstitutionalBuilding & ProjectManagement

8.1 ProjectManagement Offices ThreeProvincial, six Prefecture, ProgressReports. Governmentat differentlevels establishedand staffed;and and 34 CountyPMOs carriesout MOUseffectively. Memorandaof Understanding establishedand fullystaffed by signedwith LineAgencies by mid mid 1999.Annual 1999. ImplementationPlans on schedule.

8.2 PMOsat provincial,prefecture MEDplan adopted by 2000. and countylevels, and township projectworkstations and village projectimplementation groups established.

8.3 TechnicalAdvisory Groups Trainingprogram launched and ProgressReports. PMOsassign training specialists established. full coverageof projectstaff by on time. 2002and farmers/herdersby 2004.

8.4 PovertyImpact Monitoring, SSB Report. SSB assignsexperiences staff. Evaluation,and Dissemination (MED)plan prepared.

8.5 Projectstaff, farmers/herders trained. 47

X~~~~~~~~~~ , A. Land and Household Total: 141.0 mill; IBRD/IDA: Project Launch Workshop Central, provincial, prefecture Development 85.0 mill; local co-financing: Report. and county governments approve 56.0 mill. and release project counterpart 1. Crops 44.0 mill. Six-monthly Project Progress funds on time. 2. Livestock and Fishery 63.0 mill. Report. 3. Forest and Grassland 23.7 mill. Chairs of Provincial Poverty 4. Support Services 10.3 mill. Bank supervision mission Leading Groups ensure B. Irrigation and Land Total: 86.3 mill; IBRD/IDA: reports. execution of PIP. Improvement 42.2 mill; local co-financing: 44.1 mill. Mid-term Review. Village committees mobilize Implementation Completion resources (cash and in kind) to C. Rural Infrastructure Total: 28.7 mill; IBRD/tDA: Report. meet share of project financing- 11.9 mill; local co-financing: plan. 16.8 mill. 1. Rural Road 5.6 mill. Village committees encourage 2. Drinking Water Supply 8.7 mill beneficiaries participation in all 3. Rural Electrification 14.4 mill aspects of the project. D. Rural Enterprises Total: 13.6 mill; IBRD/IDA: 6.8 mill; local co-financing: ACWF provides institutional 6.8 mill. support to village committees to ensure equitable female E. Labor Mobility (Gansu) Total: 15.7 mill; IBRD/IDA: participation. 7.7 mill; local co-financing: 8.0 mill.

F. Voluntary Settlement Total: 6.5 mill; IBRDIEDA: (Qinghai) 0.1 mill; local co-financing: 6.4 mill.

G. Social Sector Services Total: 8.7 mill; local co- financing: 8.7 mill. 1. Education 3.6 mill. 2. Health 5.1 mill. H. Institutional Building & Total: 10.6 mill; IBRD/IDA: Project Management 5.7 mill; local co-financing: 4.9 mill. 1. Institutional Building 6.5 mill. 2. Independent Monitoring 1.3 mill. 3. Project Management 2.8 mill. TOTAL PROJECT COST 311.1 million

IBRD/IDA 160.0 million

Local co-financing 151.1 million 48

Annex 2: Project Description

CHINA: Western Poverty Reduction Project

A. Overview

Project Development Objective

The objective of the proposed Western Poverty Reduction Project (WPRP) is to alleviate chronic poverty for the absolute poor living in remote and inaccessible villages of Gansu and Qinghai Provinces and Inner Mongolia Autonomous Region (hereinafter referred to as Project Provinces - PPs). The project would enable poor households in the selected poor counties to raise their incomes through increasing grain and livestock production sufficient to meet their basic food and clothing needs and, in many cases, also generate marketable surplus to improve their living standards. Access to health, education and other rural services would also be improved for the poor households. The development objective has emerged from the expectations and aspirations of the target population surveyed during project preparation. The proposed project is the fourth in a series of Bank-assisted, targeted poverty reduction projects in China supporting the Government's Eight-Seven Poverty Reduction Plan (8-7 Plan).

Project Components

The project includes eight complementary components with the following budgetary allocations (for a total of US $311.1 million): Component 1 - Land and Household Development (U$$ 141 million) Component 2 - Irrigation and Land Improvement (US$ 86.3 million) Component 3 - Rural Infrastructure (US$ 28.7 million) Component 4 - Rural Enterprises (US$ 13.6 million) Component 5 - Labor Mobility (US$ 15.7 million) Component 6 - Voluntary Settlement (US$ 6.5 million) Component 7 - Social Sector Services (US$ 8.7 million) Component 8 - Institutional Building and Project (US$ 10.6 million) Management

B. Component 1 - Land and Household Development (US$141.0 million)

Goal and Strategy

The goal of the land and household development component is the promotion of environmentally sustainable agricultural and rural development as the primary means of reducing poverty among the target population. The development objective would be to put in place the institutional arrangements and management and operational systems that would give the target 377,800 project households greater access to productive land, water, inputs and services needed to effect sustainable agricultural and rural development, and to improve their economic and social well- being. The broad strategy proposed under the project in Gansu and Inner Mongolia would aim at progressively changing, over time, the farm socio-economic system from one which is almost entirely focused on food self-sufficiency to one that stresses food security and maximizing cash income. An important project goal of the Land and Household Development component would be the promotion of agricultural land use sustainability by shifting free-ranging or herded 49

livestock-an important cash resource for poor households-from grazing on deteriorating natural pastures to cut-and-carry feeding using crop byproducts, forages, and home-grown grain products. As productivity and incomes increase, steeper slope land and other areas subject to severe wind erosion and not suitable for sustainable crop production would be progressively retired from cultivation and sown to forage legumes and grasses, or protective trees. The consequent improvement in soil and water conservation and progressive reduction of soil erosion (wind, sheet, gully) would also, over time, contribute towards raising sustainable crop production. In Qinghai the strategy would be to reduce population pressure in the six poverty counties in the mountainous eastern region through voluntary settlement of about 57,800 persons on newly developed irrigated land. This would also improve the living standards of those who remain behind through increasing farmland and thus income and grain production per capita. Simultaneously, steep land in these areas would be progressively retired from cropping thus reducing continued soil erosion and environmental degradation.

Selection of Project Activities

Beneficiary villages and households participated in the determination of a 'menu' of 45 different land and household development component activities; they will continue to play the key role in actual selection of these activities during implementation. The development of several activities by a single household or single village would promote improvement of the households' farming systems. The activities include both short-term and longer-term investments, such as integrated crop and animal production, and tree crop development. The project combines the activities into individual village development plans, the design of which has reached an advanced stage of preparation. The average investment per household is sufficient to give each household in each of the 3,171 project villages the opportunity to participate in the component activities. An upper limit to the investment per household has been established to further ensure access to project benefits by all households.

Component Activities

To achieve desired project outputs in accordance with the development objective, the investment activities would include: i) intensifying crop production by expanding the use of proven improved and environmentally-friendly technologies to economically raise crop yields; ii) introducing and expanding cash crops--vegetables, tree crops, fruits; iii) converting from free-range to confined- feeding production (sheep, goats, cattle, pigs, rabbits, poultry) through small-scale, integrated crop/livestock production; iv) in Gansu and Inner Mongolia, expanding gravity and groundwater irrigation systems and rainwater collection cisterns, and improving soil and water conservation on slope land through terracing and other conservation measures, and in Qinghai constructing a new irrigation system to develop 19,200 hectares by reclaiming desert wasteland for an integrated oasis agricultural and forestry production system (see Irrigation and Land Improvement Component); v) improving the quality of farm level technical services through skill upgrading of County, township and village technicians; and vi) developing improved and more economical and environmentally sound technologies through adaptive research and on-farm demonstration.

Crop Production

The crop production sub-projects would include investments in improved technology to increase food, cash, and tree crops production. In Inner Mongolia the main focus would be on food crops (maize, wheat, potato, rice, buckwheat, soybean, sunflower). Gansu would include food crops (maize, wheat, potato, broad bean) and cash crops (vegetables, hemp and Chinese medicinal herbs). 50

The project would promote improved food crops technology packages for 30,900 hectares in Gansu and 87,200 hectares in Inner Mongolia. It would supply households with the first year's requirements of quality inputs (seeds, plastic mulch, and pesticides) for corn (36,700 hectares), wheat (1,400 hectares), and potato (32,000 hectares). Improved technology packages would also be provided for irrigated corn (1,900 hectares), irrigated and dry land wheat (30,600 hectares), buckwheat (6,300 hectares), soybean (3,200 hectares), broad bean (3,090 hectares), hybrid rapeseed (1,980 hectares), wheat/corn inter-crop (580 hectares), and paddy (330 hectares). In Gansu, about 9,000 hectares of cash crop development would be supported. Hemp (1,040 hectares) and four species of medicinal plants and tubers would be introduced in project villages on 7,780 hectares and 220 hectares of commercial vegetables would be developed.

Production investments in Qinghai would include the initial land development (clearing, leveling and cultivation) for crops, forage and tree planting on 19,200 hectares; soil fertility improvement (green manure crop before crop production begins); investments for the first crops to be sown by the migrants (seed, fertilizer, plastic mulch, pesticides, etc.) and establishment of an area (679 hectares) for the production of various crop and forage grass seeds; two technical service centers, applied research and demonstration. The main crolps to be grown include grain (wheat and barley), potatoes, rape, beans and vegetables (onions, garlic, cabbages). Forage crops (astragalus, vetch, elymus, trifolium, oats) would be grown as an integral part of crop rotation.

Livestock Production

Qinghai. Some 13,600 households would be settled in the project area and engage in irrigated crop and forage production. Nine livestock models have been prepared (Table 2-1), based on different household size and cropping patterns. Winter forage will consist of ensilaged greenfeed and byproducts and ammoniated straws and stalks. Oilseed cake, wheat bran, and barley will be available for grain feeding. Livestock production in the Qinghai settlement area would be based entirely on the stall-fed, cut-and-carry feeding system. No natural grassland would be utilized. Pig breeding and fattening, sheep and cashmere goat breeding and fattening, and household chicken and egg production, would constitute the 'menu' of investments. The majority of households would raise a small number of animals, largely for household consumption but with one or two surplus animals sold for cash income. A number of households would, however, specialize in one type of animal production system such as pig fattening, pig breeding or sheep/goat breeding. Feed and forage resources would be derived from irrigated forage and crop lands and would include ammoniated crop straws (a technology widely used on small farms in China), grain byproducts (bran), farm-grown forage, and, where required, purchased or home produced concentrates. Small forage areas would be planted on each farm while specialized forage production farms (about 2,880 hectares in total) would be designated within the settlement area. Each household would be allocated land (about 0.256 hectares) including the forage production farms. Forage would not include alfalfa for which sufficient precipitation is not available, but would include more hardy forage varieties, including forage oats, astragalus spp., medicago spp., elymus spp., and other anid grass species. A conditionality of household-based livestock financing would be a proven self- sufficiency in forage or grain. Livestock numbers would be based on feed balance calculations. Grazing on the natural grasslands surrounding the settlement area would be prohibited. Local Grassland Monitoring Stations (GMS) supervised by Provincial Project Management Offices (PPMOs) and Provincial Environmental Protection Bureaus (EPBs) would spot check the accuracy of feed balances and compliance with confined feeding.

The project would finance low-cost, adobe and plastic cover livestock sheds, ammoniation pits, capital stock purchase, forage establishment costs and, for specialized households, one third of the cost of initial feed purchases. A number of households would specialize in forage processing with small low-cost straw and chaff cutters. At full production in project-year twelve incremental livestock population in the Qinghai settlement is expected to be about 15,600 draught animals, 51

64,000 cashmere goats, 116,000 fattening sheep, 23,000 pigs and some 221,000 poultry. Annual output is estimated to be about 1,600 tons of pork, 3,300 tons of mutton/goat meat, 400 tons of wool, 23 tons of cashmere, 1,030 tons of goat milk, and 1,240 tons of eggs.

Gansu. The livestock sub-project in this province would be implemented under two conditions that differ from those in Inner Mongolia. The Gansu component is designed to strengthen integrated crop/livestock production system (ICLS), based on household production, rather than greatly expanding ruminant livestock numbers. First, crop production, on which project livestock will depend heavily for byproduct and planted forage feeding, faces harsh environmental conditions in the form of frequent droughts and short growing seasons. This can lead to periodically depressed crop yields during drought years which may increase forage shortages.with which to sustain livestock. Second, production technologies in Gansu are not as far advanced, requiring more extension, demonstration and training inputs to make herders/farners more familiar with ICLS. It is the cropping system under this sub-project, rather than grassland integrity, which would be the limiting factor with respect to ensuring acceptable crop yields and adequate animal forage supplies. As a consequence, planned incremental project livestock numbers have been kept relatively small, and the project will focus more on establishing and refining ICLS-and having this operate more efficiently-than on producing more livestock.

Livestock development interventions in Gansu include pig and cattle fattening, sheep and goat production by individual households to supplement farm crop income, mixed livestock production systems using remaining natural grassland areas, commercial large scale breeding pig and cattle fattening production operations (to be financed under the rural enterprises component), soft-shelled turtle production, and fish production. Most feed for ruminant animals and pig production and turtles and fish production would be obtained from crop byproducts and manufactured feed sources and have little additional impact on natural grassland ecosystems. Fodder legumes would be fed directly to livestock or processed into higher quality feeds. Byproducts of maize and broad bean and ammoniated grain straw are useful feed. A large amount of surplus crop byproducts and the technology to improve feed quality are available. Based on 1999 forage balance determinations the project would have, at full production, a 42 percent forage surplus over incremental livestock demand.

In Gansu, grazing on natural grasslands is planned in three of the twelve project counties. This represents a total of 8,050 hectares (of 43,200 hectares) of natural grassland area (19 percent) in these counties. These grasslands are located in southern Gansu occupying valleys in hill country. Crop cultivation which took place on the slopelands adjoining these grass-covered valleys has caused erosion in the past but will now be stopped as the result of newly revised and strengthened grassland management regulations. Rainfall in these areas ranges from 400-750 millimeters/year. The project would rehabilitate these abandoned slope-croplands by providing ground cover with cultivating alfalfa on the higher, 15-25 degree slopelands. Cropping would then be confined to base (terraced or non-terraced) slopeland only, which directly adjoins the valley grasslands below-. the cultivated forage slopeland above.

The project grassland itself remains in good to excellent condition, mainly as the result of a rotational summer grazing program (from June-September only) which is already in place. Grazing is by sheep only, and at acceptable stocking densities. Land deterioration in this area is caused more by crop production on slopelands than by grazing. The project would make a positive impact by relocating cropping and covering the higher-level, formerly cropped slopelands with alfalfa and other legumes and grasses (2,300 hectares in total). This would provide ground cover by perennial forage crops, to be used as livestock forage.

Inner Mongolia. "New thinking" is developing in Inner Mongolia with the realization that further deterioration of grassland will inevitably result in a progressive loss of livelihood in an 52 area where 15 percent of its GDP comes from livestock production, which employs 70 percent of its people. Eight percent of the project area, equal to 1,370 hectares, comprises natural grassland to be grazed under the project. This grassland is used by participating households currently grazing livestock under non-project conditions. Participating households will be engaged in mixed crop and livestock farming, utilizing their own crop byproducts as the main source of ruminant forage for cattle and sheep, and returning manure to their croplands. The management of this land would fall under the villages' grassland management plans to ensure sustainable use. With proper implementation, the use of these lancls would become a pilot demonstration of sustainable land use. In addition to the above, 2,000 hectares of deteriorated grassland would be rehabilitated and reserved for forage harvesting only, with an additional 2,200 hectares to be fenced off for controlled grazing.

The Inner Mongolian government represented by the GMS has given priority to the project areas in its own reseeding and other grassland rehabilitation programs currently underway. The village project implementation plans for the livestock sub-project would incorporate existing measures to protect agricultural lands. The project would finance beef cattle, sheep, goats, meat pigs, rabbits and poultry, construction of livestock and feed improvement facilities such as fences, silage pits, and animal sheds, necessary to support livestock production.

Funds from each investment category would be used to develop a basic livestock production system for farm and herder households who qualify for loans. Since most poor households have no livestock or have insufficient numbers, they will need access to project funds to purchase the necessary livestock and materials to construct livestock shelters and feed storage facilities. Construction of livestock or feed preparation facilities at farm households would focus on livestock shelters and feed treatment and storage facilities since all roughage used as livestock feed would be obtained as fodder or crop byproducts. Project beneficiaries must also have access to feed resources of sufficient quantity and quality to satisfy animal feed demands and the necessary husbandry inputs to produce a marketable livestock product.

Household livestock production would differ according to area and predominant agriculture production system. Farm household beneficiaries' cropping areas would focus on fattening beef cattle and raising small stock, while herder households in pasture areas would focus on the breeding and fattening of large stock.

Inner Mongolian herder households would develop integrated household pasture and fodder production systems called caokulun. A caokulun involves one hectare of land protected by a wall and tree-shrub shelterbelt in which irrigated fodder and grain crops are produced to benefit two households. Associated with the caokulun is an area of natural pasture assigned to each household, thatis, each household has a 30-year lease on the area on which household livestock are permitted to graze, and which will fall under the prevailing grassland management regulations to be monitored by the project, under the guidelines outlined below. The livestock feed produced- by the caokulun is used to supplement natural pasture forage obtained by grazing during the summer and fall and to provide a full feed ration to household livestock during the winter and spring. Rehabilitation of degraded natural pasture through seeding and grazing management using deferment and rest at critical plant physiological stages is a primary goal of the caokulun. Supplementing forage obtained by grazing during the grazing season and providing full feed during other seasons is designed to relieve grazing pressure at critical times to improve the ecological condition of natural pasture ecosystems.

Basic Implementation Principles. A set of basic principles and operating guidelines has been developed that will ensure successful implementation of the livestock sub-projects in each province. These guidelines are based on existing legaLland regulatory frameworks and contain the necessary conditionalities to ensure compliance. Implementation will take place in accordance 53

with the following five basic principles which can be used as the basis for livestock investment decision criteria: * Natural pasture land use will conform to existing grassland management and monitoring laws and regulations; * Ruminant livestock feeding will be based foremost on available crop byproducts and forages rather than on grazing natural pasture in order to minimize pressure on fragile grasslands; * Project participants who lease crop and/or grasslands will receive long-term (30+ years) land leases under which they will have the rights but also the responsibilities of land use; * Livestock densities will be based solely on realistically prepared and annually updated feed balances for green forage, crop byproducts, and feed grains; seasonal highs and lows in spring and fall; and types of livestock to be raised; and * The project village and its local government will be the basic unit to develop, implement and update grass and cropland management plans, and be responsible for their implementation in accordance with the above listed guidelines.

Monitoring and Enforcement of Agricultural Land Use. The five principles outlined above will especially concern the use of natural grasslands, and will be monitored and enforced by existing regulatory agencies, namely the project the County-level GMS and the local EPB. Village Project Implementation Groups (VPIGs) will bear primary responsibility for preparing and annually updating village area workplans. These will include grassland and livestock production monitoring indicators; updated feed balances; livestock inventory limits for the next implementation period in accordance with these balances; updated County livestock marketing reports; and other physical, financial and institutional performance indicators. County-level GMSs are committed to spot check overstocking and enforce the existing regulatory framework.

Provincial governments will ensure that GMSs receive adequate budgeting and resources to carry out their tasks. Bank supervision missions will review the workplans and implementation of the livestock sub-project and monitoring activities at 6-month intervals. Quality assurance spot checks will be carried out by technical staff from independent provincial-level institutions and by Regional PMO staff. Existing monitoring indicators from the Environmental Impact Assessment report pertaining to grassland and agricultural land use will also be employed. Guidelines for the preparation of feed balances and monitoring indicators for grassland management and livestock production are included in the PIP.

Household Livestock Production Models. The three provincial project areas each have a number of different livestock production models. Cattle, sheep, and feeder pigs will be included in all project areas. These form the "core" livestock in the project. Inner Mongolia will alsQ introduce meat chickens and rabbits, Gansu Province will introduce fish and soft-shelled turtle; and Qinghai Province will introduce cashmere goat and laying chickens as well as draft animals. Although the core livestock in the household models are similar, considerable difference is found between sub-project areas in the modeled livestock components. Inner Mongolia has variable models while Gansu has eleven basic variants and Qinghai has nine. Although all models rely primarily on feedstuffs obtained from sources other than natural pasture, models that have cattle or sheep/goats, especially the Inner Mongolia sub-project area, are programmed to-use limited natural pasture, attached to the caokulun, during the summer and fall seasons only. 54

Table 2-1: Household Livestock Production Schedule by Project Province

InnerMongolia 138,820 Pig,Sheep/Goat, a. IOS/2P a. 12 a. NPG/CC Cattle,Rabbit, b. 2C/IPs b. 13 b. NPG/CC Chickens c. IOG/IPs c. 11 c.NPG/CC d. 4S12P d. 6 d. NPG/CC e. 20S/2P e. 22 e. NPG/CC f. 2C/2P f. 12 f NPG/CC g.3P g.3 g.CC h. 2C h. 10 h. NPG/CC i. 2C/4S/2P i. 16 i: NPG/CC - j. 4S/IPs j. 7 j. NPG/CC k. 20S k. 20 k. NPG/CC GansLI 195,150 Pig,Sheep, Cattle, a. lC/2S/2P a. 9 j. 18197 a. CC/PF Fish,Turtle b. IC/2P b. 7 k. 24081 b. CC/PF c. 0.3C/0.5S/3.3P c. 7 1.31726 c. CC/PF d. 3S/2P d. 5 m. 21284 d. CCIPF e. 4S/2P e. 6 n. 17408 e. NPG/CC f IC/3S/2P f 9 o. 24831 f. CC/PF g. 2P g. 2 p. 33781 g. CC/PF h. IC/3S/2P h. 10 q. 22657 h. i. IC/2.5S/2P i. 10 r. 193985 NPG/CC/PF j. 1F s. 900 i. k. IT t. 930 INPG/CC/PF j. PF k. PF Qinghai 13,637 Pig, Sheep,Goat, a. lD/IlS/2P/20-f a. 15 j. 3089 a. CC/CFG Poultry,Draft b.lD/5G/I lS/2P/20H b. 17 k. 4324 b. CC/CFG Animals c. ID/IOG/6S/201i c. 18 1. 1730 c. CC/CFG d.lD/5G/2S/20Hi2P d. 10 m. 875 d. CC/CFG e. ID/8G/14S/20]i e. 25 n. 722 e. CC/CFG f. D/5G/21S/20R'2P f. 29 o. 656 f. CC/CFG g. ID/15G/9S/201. g. 25 p. 661 g. CC/CFG h. D/5G/27S/2P/20H h. 35 q. 663 h. CC/CFG i. ID/5G/12S/2P/20H i. 20 r. 854 i. CC/CFG 11 C = breeding cows and feeder cattle, S =breeding ewes and feeder lamnbs,G= goat (usually cashmere goat), P = feeder meat pigs or breeding sows, F = Fish, T = Turtle, R = Rabbit, D = Draft animal (Horse, Donkey, Yak, Mule),. H = chicken. 2/ Household livestock models in Inner Mongolia are variable according to village potential compared with livestock models in Gansu and Qinsghaisub- projects. 3/ CC = cut and carry, NPG = natural pasture grazing, CFG = crop field aftermath grazing, PF = prepared feed

Feed Balances. To ensure that the impact of incremental project-supported livestock production will not contribute to further degradation of natural pasture ecosystems, carefully designed feed balances will be employed to monitor feed availability and livestock needs. These have been designed to take into account livestock species, forage and grain balances, and seasonal (spring highs and fall lows) balances. Meeting the nutritional requirements of incremental livestock generated under the project solely by grazing during all seasons of the year would exert unacceptable grazing pressure with severe impacts on the ecological conditions of natural pasture lands. A considerable effort has been made to develop accurate feed balance estimates. These feed balances would ensure that feed requirements of incremental project livestock would be met without increasing grazing pressure. Annual feed requirements of livestock purchased by project beneficiaries in the three project provinces will be met from several different sources. These include: legume fodder crops (astragalus and alfalfa) developed on marginal cropland in Inner Mongolia and Gansu, fenced and irrigated fodder and grain crop areas developed by individual households in Inner Mongolia and Qinghai, natural pasture leased to individual households to 55

graze livestock during portions of the year in Inner Mongolia, crop byproducts such as grain straw, corn stover, and potato stems which will be ensilaged or treated to improve nutrient content in the three project provinces, and grazing of interstitial vegetation between crop fields in Gansu. Further details on feed balances and feed balance monitoring are located in the PIP.

Livestock Phasing. The annual increase of the number of livestock is indicated in Table 2-2. Converted into Sheep Units (SU), this increase provides a broad guideline with which to assess available forage and grain resources, as prepared by the provinces. As indicated above, livestock numbers will be determined on the basis of available feedstuffs and not vice versa.

Table 2-2: Incremental Livestock Numbers, by Project Year, Expressed in Actual and SU values

Inner Breeding Cow 6,950 12,373 11,817 6,918 3,295 1,647 43,000 215,000 Mongolia Breeding Sheep 34,369 62,443 63,414 53,143 24,142 12,077 249,588 249,588 Breeding Goat 5,554 10,068 6,988 7,702 2,688, 1,147 34,147 27,318 Breeding Sow 2,642 5,288 5,284 4,475 2,033 978 20,700 62,100 Feeder Pig 32,920 61,608 61,337 49,532 22,529 11,235 239,161 239,161 Total Livestock 82,435 151,780 148,840 121,770 54,687 27,084 586,596 Total SU 114,408 209,834 205,278 156,852 71,395 35,399 793,167 Gansu Breeding Cow 3,339 5,140 3,940 3,340 2,756 1,259 19,774 98,870 Feeder Cattle 158 211 211 211 158 107 1,056 4,224 Breeding Sheep 16,071 19,582 18,682 17,482 10,892 6,603 89,312 89,312 Breeding Sow 496 662 662 662 496 330 3,308 9,924 Feeder Pigs 42,125 56,071 49,831 46,127 34,161 18,048 246,363 246,363 Total Livestock 62,189 81,666 73,326 67,822 48,463 26,347 359,813 Total SU 77,011 104,183 91,043 83,139 60,953 32,364 448,693

Qinghai Draft - 212 1,799 3,222 4,357 2,565 12,155 48,620 Feeder Sheep - - 891 2,175 8,276 20,439 31,781 15,891 Cashmere Goa - 131 1,256 4,182 9,290 21,536 36,395 18,198 Feeder Pig - 213 632 1,796 5,935 8,997 17,573 17,573 Total Livestock - 556 4,578 11,375 27,858 53,537 97,904 Total SU - 1,127 8,902 17,863 32,146 40,245 100,281 Total Livestock 144,624 234,002 226,744 200,967 131,008 106,968 1,044,313 SU 191,419 315,144 305,223 257,853 164,494 108,007 1,342,141 1/1 he sheep unit (SU) represents a feed intake of 2 kg/day (730 kglyear) with I sheep =1 SU, I mature breeding Cow= S SU, I mature breeding goat = 0.8 SU, I feeder pig = 1 SU, I sow = 3 SU, and I draft animal = 4 SU. 2/ In Gansu Province, goats are treated as being equal to I SU. 3/ In Qinghai Province, feeder ewes and lambs will be purchased and fattened over a 150 day period and are equal to 0.5 SU.

Forestry Development

The main forestry development would be in the Qinghai Voluntary Settlement area. The Qinghai Forestry Bureau has considerable experience, and success, with forestry development, including sand dune stabilization in this high altitude, semi-arid region. A comprehensive program for the settlement areas includes wind shelterbelts on the perimeter and elsewhere in the project areas, sand stabilization, road and canal protection, and stream and gully protection. This system would provide comprehensive protection from the prevailing winds and over time, contribute significantly to stabilizing and increasing crop production. Fuel wood plantation would be an integral part of the forestry development.

In Qinghai, the effect of wind is a major constraint to crop production in the project area. It is also a major factor in soil erosion and desertification. Extensive experience in the Qaidamu basin and the project area has demonstrated the important contribution that forest protection can make to 56 both reducing wind damage and improving crop production. Under the project, a comprehensive afforestation program would be implemented. To the west of the main project area, a sand dune stabilization program would be implemented. The whole project development area would be completely surrounded by a forest protection belt. This would consist of four supporting belts. First, on the outside, a belt closed to all grazing ,md fuel wood collection in which suitable stabilizing grasses and shrubs would be planted; then a belt of low shrub bushes; third, three rows of higher shrubs; and fourth, three rows of trees (poplus) planted in relation to the prevailing wind (northwesterly). On the windward side it would be about 500 meters wide (4 belts). On the leeward side it would be about 350 meters. The poplar trees in the shelterbelt would be expected to reach a height of about 4 - 4.8 meters within 5 years. Species to be used in the bush belts include elaeagnus angustifolia; holoxylon sp; tamarix; and nitraria sibirica.

Within the irrigated crop production area, protective shelter belts would be planted along all canals and roads. The main canal and branch canals would be protected. The flood channels that pass through the area from south to north would be contained with tree belts. Cropland would be subdivided into rectangles each about 8- 15 hectares. Vertical intervals between the belts would be 400-600 meters and horizontal about 200-250 meters. The trees would comprise mainly poplar and willows.

An area would also be planted for fuel wood harvesting. This would include a range of species including poplar, tamarix, caragana, and haloxylom, among others. All of the trees and shrub bushes including those to be used for sand stabilization have been proven to be effective in the project area. Areas would be set aside for cash tree crop production using mainly wolfberry (lycium chinensis) and, to a lesser extent, seabuckthorn (hippophae rhamnides). The latter is particularly drought tolerant. Wolfberry is expected to yield the first fruits in about three years and reach full production in 6-7 years. Each household would be allocated a part of the tree crop and fuel wood plantations. Forestry development would include an extensive program of adaptive research and especially demonstrations of new planting techniques, new and improved species and irrigated forest management. The total area of afforestation would be about 2,862 hectares including farmland protection-1,204 hectares; sand-wind break-820 hectares, road protection-73 hectares, cash tree crops-400 hectares, and fuel wocd-346 hectares. In addition, it is proposed to establish a 20 hectare nursery. In the move-out area, 85 percent of the land vacated by the migrants would be converted into forests (30 percent); natural pasture (10 percent); closed hilly and mountainous areas (45 percent) for natural revegetation; and the balance (15 percent) would be terraced and assigned to poor households. The Water and Soil Conservancy Station in collaboration with the Agriculture, Forestry and Animal Husbandry Bureaus would prepare the plan and would be implemented by the Qinghai Water Resource Bureau.

The project would support the development of 1:5,538hectares of perennial crops in Gansu including apple and pear (3,148 hectares), prickly aLsh(7,862 hectares), walnut (2,289 hectares), Japanese larch (1,228 hectares), and mulberry, olive and tea (1,011 hectares). Technical packages- for each crop would include variety selection, pruning, fertilizer application rates, and pesticides. The packaging has been designed according to local environmental conditions and long terrn market needs and risks. The project would supervise the provision of certified quality seedlings, annual application of chemical fertilizers and pesticides until the first year of production.

Forestry investments for Inner Mongolia would take place on land either too steep or unsuited to cultivation. Most of this land is severely degraded and subject to severe wind erosion. It is also planned to continue the current policy of retiring unsuitable land from dry land cropping. These areas would either be sown to suitable pasture species (astragalus, caragana, elymus, etc.) or planted in "protective" forest or bush. The protective bush areas are able to provide some forage for livestock, mostly on a cut-and-carry system, or as controlled grazing depending on the locality and species. The most widely used species for this purpose are those of the caragana varieties. 57

Where suitable, steeper land would also be planted in "wild apricot", which can provide some economic return. A smaller area would be planted in larch trees for pole and timber purposes. The total area planted over six years would be about 15, 144 hectares comprised of 396 hectares of larch trees, 2,076 hectares of wild apricot, and 12,672 hectares of protective forest or bush.

Input Access and Supply

Agricultural Bank of China (ABC), acting as a financial agent for the Provincial Finance Bureaus (PFBs), would provide credit to farmers for purchase of fertilizers. Credit needs in subsequent years would be financed by ABC's own sources. The Project Monitoring Offices (PMOs) would organize the procurement and delivery of other inputs including improved seed, seedlings,.root stock, plastic film, pesticides, and other agro-chemicals. This system is already being successfully used in much of the project area. Farmers would thus be free to choose from whom they purchase fertilizers, whether they are state or non-state sources. Currently, in both Qinghai and Inner Mongolia, fertilizer and most farm inputs are supplied largely by state outlets. Providing farmers with cash rather than inputs directly is expected to accelerate the development of input supply services in both the state and the non-state sectors. Fertilizer, seed and other key inputs are readily available at the government owned fertilizer and seed supply companies, which have depots in every township. In Gansu, non-state shops also sell fertilizer and other inputs. Some act as agents for the State owned companies and sell on commission while others are traders.

Adaptive Research and Demonstrations

A program of adaptive research and demonstrations designed to develop improved and more economically and environmentally sustainable technologies and systems would be implemented. Lower cost and risk technologies and systems would be explored. While the detailed design and implementation would be done in conjunction with the respective research institutions, the programs would, to the extent possible, be conducted on farms under the supervision of County and township technicians. As appropriate, the implementation would be done by farmers under practical but controlled farm conditions. The research would be carried out within a range of agro-ecological zones and farming systems. To the extent possible, the applied research programs would be packaged within Farming System Research (FSR) into three main categories: crop production technologies and systems; animal production technologies and systems; and water utilization efficiency. These systems would integrate crop and livestock production technologies (including grasses and forage) and management on both rain-fed and irrigated farms. In all cases, the technology being tested would be compared to the benchmark system using standard technology. For each program, the purpose, the specific objective of the research, and the target area (agro-ecological zone and farming system) would be defined. Key technologies to be tested would include, but not be limited to: * Minimum tillage practices on dry land and the effects over time on soil water conservation; crop yields, soil organic matter, and most importantly, labor and costs. The aim would be to develop more economically sustainable cultivation systems which would improve both soil and water conservation and, over time, crop yields; * Trials and demonstrations designed to accelerate the search for new drought tolerant varieties of the main crops, which are more suited to the drier, drought prone soils and areas, have greater drought tolerance, lower water requirements, shorter growing seasons, and lower fertilizer needs; * Fertilizer needs, application rates, frequency of applications at various stages of growth, and most critically, monitoring the cash surplus per hectare from a range of applications, types of fertilizer and crop. The aim would be to rationalize fertilizer recommendations so as to improve the efficiency of utilization, reduce costs and reduce or eliminate the potential for soil and water pollution; 58

* Improved water saving and utilization techniques for both dry land and irrigated crop production systems including those for mulch crop production. The aim would be to test techniques for improving water utilization efficiency and thus the output per unit of water used. Trials would test newv or improved economical and practical technologies and systems for reducing seepage losses in water cisterns; portable spray irrigation techniques; improving water utilization efficiency for irrigated mulch crop farming; * Higher quality fruit trees, mulberry tree improvement and virus free potato production in Gansu; * In Gansu, applied research conducted in alfalfa and olive tree production, and processing sub-projects; and * Farm system research aimed at developing farming systems which would change the focus from one of subsistence with maximum grain production, to one which minimizes subsistence grain production and maximizes income from cash crops such as livestock and where applicable, vegetables, or some other higher value crop. The system would maximize the area of forage legume-astragalus sp., vicia sp., alfalfa (where applicable)-as an integral part of the crop rotation. Key performance indicators would include, among other things: farm income, labor requirements, soil fertility status including water holding capacity, crop yields, livestock production coefficients and where applicable, irrigation efficiency coefficients. The indicators would include technical, financial and risk analysis.

The FSR program would be conducted in two main ifarming systems in Inner Mongolia: one, the predominant crop production with some livestock system; and two, animal production areas where livestock grazing with limited cropping is the predominant system. An international study tour would be included in the FSR program.

Within Gansu, it is proposed that each Project County would establish two demonstration villages where improved technologies and practices for integrated crop and livestock production including economic terracing and irrigation techniques, wou]d be demonstrated. Most of the applied research work would be done within these demonstration villages. Villages would include a range of farming systems and main agro-ecological zones.

In Qinghai, the project would support applied research and demonstration on crop shelters, economic plants, sustainable tillage, soil nutrient monitoring, artificial grassland and water saving.

Animal Husbandry and Pasture Improvement Demonstration and Extension (D&E)

New techniques or products that increase production and/or reduce production risk will be readily accepted by poor households if the beneficial attributes can be demonstrated to apply to their production situation. For that reason, the adaptive research program outlined below focuses on activities that demonstrate and extend new productiiontechniques or agriculture products rather than engaging in a traditional research approach.

The proposed D&E program in Inner Mongolia includes four animal husbandry and pasture improvement proposals and several farm product and farming technique proposals, which include multiple components: * Pasture improvement techniques including introduction of new forage species and demonstrating techniques to rehabilitate degraded pasture; 59

* Animal management to reduce grazing impacts on natural pasture by selecting the appropriate time to begin grazing after growth of forage species is initiated; * Improved animal husbandry techniques including selection of feeds and feeding methods; and * Using improved animal and pasture management techniques to reduce grazing impacts including time and duration of grazing and deferred grazing systems to improve forage plant vigor.

Four counties/Banners have been selected for demonstrating animal husbandry, pasture, and farm improvement proposals. Selection was based on locations representative of the different conditions in the 15 project counties/Banners. The selected counties are Siziwang and Chayouhou Banners in the western portion of the sub-project area and Taibusi and Wengniute Banners in the eastern portion of the sub-project area. Activities related to demonstration and extension in all counties/Banners will be the responsibility of Inner Mongolia Agricultural University and the Demonstration and Extension Group of Wengniute Banner acting as a single group. Outside funds from the China Council for International Cooperation on Environment and Development (CCICED), if obtained, may be used to demonstrate resource management planning at the level of a village in Wengniute Banner. These funds would be linked to the WPRP D&E program but would also facilitate meeting CCICED objectives of establishing sustainable agriculture demonstration areas.

In Gansu, the D&E program is primarily oriented towards demonstration of farming systems. Included in this program are four animal husbandry related sub-projects. Livestock-related proposals include: - Demonstration of nutrient blocks to improve animal nutrition; and - Health programs for penned sheep. The demonstration of nutrient blocks to improve animal nutrition was selected as an appropriate D&E project. The demonstration of hormone injection to improve twinning in sheep was excluded from consideration as a D&E project because of its theoretical nature and unproven technology aspects.

In Qinghai, most components will need to be applied in the project as a matter of course since it involves oasis development. Three D&E projects suggested as being suitable in the Xiangride Balong development and settlement area were: * Demonstration of different types and sizes of feed storage facilities to improve winter and spring storage of animal feed prepared by the household; * Introduction of new and improved forage and fodder species that have high yields under conditions found in the project development area, especially high sodic soil conditions; and * Demonstration of the use of broad spectrum parasite control medicines such as Ivomectin to assist in meeting animal health requirements for mostly confined animals such as sheep and goats. Marketing studies would also be conducted to analyze likely supply and demand for the main agricultural commodities produced within the project areas. The aim would be to provide an analytical basis for assisting with formulating appropriate agricultural development policy and strategies. 60

Technical Services

Technical support services would include those provided from the County and Townships. In addition, depending on the size, each village would include one or more 'village technicians' who have received special training in one or more disciplines such as animal husbandry, crop production, water management, and farm machinery. These households would provide low level technical support and act as demonstration farmers. In some cases they would also act as agents for the township based fertilizer and seed companies. The County technical services would be provided from a series of 'Stations' according to discipline: grassland, crop production, crop protection, water conservancy (irrigation and soil and water conservation), animal husbandry and animal health, agricultural machinery and farm economy. Staff at the County level has usually completed a diploma and in some cases, are university graduates. Apart from planning and monitoring, the role of the County staff would be to provide technical support to the township based staff. They would also conduct technical training for technicians and assist with farmer training. The structure of County technical services is largely replicated at the township. Most township staff has a tertiary certificate or less. Agricultural graduates are rare. Depending on the season, township staff spends a high proportion of their time in the field visiting farm households and providing technical advice. During the winter imonths, some farmer training seminars are conducted.

In Qinghai, one comprehensive technical support station including all of the farm technical services would be established in each township, hea,ded by a Director. The long-term aim is for the technical services to be provided through a farmer owned company. A series of farmer owned and managed demonstration farms would be established in Qinghai in the early stages of development.

Technicians and Farmers Training

Training would be provided at three levels: senior technician, township technician, village technician, and farmer. In Inner Mongolia three senior technicians, one each from the three main disciplines-crop production, animal husbandry, and forest and grassland production-from each participating County/Banner would receive short term intensive refresher training (about 1 month) at the Inner Mongolia Agriculture and Animal Husbandry College. This would be followed by 2 weeks in project year 2 and 1 week in project year 3. At the follow-up training, each group (15) of technicians would present a brief summary of their experiences during the past year and discuss technical issues that have arisen during the year. They would also comment on the effectiveness and relevance of the training.

The training would focus on three main areas: one, an introduction to production economics and farm system analysis including risk analysis; two, crop, animal and forestry and pasture production and management as the case may be; and three, effective communication and training-. This training would be conducted in the first project year.

The three senior technicians (45 total) would lead the training of other senior County/Banner technicians and those at the township level. The iin-service technical training of other senior technicians and township technicians would be comlbined.The technical training seminars would be up to 5 days in length. Depending on the source of trainers and the available facilities, training would take place either at the League center, the County, or in the townships. Training at these levels in both provinces would be of a practical content and be relevant to the project investment program and the skill needs of the respective staff. All technical training including that for farmers would be conducted during the winter season. An introduction to elementary production economics and farm system analysis would be a special feature of the technical training program. The concept of the farm as a total integrated socio-economic system would be stressed. 61

A similar training program would be implemented in Gansu. Farmers would receive a minimum of two days technical training annually. One member of each household participating in the project would be required to attend training on the particular technology to be used, among other things. Township technicians would be trained for three days in each of the first four years, while County technicians would be trained for five days annually. Training would be conducted at training facilities at the province, County and township levels by staff and specialists from the Gansu Agricultural University and Provincial Bureaus, research specialists, and Prefecture and County technicians. Farmer training would be conducted in both the villages and the townships. In each village the project implementation group would ensure that at least one person from each household received technical training. Annual technical training would take place each winter or spring and be for 1-2 days in each case. In addition, selected farmers would receive additional training and act as village technicians. Thus each village would have at least one farmer specialized in one particular discipline. These village technicians would provide voluntary technical services to village households. Apart from formal training workshops, farmers and village technicians would receive continual on-the-job training from township technicians. During the cropping season, township technicians would spend a minimum of about four months continuously in the field.

In each County/Banner and township, 1 or 2 persons would be assigned to act as the training coordinators. They would receive some elementary training in effective communication. The training coordinators would be responsible for coordinating and organizing the training programs including preparing the training contents and ensuring the trainers, materials and equipment required are available. The project would also provide a limited package of visual aids (flip charts, overhead projectors, video screens, materials, etc.) to assist training implementation. In Gansu one visual aid package would be provided for each County PMO. The visual aids would be only those needed beyond what is already available. Training coordinators would be trained in the use of visual aids.

Study Tour

As an integral part of the Inner Mongolia Farming Systems Research program, about three senior technicians involved in the program would make an intemational study tour. They would study farming systems research and the technologies proposed for testing within that program. Minimum tillage systems, for example, would be one of those technologies. A more extensive study tour program is planned for Gansu. About 104 technicians, management staff and farmner representatives would complete domestic study tours within China. Some 19 senior technicians and management staff would make an international study tour. The purpose of the study tours and the results to be achieved would be carefully defined. Following completion of the study, the participants would complete a written report on their findings and recommendations, especially those considered relevant to the project target areas. They would also deliver a series of seminars- on their findings and recommendations. The project would finance all training costs including those for technicians and farmers (travel, subsistence, accommodation, study fees, etc.) and the cost of study tours both domestic and international.

C. Component 2 - Irrigation and Land Improvement (US$86.3 million)

Irrigation Development in Qinghai

Storage Development. The irrigation system developed under the project would comprise five irrigation sub-districts with an aggregate area of 19,200 hectares. Irrigation supplies would depend largely on diversion from the Xiangride River, with its upper catchment flow being 62 regulated by two natural lakes, of which Tuosuo Lake is the more significant. The project would finance the construction of the Keri Dam to provide an effective 23 million cubic meters of off- river storage for additional flow regulation and irrigation release at the beginning of the planting season in March. This would be a 40 meter high concrete-face sand-and-rock-filled dam with a crest length of 391 meters. There would be two saddle dams, also sand-and-rock-filled, the left with a height of 21 meters and a crest length of 135 meters and the right a height of 26 meters and a crest length of 480 meters. A 10 meter high masonry gravity dam with a 5 meter wide spillway would be provided for flood discharge. A low weir w ould be constructed on the Xiangride River to divert its flow into the Keri Dam via a 28 kilometer intake canal. A 59 kilometer long main canal emanating at the dam would bring irrigation supplies from the dam to Balong Irrigation Sub-district (14,700 hectares), and part of the Qingshuihe Irrigation Sub-district (1,660 hectares).

The project would also finance the reconstruction of ;heexisting control gate at Tuosuo Lake and the improvement of Hatu Dam. Hatu improvement would involve the raising of an existing small dam to 10 meters and seepage treatment of the reservoir floor to serve the Hatu Irrigation Sub- district (760 hectares). The other two gravity systems would serve 740 hectares in Yikegaoli Irrigation Sub-district and 1,3 10 hectares in Xiaoxiatan Irrigation Sub-district. The Tuosuo head works would replace the gate to improve operational efficiency.

Dam Safety Review. A panel of experts reviewed the safety aspects of the design of the Keri Dam, the existing Hatu Dam and the control gate of Tuosuo Lake in August, 1998. Based on the recommendations of the review panel, the design of the Keri right saddle dam has been revised and the second Panel made its second visit to Qinghai in December 1998, and has approved the design.

Irrigation and Drainage Facilities. The project would finance three separate conveyance systems comprising intake, main and branch canals, laterals, sub-laterals and farm ditches and associated canal structures for the five irrigation sub-districts. For water control and management at the farm level, water measuring structures (Parshall-Flumes) would be installed at lateral and sub-lateral levels. The conveyance system for the Qingshuihe Irrigation Sub-district would also include a diversion headworks. All canals, with the exception of farm ditches, would have concrete lining to prevent seepage. In the Balong Irrigation Sub-district, which comprises 76 percent of the irrigation area, typically a sub-lateral would serve an area of 15 hectares and a lateral would serve 120 hectares (8 sub-laterals). On average, a branch canal would feed 7-8 laterals. Farm ditches (compacted earth ditches) woul]dbe constructed within sub-lateral irrigation blocks, and each farm ditch would irrigate about 1.25 to 1.5 hectares. Irrigation would be by the border-strip method. In view of high intake rates of the soils in the project area, the length of border-strips would be relatively short, between 50 and 100 meters, and farm ditches would be set at these distances apart, perpendicular to the sublaterals. Border-strips would generally be 3 to 4.5 meters in width. Project financing would cover land leveling, both to remove the numerous sand mounds and to smooth and grade the border-strips to the required slopes.

A complementary drainage system consisting of main, branch, lateral, and sub-lateral drains would be provided to remove excess surface water. Except for the main drain, these drains would be constructed alongside the canals. The main drain running on the northern boundary of the Balong and Qingshuihe Irrigation Sub-districts would discharge into the wastelands and subsequently an inland lake to the northwest.

Water Table Control and Groundwater Development. Groundwater abounds at Balong and Qingshuihe Irrigation Sub-districts, with the underlying aquifers being recharged by several streams including the Hatu and Qingshuihe Rivers on the south and discharging into marshy land on the north. There will also be recharge from canal seepage and deep percolation of irrigation water. Drainage wells would be constructed to check progressive rise of water table and 63 secondary salinization resulting therefrom. Drainage discharge would be used conjunctively with surface water for irrigation. The wells, generally located along side branch canals would vary from 80 to 130 meters in depth and would be evenly distributed in the irrigation area. About 70 wells equipped with electrically operated pumps averaging 150 cubic meters/hour in discharge capacity would be installed. In addition to the drainage wells, sub-surface drainage would be installed in areas with shallow water table (less than 2.5 meters).

Water balance and reservoir operation studies incorporating groundwater model simulation showed that the construction of Keri Dam together with the installation of 67 tube wells is the optimum among several options considered including one with 287 wells but without the Keri Dam. The optimum scheme, besides being economical, has the merit of avoiding excessive draw down of the groundwater table and the consequential adverse impact on the adjacent grazing areas. Under this option, the annual evacuation of drainage water would be of the order of 23 million cubic meters. To ensure sustainable development, pumping would be carefully controlled to maintain the present groundwater table. A program of water level and water quality monitoring would be put in place from the start of project implementation.

Reclamation of Salinized Soils. Moderate salinization occurs in the Balong and Qingshuihe Irrigation Sub-districts. There are about 3,500 hectares of soils with salinity ranging from 1.5 percent to over 7 percent. About 90 percent of these are with salinity less than 5 percent. The soils would be improved and reclaimed through leaching in combination with agricultural measures. Experiments conducted elsewhere in the Qaidamu Basin showed that soils with similar salinity levels had been successfully reclaimed through applying leaching water at the rate of 7,500-9,000 cubic meters per hectare. Leaching is most effective when applied at high temperature during summer. This leaching rate would be adopted for the project area.

Water Supply, Demand and Quality. The project irrigation area would comprise 70 percent crop land, 15 percent forestry, and 15 percent pasture. Based on a weighted average net irrigation requirement of 5,468 cubic meters/hectare, and an overall canal system efficiency of 61 percent, the total gross irrigation demand would be 175 million cubic meters. Of this amount, about 79 percent would be met through the regulated flow of Xiangride River, 8 percent through the impoundment and diversion of Hatu and Qingshuihe River, and 13 percent from groundwater (drainage water). Apart from meeting the project irrigation requirements, the regulated flow of Xiangride River would also compensate the supply of 115 million cubic meters to existing irrigation areas downstream. Operational studies carried out using the continuous flow record of Xiangride River (1962-69), the driest period in 40 years- 1956-96, showed that the regulated flow of Xiangride River satisfied the demands in all years. Flows of Hatu and Qingshuihe River are also sufficient to meet demands in 3 out of 4 years. The overall supply situation is therefore satisfactory. Water samples tested showed that the quality of both surface and groundwater fully meets the national irrigation water quality standards in China.

Small Scale Irrigation Works in Inner Mongolia and Gansu. The proposed project would also finance the construction of small-scale irrigation works to serve 34,050 hectares of existing and new crop land in Inner Mongolia and Gansu. Irrigation works would include 6,460 tube wells and 5,060 shallow wells for 22,050 hectares in Inner Mongolia. In addition, a further 3,760 hectares served by existing groundwater schemes would be equipped with on farrn water distribution facilities comprising irrigation pipe lines and concrete lined canals to improve irrigation water use efficiency. In areas with a lack of groundwater resource, rain water would be harvested and 3,300 individually operated micro-irrigation facilities for 3,535 hectares would be constructed in Gansu and Inner Mongolia to tide over critical periods of water shortage and to lessen the impact of drought. Other small irrigation works would include canal lining, small pumping stations and moveable pumps and small flood protection schemes for 8,470 hectares. 64

The development of groundwater for irrigation in Inner Mongolia would be the single most effective strategy for increasing food production and improving farm income, while complementing ongoing soil and water conservation programs and mitigating pressure on overgrazed grassland in the project area. To ensure that there will be a sound technical basis for groundwater development in the project area, water balance and technical criteria have been carefully reviewed. The groundwater balance study lhas shown that approximately 25 percent of an exploitable volume of some 225 million cubic mnetersis presently abstracted in the project area. There is thus sufficient potential for the abstraction of an additional 83 million cubic meters per year by project wells. Groundwater salinity is less than 1000 milligrams/liter and suitable for irrigation purposes.

Site selection for groundwater development would be made using a uniform technical procedure and relevant hydrogeological, geological, soil and topographic maps and reports, which are available in all project County/Banner Water Conservancy Bureaus. The technical criteria for the design and construction of irrigation wells would conform to relevant technical standards. Irrigation wells would be of three types: tube wells JTW), shallow wells (SW), and shallow tube wells (STW). The technology for construction for all types is well adapted to the area. TW would be installed under a drilling contract while SW and STW would be constructed by the farmers. Wells would be equipped with centrifugal or submersible pumps, depending on depth to water level, driven by electric motors or, where electricity is not available, diesel engine. To guard against over-exploitation, a groundwater monitoring program would be put in place. Monitoring of water levels, abstraction and water quality would be carried out by County/Banner Water Conservation Bureaus (WCBs). Water levels and power consumption would be measured in all project wells at monthly intervals and water quality at annual intervals.

Land Improvement

Terracing. Aside from irrigation, terracing of sloping land would be a key strategy to bring about a substantial increase in the agricultural production and income of project households. New terraced land would allow project farmers to increase production of grain and cash crops and animal feed, thereby mitigating pressure on steeply sloping land. Terracing would also contribute to soil and water conservation in the degraded environment of the project area. The project would therefore finance about 18,000 hectares of terracing in Gansu and 8,500 hectares in Inner Mongolia. About 90 percent would be with earth riser and the remainder with rock riser. Terracing in Inner Mongolia would be mechanically constructed on new and cultivated sloping land below 15 degrees. In Gansu, terracing with both earth and rock riser would be constructed with manual labor on cultivated sloping land below 25 degrees. Benches of terraced land would vary from 8 to 15 meters in width depending on slope inclination, with wider benches being provided on flatter slopes. Terracing would take place on contiguous areas of 15 to 20 hectares, and civil works would include drainage and other soil and water conservation measures for environmental protection.

Soil and Water Conservation. About 5,900 hectares of trees and shrubs would be planted in Inner Mongolia to protect against wind erosion, to rehabilitate degraded land, and to improve the eco-environment for crop production and human habitation and for fuel wood. Planting would take place on ridges to protect farms and grassland against wind erosion; on waste land and degraded sloping land; in gullies; and alongside roads and waterways. Trees would be planted in belts with a density of 4,000 to 6,000 trees/hectare and spacing of 1 x 2 meters or 2 x 2 meters. In Gansu some 194 square kilometers of watershed subject to severe soil erosion would be rehabilitated through soil and water conservation measures comprising terracing, check dams, and tree shrub planting. About 250 hectares of gully land in Damao County in Inner Mongolia would be reclaimed for cultivation through construction of small check dams. 65

D. Component 3 - Rural Infrastructure (US$28.7 million)

The project would support labor-intensive construction of rural roads, water supply systems, and rural electrification works.

Rural Roads

In Gansu and Inner Mongolia, the project would construct 216 kilometers of all-weather gravel roads and four bridges to link 289 administrative villages to the existing network of County and township roads. Roads would be included for implementation using similar screening..and selection procedures and criteria adopted for the Bank-assisted Second Henan Provincial Highway Project and Tri-Provincial Highway Project assisted by the Bank. Based on the anticipated daily traffic volume, 171 kilometers of roads would be constructed to Class IV road standards in Gansu and 45 kilometers would be constructed to Class III road standards in Inner Mongolia.

The project would finance 116 kilometers of all-weather Class IV gravel roads in 4-kilometer x 4- kilometer grids to serve 21 new administrative villages and newly developed irrigation area in Qinghai. These roads would be constructed alongside branch canals. Financing would also be provided for 426 kilometers of farm roads, constructed alongside irrigation laterals and linked to the gravel roads. Farm roads would have a formation of 4.5 meters with a 3 meter carriage way designed for the same vehicular loading as Class IV roads.

Rural Water Supply

Low-cost water supply facilities would be provided to about 308,200 people in 760 administrative villages in Gansu and Inner Mongolia. The facilities would tap from springs, utilize groundwater, and divert or pump from streams, and would include piped systems both with and without household connections. In localities where neither surface nor groundwater is available, rain collectors serving individual households would be provided. Facilities would be designed to supply 15 to 35 liters of water per capita per day, and would meet current water quality standards for rural water supply in China. Drinking water for animals would also be provided in areas where water source is sufficient and marginal cost of construction is reasonable. Daily consumption of 5 liters per sheep and 30 liters per head of cattle would be allowed. For piped systems, a design period of 15 years would be adopted and 1.2 percent taken as the annual population growth. A factor of 1.5 would be used for maximum day demand and the peak hour demand factor applied would vary from 2.5 to 4.

Rural water supply facilities would also be provided to 61,800 migrants in the voluntary settlement area in Qinghai. The facilities would utilize the good quality groundwater that occurs- in abundance in the area and would include public hydrants for centralized supply to villagers.

Rural Electrification

The project would enable about 410 administrative villages in Gansu and Inner Mongolia to gain access to electricity. To satisfy power requirements for households, agro-processing, other light industries and mining, 1,832 kilometers of 10 kV and 73 kilometers of 0.4 kV distribution lines would be installed. To meet similar requirements in the voluntary settlement areas in Qinghai, power would be drawn from the proposed 110 kV power line at Xiangride Town. Two transformer stations, one each at the two new project Townships of Dongba and Tawen, 126 kilometers of 35 kV and 211 kilometers of 10 kV power transmission lines, 1,238 kilometers of 220 V and 380 V distribution lines would be installed. 66

E. Component 4 - Rural Enterprises (US$13.6 million)

The rural enterprise component comprises three investment packages covering the three PPs. The objectives, types of sub-projects, lending terms, and implementation arrangements are described separately for each PP.

Gansu

Objective and Proposed Sub-projects. The objective of the rural enterprise component in Gansu is to promote commercially viable investments with a view to opening markets for raw materials and generating employment opportunities for project area households. Two categories of investments have been identified: * Investments in large scale (Yuan 250,000 upwards) rural enterprises providing benefits to the poor project households either by giving farmers access to markets for their produce or by providing employment opportunities at a relatively low cost per job created; and * Loans to individual and small groups of households to set up small-scale village enterprises.

(i) Large Scale Enterprises. Loans for large-scale rural enterprises will contain terms relating to the poverty alleviation objectives of the project. These include: i) priority to project households for employment and selling their farm products to the company; and ii) provision of technical assistance to project farmers supplying raw materials, to upgrade their farming practices. Both these requirements must be met for a sub-project to be included under the project. For example the apple storage sub-projects will benefit poor apple farmers by providing them with access to storage. The carpet factories provide local employment for unskilled labor who will be trained on the job. Basket weaving, and collection of wild vegetables and medicinal herbs provide both employment opportunities and additional rural income opportunities. These sub-projects would be operated as share holding companies with a minimum of 30 percent of total investment (equity) being provided in the form of cash or assets by investors.

(ii) Small Household Enterprises. These enterprises would operate at the village level processing animal feed, oil seeds etc. for village households in return for a fee. The enterprises would be owned and operated by individual or small groups of project households to service other village households, both project and non-project. Cave storage loans for apple farmers will also be provided under this category, as the facilities would provide storage for the farmer's own fruit and also for fruit from other project farmers in the immediate neighborhood. While only a limited number of enterprises have been developed at this stage it is envisaged that the scope of- loans under this category of enterprises will be widened in later years to cover similar household enterprise activities with the same order of investment cost per family. Eligible individual households will be approved by the County PMOs. ABC, using their current criteria for household loans, would appraise and approve loans on terms for similar loans.

(iii) Appraisal Criteria, Lending Terms. Appraisal criteria differ for the two categories of enterprises. Loans will be based on standard ABC terms with interest rates equivalent to current ABC rates for similar investments. The current rates range from 9 percent for up to three year loans, to 9.72 percent for up to five years, and 10.35 percent for more than five years. A minimum investor equity of 30 percent of the investment would be required with the balance provided as project loans. The PMO would furnish all loan applications to ABC for review and 67

appraisal and the PMO would make loans to eligible investors on current ABC terms for similar investments. Rural enterprise sub-projects in Gansu would be subject to prior Bank approval.

(iv) Large Scale Processing Investments. Gansu has identified a number of processing investments to be appraised and approved by ABC. The appraisal will be conducted in two stages. An initial appraisal would be based on the project documents as prepared by the County Project Offices, Design Institutes etc. Once a sub-project has passed this appraisal by all parties it indicates that the project is considered commercially sound and in line with project objectives. When an investor is identified, the feasibility study of the sub-project would be revised, generally by the preparation of an addendum adding details of the company structure, names of shareholders and their equity contribution, updated market information and prices, any changes in details of the project proposed by the investor, updated equipment lists and other additional details as required by the appraising bodies during the first stage appraisal. On the basis of this, a second appraisal should be conducted to review the commercial viability and development (poverty reduction) impact.

(v) Commercial Study. The commercial feasibility study should look purely at the commercial aspects of the enterprise. Market and pricing sections should be developed to include full details of current market conditions with past time series if available. Equipment lists and cost estimates should be supported by sketch drawings, and calculations showing building areas required, copies of catalogues and quotations for major equipment items. At least 65 percent of items by value should be supported by written quotation. Management and organization structure descriptions need not be developed in detail since these would change once an investor is identified. Loan terms and conditions proposed are to be described and included in the cost tables based on a minimum 30 percent equity from the investors. A brief description of environmental issues should be included highlighting possible environmental impacts. All projects require Environmental Bureau approval as a condition of loan approval.

Where the sub-project involves remodeling or expansion based on an existing enterprise the full results of the current operation must be included as an annex to the study including operational and financial performance. Generally the new company should incorporate the existing operation, although sometimes this may not be possible due to different ownership systems, and it should be included in the new enterprise model if it is to continue in operation. The existing assets if they continue to be used can form part of the equity.

(vi) Poverty Impact Study. This would assess the poverty alleviation impact of the sub-project against the criteria outlined above. The impact study would highlight the direct and indirect impact of the project including: i) employment of project households; ii) expected sales of raw material produced by project households; and iii) summary of technical assistance to be provided to project farmers.

Implementation Arrangements. Rural enterprises would be independent legal entities. Potential promoters .of enterprises would be invited to participate through an open selection procedure. The selection committee would comprise the heads of the County Poverty Alleviation Office, County PMO, ABC, and Township Enterprise Administration Bureau. A vice magistrate of the County would chair the selection committee. The provincial PMO would review and approve the selection. Participants would: i) provide required amount of equity; ii) provide construction designs and procurement plans; iii) set target for poverty reduction and develop and implement technical standards for construction, operation and environmental protection; and iv) prepare and furnish progress report to County PMO. Criteria for selecting, evaluating and approving sub- projects under this component are given in Annex 3 of the PIP. PMOs would actively promote the proposed sub-projects and identify prospective investors. They would also familiarize investors with the completed feasibility studies that have been done and assist with updating these studies. 68

Once loans are approved, primary responsibility for implementation will rest with the investor rather than the project office and County bureau. PMOs, in collaboration with ABC, would be responsible for monitoring implementation, ensuring that procurement guidelines are followed, the project follows the agreed plans and that disbursement of funds matches physical progress. A manual covering these responsibilities should be prepared by the PMO. Implementation arrangements for small household loans will follow similar procedure as above.

Inner Mongolia

Objectives and Proposed Sub-projects. This component is intended to make available to project farmers a range of small machinery to increase farm labor productivity and provide transport for goods and people. The proposed investments would include: i) stationary machines for feed chopping, milling and threshing; and ii) small tractors for transport, general farm work and to power farm machinery. It is intended that a range of equipment will be financed and that farmers can tailor a package to suit their own particular needs, subject to the approval of the Provincial Agriculture and Livestock Engineering Design Institute on technical issues and the Project office in terms of financial viability. Farm machinery loans under this sub-project will be made available to the whole project area subject to meeting the defined criteria. Specifications for machinery will be prepared by the Design Institute and Project Office for competitive bidding.

These small chopping, milling, and threshing machines are used to either reduce the labor requirements at peak seasons or to carry out tasks that are difficult or impractical to perform by hand. The choppers and huggers will be used to process straw and crop residues for silage and micro-ensilage in order to upgrade these materials for animal feed. While the financial models have been prepared for fixed electric powered units, portable tractor powered units will also be available where these are better suited to local requirements.

Small tractors have also found widespread acceptance in rural China for transport (people and goods) and for agriculture where they are used for both farming and as a source of power for stationary farm machinery and pumps. These robust, versatile machines equipped with trailers provide the primary transport for both farm inputs and outputs as well as people. With the significant increase in both farn inputs and outputs as a result of the project, demand for transport will increase accordingly. Without adequate transport many of the project advantages are in danger of being lost. By localizing transport in the villages (Gacha) rather than townships more of the benefits arising from increased productivity will be captured at the village level.

A typical package comprising a tractor, trailer, plough, harvester and seed sower has been modeled by the Provincial Agriculture and Livestock Engineering Design Institute. It is not intended, however, that farmers be restricted to this one machinery package. Implements suitable for more specialized tasks can replace the standard ones and there is no obligation on the farmer to take the whole package. The returns to the fanner for this package demonstrate that the proposed loan repayment schedule can be met particularly when returns to labor are included.

Lending Criteria and Terms. Application for such loans would be subject to: i) current availability of machines at the Gacha level (either under the project or privately); ii) the proposed owners being able to bring the necessary entrepreneurial skills to the project; iii) guarantee that priority service will be available to the project Barners; and iv) tractor drivers to come from within the owners' families rather than being hired labor. Lending terms for tractors and machinery will be offered by ABC for similar investnents; that is currently 3-5 year loans with a maximum of one year grace period when interest only will be payable. ABC's interest rate reviewed annually, is currently 7-8 percent. Repayments will be made six monthly at dates set out in the sub-loan agreement. Sub-loan agreements will include rights to repossess and resell 69 machines in the event of loan conditions not being met. Where ownership is to be by a group of families the group families will be jointly and severally liable for the debt.

Implementation Arrangements. The PMO would actively market the proposed sub-projects and identify prospective investors. It would also familiarize investors with the completed feasibility studies that have been done and assist with updating studies. Once loans are approved primary responsibility for implementation will rest with the investor rather than the project office and County bureau. The project office will be responsible to monitor implementation ensuring that procurement guidelines are followed, the project follows the agreed plans and that disbursement of funds matches physical progress. A manual covering these responsibilities has been prepared by the PMO. Implementation arrangements for township markets and small household loans will follow procedures similar to the above.

Qinghai

Objectives and Proposed Sub-projects. This component would finance a range of village scale processing machinery and company owned feed concentrate mills as identified by the Provincial Rural Enterprise Bureau (PREB). Four specific small enterprises have been modeled-flour milling, fodder chopping, oil extraction and noodle making. It is intended that there will be some flexibility in the makeup of the package and households will be able to tailor a package to suit their own particular needs, subject to the approval of the PMO. The concentrate feed mill is intended to have a conventional commercial structure and will require equity investors and management identified prior to approval.

In addition to the enterprises described above, sub-projects also include small loans, mainly targeted to women, of less than Yuan 1000 to households under a standard ABC lending program. This program allows for a variety of small investments in farming enterprises, which are common in many farming villages in Qinghai. Villagers take their wheat, oil seeds, or fodder to be processed for a fee. The models have been prepared by the PREB. The operations have been scaled to suit a single household or two households combining to run the enterprise. Capital costs have been kept to a minimum to allow entrepreneurs with only limited resources entry into a business.

Lending Terms. ABC would lend to household enterprises as poverty alleviation loans from ABC funds set aside for this purpose: * Loans of up to Yuan 10,000 per household would be provided for setting up approved investments under the project. More than one household can take out a loan to jointly invest in an enterprise subject to the equity rules for the total investment being met; * 20 percent equity will be required and 10 percent of the loan must remain in the- individuals' account for the first year (making effective equity requirement 30 percent) but no other guarantees or collateral will be required; and * Loan term will normally be 3-7 years. Interest only will be paid during the term of the loan with the loan being repaid in full at the end of the loan period. Interest rates will be as set by the State for poverty alleviation loans. (Currently these range from 2.88 percent to 4.5 percent depending on term).

Implementation Arrangements. The entire rural enterprise component would be financed by ABC using its own source of funds. ABC would manage loan disbursement and would be responsible for selecting, approving, implementing, monitoring and evaluation. PMOs will provide information materials prepared by the Rural Enterprise Bureau (REB) and mobilize the households to participate. Households wishing to take part will be directed to the ABC for loan 70 processing. ABC has signed an agreement with PM\O to provide funds as required for the household enterprise program in a timely manner subject to the loan conditions being met. They will also monitor and evaluate implementation progress. REB will provide technical assistance and deliver training. In the case of the concentrate feed mill, the loan from ABC will be on a normal commercial basis at rates current at the time of borrowing and will be negotiated directly between the ABC and the investors) and cleared by the Provincial Environmental Protection Bureau.

F. Component 5 - Labor Mobillity(US$15.7 million)

Objective

The labor mobility component is aimed at providing employment opportunities to 65,700 rural workers in Gansu with a view to supplementing their farm incomes. In the project area (Gansu) about 265,000 rural laborers, or 33 percent of the total rural population, are under-employed. Discounting for age and potential project-generated employment opportunities, an estimated 122,000 surplus workers would be available for export to the outside labor market. While Gansu has a long tradition of exporting rural workers to other parts of China, a special labor mobility program would be implemented under the proposed project. The program would be designed to target poor households with surplus labor-households in remote areas that often have no access to labor market information and employment opportunities. In addition, the project would strengthen the labor market information system of Gansu's Provincial Labor Bureau (PLB) to provide timely information to job seekers and employers.

Similar programs have been supported by the Bank in Southwest Poverty Reduction Project (SWPRP) and by the government in Gansu. Lessons show that such programs may fail if labor market information is poorly provided, initial support to laborers is insufficient, monitoring of laborer conditions does not function, and institutional coordination is weak.

Component Activities

The component activities would include the identification and selection of surplus workers; provision of basic orientation on the urban environment, labor law and relevant worker's safety measures; on-the-job training and initial settlernent expenses; and establishment of a computerized labor market information system linkcedto 13 labor stations in China and the Ministry of Labor and Social Security employment information data base. The system would register job openings and profiles of job seekers, contract with employers to place selected workers in enterprises, monitor the safety and well being of workers, and recovery loans.

Implementation Arrangements

Placement of Workers. 65,700 persons would be placed in off-farm employment over a period of six years. Of these 19,600 persons (30 percent) would be for intra-provincial placement and 46,100 persons (70 percent) would be for inter-provincial placement. On average, intra provincial and inter-provincial job-seekers would receive loans of Yuan 1,240 and Yuan 1,740 respectively to cover pre-departure, training, transport and initial settlement expenses. These loans would be repaid within two years at 8 percent annual interest. PMO would make arrangements with the employers for monthly salary deductions. The annual phasing of labor mobility would be 12 percent, 11 percent, 16 percent, 21 percent, 26 percent and 23 percent. 71

Applicants for the program should meet the following criteria: * Surplus household labor as certified by the VPIG using the standard norms of farm holding and available family labor (1 labor for 0.5 hectares); * Per capita income and food grain production below the poverty level; * Basic literacy, numeracy and any technical skills; * Good health as certified in recent medical exam; and * Willingness to work in an urban environment.

Pre-departure basic orientation would be arranged by the PMO in collaboration with the Provincial Labor Bureau (PLB). This training would cover introduction to labor law with emphasis on labor rights and obligations and safety standards; living in urban settings and the contractual obligations entered by the workers to repay loans and social obligations to remit incomes to their households. To enhance skill training for the migrant workers, employing enterprises would provide on the job training at cost.

Labor Market Information System. The project would develop a labor market information system (LMIS) for registration ofjob seekers and job openings. The LMIS would be set up in the PLB with links to the Project Management Office and 13 labor stations in major cities in China. The system would be operated by the PLB and would provide information on job openings to Project County labor bureaus for dissemination to project villages.

Monitoring and Evaluation. To ensure the effectiveness of the program and the well-being and safety of the migrant workers, the project would set-up a monitoring and evaluation system. The system for Monitoring and Evaluation would be designed by a qualified institute (the Population Research Institute of Northwest China, Lanzhou University) under the overall guidance of Gansu PMO. Surveys of representative sample workers and villages of workers' origin would also be conducted to assess the amount and use of remittances and the survey results would be reported annually to Gansu PMO. The All China Women's Federation would also cany out annual reviews to ensure that female migrant workers are safe and their contracts are strictly enforced.

Institutional Arrangements. Under the coordination of the PMO's Labor Mobility Coordination Section, the Labor Bureaus at provincial and County levels would take the responsibility in the implementation of the labor mobility component. The Labor Bureaus experience in managing labor export programs but lack the resources for improving their services. The PMO has signed an MOU (Service Agreement) with the PLB covering the annual targets, selection, pre-departure training, and placement procedure. The PLB would utilize the services of Labor Stations located in 13 major cities in China to provide up to date information on employment vacancies.

Retroactive Financing. Selected activities would be suitable for retroactive financing, such as setting up the basic training sites; finalizing the operational manual for VPIGs, County PMOs, County Labor Mobility Offices, and provincial PMO and PLB to guide the selection of laborers, training, transport, placement, and monitoring; and preparing the Monitoring and Evaluation System based on Terms of Reference (TOR) to be agreed with the Bank.

G. Component 6 - Voluntary Settlement (US$6.5 million)

There are 0.7 million poor in Qinghai living in 14 national poverty counties including 6 project counties in the resource-deficient eastern arid region. In many project townships the inhabitants have to regularly depend on government relief. Since 1986, 6,150 households with 29,000 people have voluntarily moved out and settled in Qaidamu basin including Xiangride area in Dulan 72

County, the selected 'move-in' area for the project voluntary settlement. Experiences from other Bank-supported voluntary settlement programs in China have been incorporated in the project design. Rehabilitation of 85 percent and terracing of 15 percent of the land to be vacated would be financed by the project counties. The component was prepared with specific reference to the provisions of World Bank OD 4.20, Indigenous People and OD4.30, Involuntary Settlement.

Selection Criteria for 'Move-Out' Counties and Households

Criteria for Selecting 'Move-Out' Counties. Poverty incidence and degraded physical environment are the two criteria for the selection of the emigration counties and townships. The selected six emigration counties with 48 townships are all with high incidence of poverty, severe soil and water erosion, extremely deficient agricultural land, lack of drinking water, and a rapidly deteriorating ecological environment. In such villages, the prospects of development, without further destruction of the fragile natural resources, is non-existent. The selected counties include some villages covered by the IFAD assisted Agricultural Development Project. These villages would not be eligible for participation in the proposed project.

Criteria for Selecting Households. The proposed settlement program is entirely on a voluntary basis without any form of coercion or discrimination of gender, ethnic origin or religious belief, except for the loss of grazing land currently leased by 63 households in the 'move-in' area (to be compensated fully by irrigated pasture). The selection of households would be conducted in a fair, transparent manner. Initially, the selection criteria include: i) per capita income of less than Yuan 580; ii) per capita grain production of less than 210 kg; iii) households with at least two laborers and the head of the household between 18 and 50 years in age; iv) experience in agriculture; and v) good health and able bodied. Since the number of prospective migrants based on these criteria is higher than the estimated carrying capacity of the 'move-in' area, a second stage selection criteria has been introduced including: (a) per capita income less than Yuan 300; (b) per capita grain production less than 140 kg; and (c) middle school education or basic technical skills (for at least 10 percent of the prospective migrants). To accord preference to minority nationalities, the educational and skill attainment standards would be relaxed. Based on the carrying capacity of the land area and initial expression of interest by prospective migrants, the project would assist the voluntary settlement of 57,775 people, including 33,318 minority migrants (58 percent). In addition, 4,000 herders and farmers, currently either grazing or living in the project area would be eligible to lands under the project according to the provisions of the Bank's OD 4.30, Involuntary Resettlement.

Public Consultation. County PMOs would publicize the project through mass media-radio broadcast and television, mass meeting, leaflets, and public gatherings. Applicants would receive a copy of Migration Notice describing the migration and settlement plan, expected benefits, risks, project support for the migrants, land allocation and lease, basic facilities and service facilities provided in the settlement regions, preferential policies for the migrants, and estimated repayment- obligation. The Migration Notice would also provide detailed explanation on the submission of applications and the formalities of the selection procedure.

Voluntary Settlement Implementation Plan. The voluntary settlement would be phased over a six-year period. Prospective migrants would settle in two existing townships that would be expanded and two new townships. There would be 21 new administrative villages with 126 natural villages in these townships. Five new irrigation sub-districts would be established. Each migrant would be allotted 0.22 hectares of irrigated land for crop and fodder production with a lease of 30 years. Migrants would be allowed to retain their land in the 'move out' area for two years in the event that they may wish to return to their home villages though such cases are rarely expected based on the experience of voluntary settlement programs in Qinghai in the past decade. In such event, the migrants will not be requested to repay the capital investment from which they 73 would benefit directly or indirectly. During the first year, and prior to the planting of the first food crop, migrants would be able to earn cash income by participating in the construction of the project irrigation and rural infrastructure. Under retroactive financing, a pilot program complete with irrigation, rural infrastructure, and health and education facilities would be established in order to gain early experience in project implementation.

(i) Settlement Subsidies. Migrants would be exempt from land rent for the first three years. Each migrant would receive the following grants: Relocation fee Yuan 240 House construction Yuan 500 Fuel Yuan 110 Total Yuan 850

(ii) Training the New Migrants. Coming from the background of mountain agriculture, the new migrants lack the know-how of irrigated agriculture. They would receive training in irrigation, irrigated crop management, crop, and animal husbandry and production.

(iii) Minority Migrants. The predominant minority groups in the 'move-out' project counties are Hui, Sala, Tu, and Tibetan nationalities. To preserve their cultural and religious tradition, minority groups would live in a community of their choice in accordance with their original residential population composition. The project would establish a minority nationality township and several minority villages. Designated land will be allocated for religious use. Their community leaders and religious elders would participate in local township work stations and the VPIGs to assist in resolving minority nationality issues. The minority migrants would enjoy preferential policies in employment, child education, and tuition. More information on basic education services is provided in Section H of this Annex. Employment in Project supported construction will give first preference to minorities when hiring, micro-credits will target minorities and schools will charge reduced tuition or none at all for minority children. Bilingual education and community activities making use of minority languages will both be provided as appropriate.

It is estimated that about 3,500 Tibetans will participate in the voluntary settlement program. Their views and interest have been recognized by the project authorities and reflected in project design and preparation. To support this process interviews were made of a 5 percent sample population of herders which included 10 percent Tibetans.

(iv) Women Migrants. Women's participation would be encouraged in income generating activities and training. The project would fund small household enterprises such as handicrafts and other activities. Moreover, individual women migrants would be eligible for Yuan 1,000 credit and women can form groups in a village for a group loan of up to Yuan 30,000 to develop a handicraft and processing industry. An estimated 17,200 female migrants would receive training. -

(v) Pilot Program. Prior to the launch of full scale settlement, a pilot program on 300 ha for 200 households would be undertaken to test the proposed interventions at the field level with the participation of project beneficiaries. This would create a demonstration area which could be later used as a training site for project participants and provide an opportunity to show typical types of irrigation layout, village infrastructure, schools, clinics, agricultural (planting of green manure) and land management practices which would be promoted by the project. The pilot program would also include participatory evaluation techniques to provide input from a representative range of parties in the project area. The pilot would also ensure that the minority needs are fully met. The pilot program would provide an opportunity to develop "lessons learned" which could be applied to the design and implementation of the full-scale phase of activities which would start after full evaluation of the pilot program by independent reviewers. It would be carried out in 74 such a way that it does not interfere with current land use patterns of the herders or farmers living in the area, nor will it result in loss of their shelter, productive assets, or access to pastures. Irrigation will be provided through use of ground water and diesel engine driven pumps, so as to avoid the necessity of developing major irrigation works during the pilot phase. Expenditures for this activity have been approved for retroactive financing under the project. The PMO would be responsible for the implementation and supervision o:fthe pilot program. The pilot settlement will be closely monitored to measure the efficiency of the various program components and also, to identify any emerging trends in the relationship between migrants and receiving area population. Full scale settlement would only begin after the pilot program is completed and evaluated.

(vi) Management for Migrant Settlement. The amnual plan of the migration approved by the provincial leading group will be managed by the 'move-out' and 'move-in' County-leading groups. The leading groups in the six 'move-out' counties would arrange with the provincial line agencies to implement and supervise the implementation of the VSIP. The PMO would transfer project funds to the County PMOs to implement the voluntary settlement.

(vii) Settlement Monitoring. The Qinghai Plateau CGeographyResearch Institute would set up a voluntary settlement monitoring system supervised by independent experts. The system will track five percent of the households from the time of their selection and to the time when they are settled. Annual reports would be prepared to facilitate the PMO and Provincial Leading Group in taking corrective measures for problems and issues identified.

H. Component 6 - Social Sector Services (US$8.7 million)

Basic Health Services

Objectives. The objective of the health investment is to improve health service in the project area to meet the local inhabitants' basic health needs. It will be achieved through improving the quality of and access to basic health service delivery at the primary level, especially in the villages.

Component Activities. The project would finance the establishment of new health services in the voluntary settlement area in Qinghai and in the 'move-out' areas with four components: i) development of new health facilities; ii) training of health workers; iii) health education and health service management; and iv) cooperative medical financing. Financing would include civil works for township health centers and village health clinics, medical equipment and medicine supply, training of health workers, health education and management training.

In Gansu and Inner Mongolia, the health network would be improved through i) better basic health service delivery; ii) interventions for priority diseases; iii) establishment of a rural Cooperative Medical System; and iv) strengthening of health service management.

Selection of Target Villages. In the Inner Mongolia Autonomous Region, the health investment would be mostly focused on the 917 project villages. Some of these villages already have ongoing program for control of tuberculosis due to its high public health significance. The tuberculosis control program would be expanded to cover all villages in the 15 project counties. In 6 counties where the IDA-assisted Health VI is ongoing, investment would be scaled down to avoid duplication. All counties in Gansu are covered by Health V (Tuberculosis Control) and Health VII (Disease Prevention); five of the 12 project counties have also been covered by Health VIII, and three of a total of 12 project counties have been covered by the Health VI Project. To avoid overlapping coverage, the project investment will be allocated in the 1,002 project villages not covered by Health VI and VIII. 75

Basic Education Services (Qinghai)

Objectives. The project would address the education needs in and around the voluntary resettlement area and the 'move-out' counties in Qinghai and would, over a period of six years, i) provide basic education access to the school-age children of the migrants consistent with the national target of establishing universal nine-year primary education by the year 2000; ii) ensure that overall education development in the voluntary settlement area would be at the same level as that of the 'move out' area; and iii) provide vocational and adult education to satisfy all other education needs of the migrant population.

School Enrollment and Teaching Staff Requirements. Based on the targeted migrant population (61,800) and the provincial education statistical data, children falling into the primary school age group and junior secondary school age group would number 8,000 and 3,700 respectively. It is also estimated that 50 percent of the secondary school students would continue their education in senior high schools.

There will be 21 newly established administrative villages in the project area. To meet the education requirements, the component would include: i) 6 complete primary schools with 6 classes for 4-6 natural villages (half of the schools will be minority schools); ii) 5 complete primary schools (of which 3 would be for minority nationalities) with 12 classes; iii) 10 joint primary and junior secondary schools with capacity for 4,050 primary and junior secondary school students respectively (5 of them will be minority schools), and provision of one year vocational education for the graduates who will not continue their education in high school; and iv) expansion of the existing No. I Secondary School in Dulan County to admit the high schools students from the voluntary settlement area. Based on the norms of the Ministry of Education and the actual situation in Qinghai Province, the number of teaching staff required has been estimated according to the ratios of class to staff of: 1:2.2 for primary schools (1:1.7 for full time teachers and 1:0.5 for supporting staff); 1:3.7 for junior secondary schools (1:2.5 for teachers and 1:1.2 for supporting staff) and 1:4 for high school (1:2.8 for teachers and 1:1.2 for supporting staff). In all, there would be, respectively, 409, 333 and 84 teachers to be hired for primary, junior secondary and senior high schools in the project area.

Minority Schools. Minority schools would be established in the 'move-in' and 'move-out' areas to provide basic education to minority nationalities. The schools would be equipped with minority language text books and other teaching materials . Qualified minority nationality teachers would be recruited in accordance with the standard class to teacher to staff ratio. To encourage poor minority students to enroll in these schools, parents would be exempted from school fees. For Tibetans and Mongol herders in the 'move-in' area, the project would provide access to basic education for the first time. Similarly, in the 'move-out' area, poor Tibetan children would have access to Tibetan schools and as part of the cultural revival effort under the project.

Component Activities. The project would finance the construction of 38,603 square meters of new school building, including 10,062 for 11 primary schools, 23,545 square meters for 10 joint primary and junior secondary schools, and 4,996 square meters for the expansion of the existing No. 1 Secondary School in Dulan County. The project would equip the schools with 7,100 sets of tables and chairs, 830 sets of office desks and chairs, 1,390 beds for the student dormitories and 248 beds for single teacher dormitories. The project would also equip each school with one set of teaching equipment and library books, at 2.5 volumes per primary school student, 3.5 volumes per junior secondary school and 6 volumes per senior high school student. A vehicle for transporting daily necessities would also be provided. In addition, schools or education programs for minority communities would be constructed or developed elsewhere in Dulan County and the 'move-out' counties. 76

Implementation Arrangements. A leading group has been established in Qinghai Provincial Education Committee (PEdC), and a Deputy Director of the PEdC is the head of the group while the members come from all relevant Divisions of the PEdC. The provincial and Haidong Prefecture offices in charge of the implementation of the IDA-assisted Third Basic Education Project and Haixi Prefecture and Dulan County will be responsible for the overall coordination and daily implementation management of the component, including design and construction of school buildings, procurement of the facilities and books, etc. The technical standards and procedures for the Third Basic Education project vwould be applied to implementation of this component, including procurement guidelines and procedures. The Education Bureaus of the move out counties will also monitor the implementation progress, especially in assigning teaching staff, arranging teacher training, and administering poor student assistance, etc. The Bureaus will also take care of the new schools management and operations before the schools are transferred to the 'move out' counties.

I. Component 8 - Institutional Building and Project Management (US$10.6 million)

Institutional Overview

Project Leading Groups and Project Management Offices would have principal responsibility for overall project implementation. At the Provincial/Regional level, Project Leading Group (PLGs) have been set up to: i) mobilize institutional, technical, and financial resources and support for the preparation and implementation of the project; and ii) provide coordination and monitoring functions. In addition, PMOs have been established to ensure effective and timely implementation of the project. At levels from the Prefecture/League/Municipal down to Township/Sumu levels, PLGs and PMOs have been set up. Technical Bureaus at the County level directly involved in the project would, in close cooperation with the Township/Sumu government and their officers at the administrative village level, play a key role in the day-to-day implementation of the project. The Provincial Health and Educational Departmnents would take leading roles in preparation and implementation of the health and education components. Technical Advisory Groups at both the Provincial/Regional and County/Banner levels have been set up to provide technical support for project implementation. Township Water Stations and Village Project Implementation Groups have been set up at Township/Sumu and Village levels respectively. The State Council's Leading Group for Poverty Reduction and Development of the State Council and the Foreign Capital Project Management Center at the central level would be responsible for coordination, guidance, monitoring and provision of services with regard to project implementation. The State Statistical Bureau would undertake independent poverty monitoring.

Provincial/Regional Level. In Gansu, the World Blank Loan Executive Committee (WBLEC) headed by the Executive Vice Governor, is the highest decision body and exercises unified- leadership on issues related to all World Bank assisted projects. The World Bank Office (WBO), set up in the PFB, is the secretariat of WBLEC, and deals with day-to-day work on behalf of WBLEC, including coordination, supervision, external liaison and financial management.

In Gansu, the PLG is headed by two Vice Governors responsible respectively for finance and agriculture and works under the unified leadership of WBLEC. In the case of Inner Mongolia, a Deputy Party Secretary General serves as Chairman, and a Vice Chairman of the People's Congress and a Vice Governor in charge of poverty reduction serve as deputy Chairmen. PLGs are assisted by representatives from the Provincial Poverty Reduction Office, Finance Bureau, Development and Planning Commission, the several technical bureaus directly involved in the project (including the Agriculture, Forestry, Animnal Husbandry, Water Conservancy, Rural Enterprises, Health, Power Supply, Labor, Statistics and Transportation Bureaus, Economic 77

Cooperation and Foreign Trade Commission, Provincial Women's Federation and Agricultural Bank of China, etc.). In Qinghai the PLG is headed by the Executive Vice Governor, and a Vice Governor responsible for poverty reduction is the Deputy Head. PLG is assisted by Directors from the Provincial Poverty Reduction Office, Finance Department, Development and Planning Commission, the several technical bureaus directly involved in the project, including the ABC, Departments of Agriculture and Forestry, Water Conservancy, Animal Husbandry, Public Security, Health, Civil Affairs, Construction and Transportation, Commissions of Education, Minority Nationality, Science and Technology, Bureaus of Rural Enterprises, Power Supply, Labor, Statistics and Environment, Land Administration, and Women's Federation. It includes also Commissioners of the Haidong and Haixi Prefectures. The main responsibilities of the PLGs are to formulate project-related policies and plans, approve sub-project designs, coordinate among various technical departments/bureaus, and between the 'move-in' and 'move-out' areas in Qinghai, and mobilize institutional, technical and financial resources and support for implementation of the project. PLGs would also provide coordination and monitoring functions.

PMOs have been established and would serve as the secretariats for the PLGs. PMOs would be responsible for managing overall project implementation. During project implementation, most of the PMO staff in Qinghai would be stationed in Dulan County, as the Provincial Project Executive Office (PPEO) responsible for the overall project management. The PMOs' responsibilities include: i) review and approve feasibility studies prepared by the Project Counties and consolidated by Prefectures/Leagues/Municipalities; ii) draw up annual work programs and budgets; iii) review and approve lower-level PMO work plans; iv) approve disbursement of Bank and government funds to local government accounts; v) arrange the preparation of and furnish to the Bank detailed project accounts, audits, withdrawal applications and progress reports; vi) arrange common procurement; vii) provide guidance to and coordinate the work of project Prefectures/Leagues/Municipalities and Banners/Counties; and viii) undertake project monitoring of physical and financial progress and evaluation of project impact. They would also directly implement the institutional development component. The PMOs would sign service agreements with participating agencies defining their detailed responsibilities and duties, rights and obligations during project implementation. In Qinghai, the PMO and PPEO would work closely with the provincial technical agencies in implementing the various components. PMOs have total staff positions of 20, 35 and 45 in Gansu, Inner Mongolia and Qinghai, respectively.

The Xiangride River Water Conservancy Administration (XRWCA) in Qinghai would be established by the Provincial Department of Water Conservancy, following completion of the water conservancy component. XRWCA would in close cooperation with PMO and PPEO, manage large-scale water conservancy works and ensure dam safety.

A Labor Mobility Development Center (LMDC) has been set up under the Provincial PMO of Gansu, to be responsible for the labor mobility component. The Provincial Labor Department in Gansu would assist PMO/LMDC implementation of the labor mobility component.

The Provincial/Regional Departments of Health and Qinghai's Education Bureau would take the leading role in implementation of the health and education components respectively.

Technical Advisory Groups (TAGs) have been established, composed of specialists in the fields of agriculture, forestry, water conservancy, environment protection, and health. The Group's responsibilities are to: i) review feasibility studies of various project components and correctness of various economic and technical indicators; and ii) provide technical information, guidance, advice, and training during project implementation.

Rural Enterprise Working Groups are set up to: i) study and formulate policies pertaining the rural enterprises component; and ii) review feasibility studies of the component. 78

Entrusted by WBO, ABC would be responsible for provision of fertilizer credit to farmers and loans (in Gansu and Qinghia) for the rural enterprise component based on the MOU signed between ABC and the PMOs.

Prefecture/League/Municipal Level. PLGs in Inner Mongolia have been set up in all the six project Leagues/Municipalities. They are headed by the Commissioners. The Deputy Commissioners in charge of poverty reduction serves as the Deputy Head. Project related technical bureaus are members of the PLGs. In Qinaghai,PLGs have been set up in Haidong ('move-out') and Haixi ('move-in') Prefectures. They are headed by the Commissioners.

PMOs have been set up, with the number of positions varying from 3 to 5. They have the status of a section in the Poverty Reduction Office. Their responsibilities include: i) serving as a contact between the Province/Region and County/Banner; ii I channeling Bank funds to the lower level; iii) ensuring provision of government counterpart ftnd of lower levels; iv) providing technical support to the lower level; and v) financial management and ensuring repayment of loans and credit. PMOs in both 'move-in' and 'move-out' areas of Qinghai have been set up. They have the status of a section in the Poverty Reduction Office (Haidong), or the Integrated Agricultural Development Office (Haixi). The responsibilities of the PMOs in the 'move-out' areas include: i) serving as a contact between the Province and the Counties; ii) channeling Bank funds to the lower level; iii) providing guidance and coordinating work of the project counties; iv) providing technical support to the lower level; and v) financial management. Another responsibility of the PMO in 'move-in' areas is to facilitate project implementation.

County/Banner Level. PLGs have also been established to fully mobilize existing institutional resources and guarantee coordination among various agencies and units. Chaired by magistrates, PLGs would take full responsibility for project implementation in their respective County/Banner. In Qinghai, PLGs have also been established at the County level, both in 'move-in' and 'move- out' areas, to fully mobilize existing institutional resources. Chaired by the County magistrate, the County PLGs would take full responsibility for project implementation in their respective County.

Each PMO has 10-15 staff positions. They serve as the secretariats for the County PLGs, and the contact with project organizations at higher levels. They would also be responsible for project monitoring and evaluation, financial management, arrangement of common procurement and distribution of goods and services procured, and ccmmon training components in conjunction with line agencies at the County level. The County PMVOsserve as the secretariats for the County PLGs, and the contact with project organizations at higher levels. Responsibilities of the PMOs in Qinghai's 'move-out' areas would be to: i) prepare armual settlement plan and budget; ii) review and approve the migrants lists furnished by townships and villages, iii) approve moving plans for teachers, doctors and administrative staff who would work in the 'move-in' area and ensure payment of their salary; iv) establish new townships in the 'move-in' areas; v) project implementation, monitoring and evaluation; vi) financial management and loan repayment. The PMO in the 'move-in' area would: i) prepare and furnish a land development program and ensure land availability; ii) coordinate use of water, power and roads in the development area; iii) set up administrative agencies in the newly developed areas; and iv) provide land for new township seats.

Technical bureaus directly involved in the project vwould,in close cooperation with TWSs and VPIGs, be responsible for the day-to-day execution of the project in their corresponding area, based upon a service agreement that would be signed with the County/Banner PMOs. TAGs have also been set up. Their function is similar to that at the Provincial/Regional level. 79

Township/Sumu Level. Township Workstations (TWSs) are responsible for formulating the village work program, credit delivery and recovery, goods distribution, preliminary monitoring and evaluation of project activities, technical guidance and training and provision of market information to the households. Township/Sumu Extension Stations would be comprehensive technical units integrating crop production, animal husbandry, grassland, forestry, water conservancy, and other technicians. Local nongovernmental organizations, such as All China Women's Federation (ACWF), would also be involved in the implementation of the project in activities such as farmers' training and monitoring and evaluation.

Village/Gacha Level. In addition to members of the Village/Gacha Committee, VPEGs would include at least one household representative. The Technical Service Groups would include at least one technician, who is seconded from the Township/Sumu and would stay at least 4 months per year in the village, and 2-3 skilled farmers. Their major responsibilities include assisting the PMOs at the County/Banner and Township/Sumu levels to establish the "Project Implementation Manual" or "Household Beneficial Card;" select migrants; monitor and supervise project implementation by farmer households; undertake technical training and credit recovery and provide market information.

Role of Beneficiaries

Throughout project preparation, project households have been actively involved in designing a menu of options to meet their development needs. They have been involved in farm model and village design through household interviews. Beyond this participation in project activities, they would also be involved in project monitoring and supervision through provision of requested data and submission of special reports. By emphasizing integrated village development and household training, the project would demonstrate and strengthen the capacity of village organizations to manage their own resources. This would become a key factor in sustaining land and farmer development. Particular attention has also been given to ensuring that women are actively involved in the project and derive positive benefits from it. Prospective migrants in Qinghai are informed about the settlement scheme, including benefits and risks, obligations and right to refuse settlement through a widely distributed Settlement Manual and the Information for Migrants. Project objectives and activities have also been included in the Manual and Information for their selection.

Various economic entities would be engaged in implementation of the project. In addition to individual project farm households, joint household entities with various forms of ownership would be engaged in more commercially oriented agriculture and forestry activities. Joint ventures between farmers and local business would also be engaged in commercially oriented large scale and intensive production.

Staff Development

Careful implementation of the project requires great strengthening of capability of PMOs at various levels. To this end, an institutional development program would be implemented as an integral part of the project. Human resources development, including training and technical assistance, is one of the important areas. Attention would be paid to development planning; project selection and appraisal; and monitoring and evaluation of project implementation. The project would fund local and overseas training, study tours and technical assistance comprising both foreign and local consultants. Each project component includes training and technical assistance. With the exception of technical assistance for institution building, all project technical assistance would support project preparation and implementation. Special attention would be paid to the management of the large scale voluntary settlement scheme; drainage, soil and water salinity control; and operation, maintenance and management of the large irrigation schemes. 80

Implementation And Supervision Arrangements

Implementation. The Provincial PMO and PPEO in Qinghai would work in close cooperation with the Provincial Department of Water Conservancy on planning, designing and implementation of the Keri and Hatu dams, Tuosou headworks, main Canals, and power transmission line of 35 and 10 kilowatts. PMOs at Prefecture and County levels would be responsible for construction of branch, lateral and sub-lateral canals, agriculture, forestry and livestock activities, as well as the voluntary settlement scheme.

Supervision. In collaboration with Provincial and County PMOs, Foreign Capital Project Management Center (FCPMC) would be responsi'ble for supervising the project's overall progress. FCPMC would carry out its own annual on-site supervision, in addition to participating in the Bank supervision missions. Results of such visits would be reported to the Leading Group of Poverty Reduction and Development (LGPRD) of the State Council and the World Bank and distributed to the PPs. In Qinghai, the Provincial PMC) and PPEO would take major responsibility for project supervision. They would frequently visit the project sites. Technical agencies would supervise activities under their respective sector. Findings of such on-site supervisions, including problems identified and corrective actions required, would be included in the semiannual progress reports. In addition, supervision would also be done in the form of household interview, regular reporting, random sample survey, spot observation in the sample villages; sample analysis and questionnaire, through which the beneficiaries would be actively involved in project supervision.

Physical and financial progress reports would be furnished by County/Banner PMOs to the Prefecture/League/Municipal PMOs on a quarterly basis, which would then furnish the consolidated reports to the Provincial/Regional PMO. Based on these, the Provincial/Regional PMO would prepare the semiannual progress reports for submission to the World Bank. The Provincial/Regional PMO would prepare, and furnish to the World Bank no later than six months following the closing date, an Implementation Completion Report for the whole project, based on the County/Banner and Prefecture/League/Municipal completion reports.

Arrangement for Project Impact Monitoring

The project's socio-economic impact on and well-being of project farmers would be monitored, evaluated and disseminated by the State Statistical Bureau (SSB) based on an MOU which would be signed with FCPMC on behalf of the Provinces/Region. The Qinghai Plateau Geographic Institute would be responsible for settlement monitoring. The Provincial Institute of Environment Sciences would be responsible for environmental impact monitoring. The Xiangride Water Conservancy Administration would be responsible for groundwater monitoring.

Design, testing, and modification of survey questionnaires for baseline and annual follow-up- surveys would be initiated in the first half of 1999. The baseline survey would be undertaken in December 1999 and the follow-up surveys would be carried out annually in December till year 2004. FCPMC and SSB would collaborate on data processing and analysis, and the publication of an annual summary report. 81

Annex 3: EstimatedProject Costs

CHINA:Western Poverty Reduction Project

The total project costs are estimated at US$311.1 million with a foreign exchange cost of US$95.7 million or 31 percent of the total (excluding the US$0.6 million front-end fee). Cost estimates are based on quantities derived from the preliminary designs, from unit prices currently prevailing for similar construction works in the project areas and from price lists of local and overseas manufacturers of equipment and machinery. The base cost estimates of US$275.3 million are expressed in December 1998 prices and the exchange rate used to convert base cost estimates and physical contingencies is Yuan 8.3 to US$1. Price contingencies of US$18.2 million for costs incurred in foreign exchange are based on annual international price escalation rates of 1.3 percent for 1999, 2.6 percent for 2000, 2.7 percent for 2001, 2.6 percent for 2002, and 2.5 percent for 2004-2005. Price contingencies for costs incurred in Yuan are based on annual domestic price escalation rates of zero percent for 1999, 3.0 percent for 2000, and 5.0 percent for 2001-2005. No price contingencies have been applied to overhead and independent monitoring expenses and costs relating to education and rural enterprise. Physical contingencies, equivalent to US$17.6 million, are based on an average rate of 10 percent for works and 5 percent for equipment, inputs and materials and services. No physical contingencies have been applied to vehicles, miscellaneous costs, and recurrent costs. Project costs are summarized in Table 3.1 below.

Table 3-1: Project Cost Summary

A. Land and Household Development 90.2 34.3 124.5 B. Irrigation and Land Improvement 48.6 25.6 74.2 C. Rural Infrastructure 13.4 11.1 24.5 D. Rural Enterprise 8.0 5.5 13.5 E. Labor Mobility 13.4 0.4 13.8 F. Voluntary Settlement 5.8 0.7 6.5 G. Social Sector Services 4.7 3.4 8.1 H. Institutional Building & Project 6.6 3.6 10.2 Management _

Total Base Cost 190.7 84.6 275.3 Physical Contingencies 12.0 5.6 17.6 Price Contingencies 12.7 5.5 18.2

Total Project Cost 215.4 95.7 311.1

Front-end fee 0.0 0.6 0.6 Total Financing Required 215.4 96.3 311.7 82

Table 3-2: Project Cost Summary - Gansu

ON- _ A. Land and Household Development 43,518.8 16,608.7 60,127.5 B. Irrigation and Land Improvement 13,900.4 2,600.6 16,501.0 C. Rural Infrastructure 7,096.1 5,342.9 12,439.0 D. Rural Enterprise 5,923.6 4,031.9 9,955.5 E. Labor Mobility 13,371.2 415.4 13,786.6 F. Social Sector Services (Health) 1,529.5 768.8 2,298.3 G. Institutional Building & Project 3,112.9 1,428.8 4,541.7 Management

Total Base Cost 88,452.5 31,197.1 119,649.6 Physical Contingencies 5,178.9 1,776.9 6,955.8 Price Contingencies 6,070.2 1,911.2 7,981.4

.Total Project Cost 99,701.6 34,885.2 134,586.8

Front-end fee 234.0 234.0 Total Financing Required 99,701.6 35,119.2 134,820.8

Table 3-3: Project Cost Summary - Inner Mongolia

A. Land and Household Development 37,940.2 13,069.8 51,010.0 B. Irrigation and Land Improvement 10,614.7 8,513.5 19,128.2 C. Rural Infrastructure 3,333.3 2,980.4 6,313.7 |D. Rural Enterprise l 1,384.9 1,293.4 2,678.3 E. Social Sector Services (Health) 1,145.6 1,264.0 2,409.6 F. Institutional Building & Project 1,683.2 1,212.3 2,895.5 Management

Total Base Cost 56,,101.9 28,333.4 84,435.3 Physical Contingencies 3,603.9 1,762.2 5,366.1 Price Contingencies 3,740.4 1,781.9 5,522.3

Total Project Cost 63,446.2 31,877.5 95,323.7

Front-end fee 233.0 233.0 Total Financing Required 63,446.2 32,110.5 95,556.7 83

Table 3-4: Project Cost Summary - Qinghai

B. Irrigation and Land Improvernent 24071.0 14,456.3 38,527.3 C. Rural Infrastructure 2,978.3 2,721.8 5,700.1 D. Rural Enterprise 730.8 178.4 909.2 E. Voluntary Settlement 5,789.4 746.1 6,535.5 F. Social Sector Services 2,017.3 1,392.0 3,409.3 G. Institutional Building & Project 1,829.6 906.3 2,735.9 Management _ a

Total Base Cost 46,153.5 25,065.1 71,218.6 Physical Contingencies 3,192.6 2,066.0 5,258.6 Price Contingencies 2,932.3 1,780.3 4,712.6

Total Project Cost 52,278.4 28,911.4 81,189.8

Front-end fee 133.0 133.0 Total Financing Required 52,278.4 29,044.4 81,322.8 84

Annex 4: Cost BenefitAinalysis Summary

CHINA:Western Poverty Reduction Project

Summary Base Year: 1998

:Acdi tvfeeFFR, NPV/a- . - Jia2-

Gansu 26.6 7483,807 28.4 891,456 15.9 18.9 1 Land and Household Development 18.9 21 [,743 19.5 230,662 13.5 15.7 2 Rural Infrastructure 14.5 '7,213 15.3 15,351 12.1 14.2 (1) Rural Road 19.1 10,922 15.8 19.1 (2) Rural Water Supply 13.7 :2,893 13.4 2,493 11.1 13.3 (3) Rural Electrification 15.6 4,319 13.5 1,936 9.2 10.6 3 Rural Enterprise 39.0 313,508 50.7 481,028 19.5 22.7 4 Labor Mobility 52.0 81,670 35.7 55,565 9.0 10.1

Inner Mongolia 18.6 246,823 20.6 328,390 10.7 12.3 1 Land and Household Development 20.0 25B,256 22.1 327,457 11.0 12.7 2 Rural Infrastructure 14.4 4,867 14.4 6,227 11.3 13.3 (1) Rural Road 16.4 2,052 14.0 15.5 (2) Rural Water Supply 13.4 713 13.3 659 10.4 12.3 (3)RuralElectrification 14.7 4,155 14.2 3,516 10.8 13.0 3 Rural Enterprise 35.6 19,155 50.1 30,252 15.3 18.6

Qinghai 13.1 38,066 13.7 62,136 11.6 13.2 1 Land and Household Development 15.5 101,238 16.2 122,339 13.7 15.2 2 Rural Infrastructure 14.6 6,811 14.5 7,417 11.8 13.6 (1) Rural Road 17.2 1,395 14.0 16.6 (2) Rural Water Supply 14.7 756 15.0 836 10.2 12.1 (3) Rural Electrification 14.6 6,055 14.2 5,186 11.7 13.5 3 Rural Enterprise 34.8 3,234 42.9 5,117 13.9 16.0

Total Project 20.3 1,033,696 21.9 1,281,983 13.5 15.1 /a Based on 12% discount rate.

A. Projecl Objectives

The Project's objective is to alleviate chronic poverty for the absolute poor living in remote and inaccessible villages of Gansu, Qinghai provinces and Inner Mongolia Autonomous Region. The Project takes an integrated approach, investing in: i) land and household development; ii) irrigatiqon and land improvement works; iii) rural infrastructure including rural road, water supply and electrification; iv) rural enterprise; v) labor mobility in Gansu; vi) a voluntary settlement program in Qinghai; vii) education and health; and viii) institutional building and project management. 85

Beneficiaries

The main beneficiaries of the project are poor farmer and herder households. It is estimated that about 1.67 million persons in 377,800 households would benefit directly from various project activities including 1.54 million from agricultural production activities, 65,700 persons from labor mobility and 20,200 from rural enterprises. The actual size of beneficiary households would vary from county to county and among participating provinces but the average size of beneficiary household is 4.4 persons including 2.4 adults and 2.0 children. In the project analysis, five typical household sizes ranging from 3 to 7 have been adopted. The annual incremental production of the main agricultural products at full development is estimated at: 269,000 tons of grains; 150,000 tons of cash crop products; 70,000 tons of fruit; and 40 tons of fresh water aquatic products and 1,457,000 cubic meters of timber (at final cut). An annual production of 3.2 million heads of large and small animals, including rabbit and chicken, and 3.0 million tons of other animal production including wool, cashmere, silkworm cocoon, goat milk and egg, are estimated at full development. As a result of higher and improved feed production, the animal herd would increase by an average of 5 percent for large animals and 3 percent for small animals annually in Inner Mongolia. Due to limited grassland availability in Gansu and Qinghai project areas, the primary focus would be on increasing feed production and modest increase in livestock population.

B. Financial Analysis

Agricultural Farm Budgets

Illustrative production models have been prepared for main crops, livestock, and forest plantation. To reflect the variations in natural conditions and performance of crops, livestock and forest plantation among project areas in participating provinces, a total of 38 crop models, 18 tree crop models and 22 livestock and fishery models have been prepared which are computerized using FARMOD software and are available on file. Production assumptions and incremental income projections for these models would be summarized in the PIP. The with-project production coefficients are based on actual results achieved under similar conditions and practices. However, under the without project scenarios, yield increases are not assumed to reflect the severe land improvement limitations, water scarcity and lack of access to technical inputs.

Farm Models

The farm and livestock models would illustrate the impact of the project on typical farm households in terms of production costs, farm labor use, food production, farm incomes, and return to labor. These models are indicative and farmer households are encouraged to make free choice of production activities in any combinations during project implementation as long as the basic principles agreed during project preparation are followed. These would include technical norms established for these activities, upper limit of per capita investment (not to exceed the agreesd level), poverty reduction targets, resource availability (land, water and organic manure), market opportunities for produce, adequate income to meet all costs and for repayment. The per capita income (at 1998 price) without project for all these models is only about Yuan 440. With project, these households' financial situation would improve substantially to reach a per capita income of Yuan 946 for Gansu, Yuan 1,251 for Inner Mongolia and Yuan 1,614 for Qinghai at full development. Food security problems have also been addressed. With project, per capita grain availability would reach 320 kg for Gansu, 1,100 kg for Inner Mongolia and 1,250 kg-for Qinghai, all exceeding the target set by respective provincial governments for the project, sufficient to feed family members and supporting household-based livestock production activities.

The baseline survey for the project indicated surplus household farmn labor would meet the project's incremental demand for labor and other project activities would involve equipment and hire contractual labor for major irrigation works, rural infrastructure and medium size rural 86 enterprises. These project activities would offer additional employment opportunities, though short term, to surplus household labor in project areas. As a result, the return per labor day for household activities would reach Yuan 25 for Gansu, Yuan 19 for Inner Mongolia and Yuan 36 for Qinghai, much higher than the average opportunity cost of unskilled labor.

Results of Financial Analysis

Rural Water Supply and Rural Electrification. Rural water supply and rural electrification are essential components to poverty reduction. The project would support village-based and low-cost water supply facilities including piped and non-piped systems to provide safe drinking water for villagers and animals. The project would also support extension of power lines to draw electricity for household use, agro-processing, and other uses in selective project areas. Water and electricity charges are adequate to cover operation and maintenance costs and marginally recover the capital costs of these systems. The financial analysis for these two sub-projects shows marginal returns around 14 percent (between 13.4 percent and 15.6 percent). More detailed analyses for these sub- projects are carried out in economic analysis.

Rural Enterprises. The proposed rural enterprises are largely small-scale labor intensive activities for construction to process food, feed, fruit, fiber, and traditional Chinese herbs using locally available materials and technology. Some medium-sized processing facilities in Gansu would also be supported by the project that will include tea, alfalfa, and commercial pig farms. In Inner Mongolia, the project would support village-based f-atrmmachinery stations to provide improved mechanized fanning services. Based on representative rural enterprise investments, the financial projections show robust returns with financial rate of returns ranging from 39 percent in Gansu to 35 percent in Qinghai and 36 percent in Inner Mongolia. The returns would generate sufficient income to repay the loans and cover operating expenses. These investments would be attracted by an assured supply of raw materials, low labor costs, moderate market risk and favorable tax terms. The analysis of representative rural enterprises demonstrates that the prevailing price structure generates adequate incentives for investors.

Labor Mobility. The Labor Mobility component, designed to promote a voluntary employment placement program in Gansu, would provide on-the-job basic skill training and offer safe and stable off-farm employment opportunities both within and outside the province for about 65,700 surplus laborers as an effective means to reduce absolute poverty. The term "surplus laborer" implies household members of working age who are unemployed on a year-round basis. The analysis indicates that the financial rate of return for the component is over 52 percent based on a conservative estimate of the job retention rate (80, 70, 65 and 60 percent during the first, second, third and fourth year of employment respectively) and a very generous living expense allowance (over 50 percent of the total wage income). The cash flow also shows that these laborers are in a strong financial position to support their household agricultural production activities and timely loan repayment. This high FRR is consistent with the results of similar programs under other_ World Bank supported projects in China. Although the potential risk for this component is moderate, this program would not succeed if institutional coordination is weak, labor market information system not in place and monitoring of worker conditions insufficient. 87

C. Economic Analysis

The provincial ERRs range from 14 percent in Qinghai and 28 percent in Gansu to 21 percent in Inner Mongolia. The overall economic rate of return for the project is estimated at 22 percent.

Results for Project Components

Land and Household Development. The investment costs for the economic analysis include the crop, livestock, and irrigation development costs. Since water in all project areas is limited due to competing uses particularly for agriculture, the real cost of supplying irrigation water provides a reasonable approximation of its opportunity cost. With the project, water charges would increase significantly to cover all O&M costs, major repairs, as well as depreciation to reflect the economic cost of water. The ERR calculation included all investment costs for water conservancy activities and the analysis shows ERRs ranging from 16 percent for Qinghai and 19 percent for Gansu, to 20 percent for Inner Mongolia.

Rural Roads. About 340,000 rural villagers living in remote and upland project areas would benefit directly from rural road improvements. The benefits would be derived mainly from: i) transportation cost savings (lower vehicle operating costs and shorter distances); ii) higher production efficiency; and iii) time saving for traveling passengers. These benefits would further stimulate economic development in project areas by offering a wider choice for economic diversification. No toll would be collected since the roads are mainly creating access to remote townships and villages. The maintenance for these roads would depend on the revenues from existing road user charges, the road maintenance fee, and the local government's budgetary allocations. The justification for this component is not based on optimum average daily traffic, but on the provision of road access to remote rural communities and population.

Economic analysis was carried out for the entire sub-project as a whole for Inner Mongolia and Qinghai while in Gansu the analysis was based on detailed design for a representative 18.6 kilometer road. The ERRs range from 16 percent for Inner Mongolia and 17 percent for Qinghai to 19 percent for Gansu. Improved road access would generate other significant benefits including more favorable commercial activities, growth of small enterprises, better access to schools, extension services and health care, and more off-farm employment opportunities. The principal exogenous parameters that could affect project benefit include: the projected rate of economic growth in project areas, its impact on the demand for transport; the elasticity of users' propensity to use the new roads; and adequate provision of required investments.

Rural Water Supply. There is a severe shortage of water supply in all proposed project areas for both humans and animals due to low annual precipitation, uneven rain distribution, pollution or remote locations. The improved water supply will: i) reduce the amount of time and resources spent on fetching water; ii) reduce the incidence of water-borne diseases; iii) promote livestocJk production; and iv) provide water for small enterprises that will create employment opportunities and raise incomes for the local population. The proposed rural water supply component would provide both piped and non-piped water supply systems to benefit 310,000 people and 1.8 million animals in the three PPs. The economic justification of the rural water supply component is based on the users' willingness to pay since an active water market does not exist in rural China. A 1993 survey in similar areas in China under other World Bank supported projects showed that, without access to drinking water and where the water is transported in tanks towed by animal or farm tractors, people paid an average of Yuan 5 per cubic meter of water. In villages with existing water supply systems, people paid between Yuan 1.5 and Yuan 2.5 per cubic meter of water. This serves as a reasonable measure of people's willingness to pay. For the benefit calculation for this sub- project, the value between Yuan 0.5 to Yuan 2.1 (Yuan 0.3 to Yuan 1.3 in 1993 prices) have been used. This is a conservative estimate, since the observed value is much lower than the surveyed 88 value in real terms and many people would be willing to pay higher prices. Based on the above scenario, the average annual water charge paid by household for family members would account for 2,1 percent, 1.3 percent and 1.5 percent of their expected annual net income for Gansu, Inner Mongolia and Qinghai respectively. Accordingly, the rural water supply component would generate an ERR of 13 percent for Gansu and Inner Mongolia and 15 percent for Qinghai.

Rural Electrification. The rural electrification component would supply power to selected project areas where there is no or limited power supply. This is in line with the government strategy for rural electrification to meet the requirements for: i) household use; ii) rural enterprise development; and iii) irrigation and other uses. Ar. estimated 71,000 households would benefit under the project. A least cost and willingness to pay approach has been adopted for the economic analysis. The economic power prices range between Yuan 0.52 per kwh and Yuan 1.15 per kwh were used to calculate the expected revenue. These rates are within the range of actual power charges paid by villagers in neighboring villages with existing power supply. The average annual power charge for household use would account for 5.9 percent, 3.4 percent and 7.9 percent of the household expected annual net income for Gansu, Inner Mongolia and Qinghai respectively. The estimated ERRs of 14 percent reflect the returns to investment, operating and maintenance costs in economic prices.

Rural Enterprise. Support to rural enterprise woulct take a program approach. Investment in each rural enterprise would be appraised by ABC. Since imostproducts of these rural enterprises would be non-traded and competitive local markets exist in most project areas, the prevailing market prices for these products were used as economic prices in the analysis. Economic prices for traded products would be used. The financial prices for investment and other inputs were converted into economic prices using conversion factors. The ERR calculation was based on illustrative and typical rural enterprises resulting in ERRs of 51 percent for Gansu, 50 percent for Inner Mongolia and 43 percent for Qinghai.

Labor Mobility. Based on local experience, remittance for labor mobility could be a major source of cash income for poor households and the local economy. In many instances, laborers returning from off-farm employment would introduce new technology and other externalities with refreshed vision to benefit not only their own families but also the local economy. The ERR for labor mobility is estimated at 36 percent.

Fiscal Impact

The project would not lead to major modifications of existing tax instruments. However, government revenues in all three provinces would increase mainly from five different sources: i) increased water charges from newly reclaimed lancdin Qaidamu basin, improved irrigated land in Gansu and Inner Mongolia, and rural water supply; ii) agricultural tax on incremental agricultural production; iii) special agricultural and forest tax on incremental project products, iv) increased or new land lease fees collected on developed land supported by the project; and v) income tax from rural enterprises. It is projected that net present values of Yuan 156 million of water charges, Yuan 153 million of agriculture tax and land lease fee and Yuan 186 million of income tax in the 20- year time period would be expected.

EnvironmentalBenefits

This is a category B project and no serious environmental risks are foreseen. Tree plantation and grassland improvement activities would significantly ease soil erosion and reduce moisture loss and improve the micro ecological environment. Based on experience of Bank supported forestry development projects in China, the total plantation of 40,000 hectares of trees in the project areas would be able to sequester and fix 21 million tons of CO2 in total equivalent to the CO2 discharge of burning 29 million tons of standard coal during the whole growth period. 89

Social Benefits

The project would have a positive social impact although it has not been quantified. Social benefits resulting from the project would be important as experienced in many other projects in China. The project would result in: i) assured provision of good quality health care services to cover all population in project areas; ii) better teachers and facilities for school children and adult education; and iii) improved living standard and life style contributed by reliable supply of safe drinking water; available power supply; and improved access to rural roads.

Employment

Various project activities would require a large number of laborers. It is estimated that the project would create a total of 565,000 full-time jobs during the six-year project implementation period- and 2.5 million person-year jobs during the project's life time, of which about 40 percent would be for women.

Cost Recovery

The cost of chemical fertilizer, agro-chemicals, plastic film, seeds, feed concentrates, animals, and other agricultural and livestock inputs, and rural enterprises would be recovered fully from beneficiaries. The recovery of water conservancy and rural infrastructure investments would depend on project specific taxes and government budgetary allocations. The planned water charge and power charge would raise adequate funds to provide full recovery of operating and maintenance costs. The cost of research, training, consultant services, and institutional development would also be recovered either through fees levied directly on beneficiaries or through taxes paid on project production. All costs for health and education would be covered entirely by the government. Services provided by agricultural support stations and their staff would be recovered through service fees. The costs for the labor mobility component would be fully recovered through repayment made by those laborers successfully placed under the project. The cost recovery analysis of the representative households shows that at present tax rates, project charges would be able to cover project costs with a recovery index of 21 percent. Rent recovery averages 11 percent at full development and 84 percent discounted over the life of the project.

Main Assumptions

The cost-benefit analysis is based on: i) a 20-year project life; ii) a 12 percent discount rate; iii) total investment costs including physical contingencies; iv) input costs and output prices in 1998 constant prices; v) "with project" and "without project" scenarios; vi) one rotation for all timber and fruit tree species; and vii) representative investment models for the rural enterprise component.

Prices

All project outputs including surplus grains would be sold in the free markets at prices determined by market forces. In the analysis, financial prices for agricultural outputs are projected at average prevailing 1998 farm-gate prices in the project areas and those for rural enterprise products are projected at ex-factory prices. Crop byproducts and animal manure are valued at their market prices, though most of them are used on farm. All inputs used in the project are also based on free market prices. The financial prices of pesticides and passenger vehicles are based on actual import prices obtained from recent imports. The financial price of labor is a weighted average of the various daily wage rates currently prevailing in the project areas.

World Bank Global Commodity Markets price projections or average export and import prices applicable to China, with adjustment for processing, handling and transportation costs, were used 90 to estimate farm-gate economic prices in 1998 constant terms for major traded inputs and outputs. Investment costs have been expressed in economic values using the applicable conversion factors. For non-traded commodities, a standard conversion factor of 1 has been assumed in the analysis. To estimate the shadow wage rate, a conversion factor of 0.9 has been used for unskilled labor hired locally, reflecting prevailing surplus agricultural laborers with limited alternative employment opportunity in project areas, leading to a marginal value product of labor below market wages. All direct transfers (taxes, debt services etc.) are not included for the economic analysis.

All prices are in or converted to local currency at the prevailing official exchange rate of US$1 to Yuan 8.3. All conversion factors and the prices of inputs and outputs used in the project financial and economic analysis are presented in detail in the PIPs.

Taxes

The following taxes and fees levied by either the central or local governments have been incorporated in the financial analysis: * Agricultural tax varies by province and land classification ranging between Yuan 8 to Yuan 15 per mu of cropped land. Under the Project, farmers in Qinghai would be exempted from paying agricultural tax during the first three years; * A land lease fee would be introduced to recover the cost of land development. The rates vary according to location, soil quality, and access to irrigation. On average, Yuan 80 per mu would be charged annually after a 2-year period of exemption for crops (including green manure crop plantation in the first year); * A special agricultural and forestry producittax is levied at rates of about 10-15 percent for fruits, 10 percent for aquatic products and 8 percent for timber; * A livestock sales tax is generally levied at 5 percent of animal sales; and * Income tax is an applicable and unified tax levied on all rural enterprises at a rate of 33 percent of net income. However, reduction or exemption of income tax for small scale and household-based rural food and feed processing activities is often granted by local governments.

Marketing Outlets

Continuing deregulation and expansion of the market economy throughout China have witnessed a parallel expansion of the market (and therefore opportunities) for all agricultural products. The number of individual and commercial traders has expanded enormously throughout the country. Similarly, while the expansion is not so marked, the transportation sector has also witnessed growing private sector involvement, especially for local trade where smaller quantities prevaif These trends have greatly increased the marketing opportunities for farmers. Moreover, the freeing up of trade means that surpluses can move to deficit areas where the demand is higher, much more rapidly. Thus the danger of oversupply resulting in rnarkedly depressed prices has, in general, been greatly reduced. Localized problems of imbalances in supply/demand will no doubt emerge from time to time, especially in more remote areas. Notwithstanding this, the overall market situation for all agricultural products, given the steadily increasing per capita demand and comparatively rapid urbanization, should remain satisfactory during the life of the project.

The dangers of market imbalances within the two provinces is also somewhat mitigated by the comparatively small proportion that projected project output represents in the total provincial production of the main products concerned. The exception could be wheat, beef and mutton, rabbit 91

and possibly corn in Inner Mongolia where these products represent a significant proportion of current provincial production.

For grain, the State Grain Purchasing stations are present in all of the County/Banner towns. During the purchasing season, temporary buying stations are established in the main grain producing townships. Farmers sell wheat, maize or rice to the State at the official procurement price, which tends to be below market prices. Officially, the State purchasing agency is required to purchase all grain delivered by farmers. During the last two years in particular, however, this requirement has not been honored in many cases due to the bumper harvests experienced and the resulting surplus grain offered by farmers. Despite the government attempt to 'protect' farmer prices, they have, in many cases, been forced to sell in the market at prices below the official 'protected' price.

In addition to the State Grain Purchasing agency, private traders operate throughout the project area and the region. Open markets exist in all townships. Traders will buy in these markets and, in many cases, travel to the villages to negotiate purchases. Although there are many exceptions, villages with road access are within reasonable distance to the township markets. Depending on the volume, products are transported to the markets either by bicycle, tractor trailer or, in many cases, by private trucks. Where a marketable surplus is produced the volumes are invariably small.

For Inner Mongolia, within the project area, most farm products are able to be sold within local markets or to traders who transport them to more distant markets. Marketing of project area products is not expected to present a constraint in the medium term. As part of the research program in Inner Mongolia, a marketing study for all the main project products would be conducted under the project.

Livestock Marketing.

The important aspects of marketing have been addressed by the PPs using two approaches: The signing of forward contracts with livestock and meat processors and buyers; and * The inclusion into local PMOs of marketing specialists with specific tasks of market development and sales. Inner Mongolia. Significant numbers of sheep, cattle, pigs, rabbits and broilers are to be marketed under the project, and marketing arrangements are largely in place. The Shuanghui Group, China's largest livestock processing group and joint-owner enterprise with 50 processing plants across China, has recently purchased and renovated a combined beef, sheep and pork slaughtering and processing plant in the project area. The Group plans to procure sheep, cattle, and pigs from the project and has already signed supply contracts with seven of the fifteen project counties or Banners. Broilers will be produced on contract to a local poultry-processing plant, which provides feed and day-old chicks and also purchases the broilers. Rabbit marketing, in contrast to earlier reservations will not be problematic in Inner Mongolia where rabbit meat is popular with minority groups. The provincial PMO has ensured that livestock marketing specialists will be attached to all county or Banner PMOs to i) develop and maintain contacts with livestock processors and buyers; ii) assist in livestock marketing through organizing producers into marketing groups and setting up feedlofs for easy access to producers; and iii) prepare and update project county/Banner marketing reports and price information for project producers.

Gansu. Considerable efforts have been made to ensure that the marketing of livestock will not pose a bottleneck for the project. Demand for pigs, sheep, and cattle currently far outstrips supplies. The provincial PMO has already signed long-term (3-6 year) forward delivery contracts with nine out of twelve project counties that include guaranteed floor prices to protect producers. 92

Qinghai. Distances to adjacent markets are considerable and can pose a constraint to effective marketing. Fortunately, livestock output is relatively limited and largely targeted for household consumption. The balance will therefore be easier to market. At present there is no livestock in the project settlement area and livestock numbers and markets will increase gradually with growing livestock numbers. A forward contracting system will be set up between project households or villages and service centers, which in turn will pay the farmers and market the livestock. A marketing specialist will be attached to the PMO for this purpose. Other marketing channels include middlemen and marketing to the 'move-out' areas from whence the migrants came.

Sensitivity Analysis / Switching Values of Critical Items

Sensitivity analysis was carried out for each project component to test the impact of changes in costs and benefits. The switching values show that, except for rural infrastructure, the ERRs are relatively robust to wide changes in revenues and costs. Costs would need to increase by 15 percent or benefits to reduce by 13 percent, for the overall project ERR to drop down to 12 percent. All rural infrastructure components showed greater sensitivity to changes in both revenue and costs which reflects the concentration of the benefit stream on the limited quantifiable benefits from rural roads and the combined effect of limited benefits from large investments and the "willingness to pay" approach for the rural water supply and rural electrification sub-components. 93

Annex 5: Project Financing Plan

CHINA: Western Poverty Reduction Project (in US$ millionequivalent)

Protject Costs =-.*=== Investment Costs 46.0 72.9 64.0 56.4 42.5 26.4 308.1 99 5.1 . 0 Recurrent Costs 0.4 0.6 0.4 06 0.5 0.5 3.0 1 1,732.8 100 Total 46.3 73.5 64.4 57.0 43.0 26.9 311.1 100 1,737.9 100

Financing Sources /a IBRD/IDA/b 24.5 38.7 32.7 28.6 21.6 13.9 160.0 51 Government /c Central Gov't 5.2 8.7 7.6 7.4 6.4 3.8 39.1 13 Provincial Gov't 10.5 16.1 14.7 13.1 9.7 6.0 70.1 22 34.8 2 Pref/County Gov't. 2.3 3.5 3.3 2.7 1.7 1.0 14.6 5 139.0 8 Beneficiaries /d 4.4 6.5 6.1 5.2 3.6 2.2 28.0 9 1,564.1 90 Total 46.9 73.5 64.4 57.0 43.0 26.9 311.7 100 1,737.9 100 Main Assumptions: /a Financing plan is based on total project cost by year including contingencies. /b Bank financing includes US$60 million of IBRD loan and USSI00 million equivalent of IDA credit. The first year financing includes required front-end fee of US$0.6 million. /c Government financing includes funds from various sources in both cash and kind. /d Beneficiaries' contributions are mainly in the form of labor and organic manure.

NB: The disbursement is by calendar year. 94

Annex 6: Procurementand DisbursementArrangements

Annex 6, Table A: Project Costs by Procurement Arrangements Procurement Method Expenditure Category ICB NCB Other /a NBF /b Total /c

1. Works 37.7 99.5 5.2 142.5 (21.9) (37.0) (0.0) (58.9)

2. Goods 3.7 6.6 98.3 14.1 122.8 (3.7) (6.6) (73.7) (0.0) (84.1) Equipment 2.8 12.2 2.8 17.8 (2.8) (9.2) (0.0) (12.0) Inputs and Materials 2.8 3.6 86.0 10.1 102.5 (2.8) (3.6) (64.5) (0.0) (70.9) Vehicles 1.0 0.2 0.1 1.2 2.4 (1.0) (0.2) (0.1) (0.0) (1.2) 3. Services 8.1 0.5 8.6 (8.1) (0.0) (8.1) Training and Study Tours 5.5 0.5 6.0 (5.5) (0.0) (5.5) Consultant Services 2.6 2.6 (2.6) (2.6) 4. Miscellaneous /d 15.4 22.0 37.3 (8.3) (0.0) (8.3) Total /e 3.7 44.3 221.2 41.8 311.1 (3.7) (28.5) (127.1) (0.0) (159.4) /a Other procurement methods include works for well and purrmpedirrigation, bridges, piped water supply, rural electrification in Gansu and Inner Mongolia, minor office buildings in Qinghai and construction of processing units in Gansu, force account and beneficiary participation for crop and livestock establishment, rural infrastructure, shopping for minor equipment, vehicles, inputs, sub-loans for chemical fertilizer, training and consultant services and monitoring activities. /b NBF denotes non-Bank financing. /c Total cost includes contingencies but excludes front-end fee. /d Including US$1.3 million for monitoring activities for project monitoring, poverty monitoring, environmental monitoring and voluntary settlement monitoring, US$14.1 raillion for labor placement, US$5.6 million for overhead and project implementation, US$2.4 million for survey and design, US$ 1.1 million for other monitoring,_ and supervision, US$6.3 million for settlement allowance, US$3 million for recurrent costs and US$3.6 million for other non-Bank funded activities. /e The totals may not tally due to rounding. Note: Figures in parentheses represent the amounts financed by lhe Bank including contingencies. 95

Annex 6, Table Al: Consultant Selection Arrangements (in US$ million equivalent) Consultant Services Expenditure Selection Method /a Total Cost (including Category QCB QBS SFB LCS CQ Other NB. contingencies)

A. Firms 0.5 0.5 (0.5) (0.5)

B. Individuals 2.1 2.1 (2.1) (2.1)

Total 0.5 2.1 2.6 (0.5) (2.1) (2.6)

/a QCBS - Quality and Cost-Based Selection, QBS - Quality-Based Selection, SFB - Selection under a Fixed Budget, LCS - Least-Cost Selection, CQ - Selection Based on Consultant's Qualifications, Other - Selection of Individual Consultants, Commercial Practices, etc.

Figures in parenthesis are the amounts to be financed by the Bank. 96

Annex 6, Table B: Thresholds for Procurement Methods and Prior Review Expenditure Category Contract Value Procurement Method /a Total Value Subject to Prior Review /b US$ thousands US$ millions 1. Works 2.0 = or >10,000 ICB 0.0 = or >200<10,000 NCB 2.0

2. Goods 5.4 = or >250 ICB 3.7 = or >100<250 NCB 1.7 <100 NS 0.0

3. Services 0.7 >100 for firms CG 0.5 >50 for individuals CG 0.2

Total 8.1 /a ICB - International Competitive Bidding, NCB - National Competitive Bidding, NS - National Shopping, CG - Selection method specified in the Bank Group Guidelines for Selection and Employment of Consultants by World Bank Borrowers. /b Thresholds for Bank's prior review are: all ICB contracts fcr goods, the first three NCB contracts each for works and good sin each PP, all contracts for consultant services greater than US$100,000 for firms and US$50,000 for individuals. 97

Annex 6, Table C: Allocation of Loan Proceeds IDA IBRD Total Category Amount Percent Disbursement (SDR (US$ (US$ million) million) million)

Gansu (1) Works 15.020 20.319 42% (2) Goods 16.443 16.126 38.435 100% of foreign expenditures; 100% of local expenditures (ex-factory cost) and 75% for other items procured locally (3) Consultants' Services and Training 2.679 3.634 100% (4) Monitoring Activities/a 0.248 0.337 100% (5) Labor Placement 7.040 7.040 50% (6) Front-end Fee 0.234 0.234 100% Sub-total 34.391 23.400 70.000

Inner Mongolia (I) Works 6.108 8.253 22% (2) Goods 11.008 23.067 39.001 100% of foreign expenditures; 100% of local expenditures (ex-factory cost) and 75% for other items procured locally (3) Consultants' Services and Training 2.304 3.126 100% (4) Monitoring Activities /a 0.285 0.387 100% (5) Front-end Fee 0.233 0.233 100% Sub-total 19.705 23.300 50.000

Qinghai (1) Works 17.153 7.126 30.363 59% (2) Goods 1.176 6.041 7.637 100% of foreign expenditures; 100% of local expenditures (ex-factory cost) and 75% for other items procured locally (3) Consultants' Services and Training 0.957 1.299 100% (4) Monitoring Activities /a 0.419 0.569 100% (5) Front-end Fee 0.133 0.133 100% Sub-total 19.705 13.300 40.000

Total 73.800 60.000 160.000

/a Including expenses for project progress monitoring by FCPMC, poverty monitoring, and environment monitoring by provincial EPBs and voluntary settlement monitoring by the Qinghai Plateau Geography Research Institute. 98

Annex 6, Table D. Estimated Disbursement Schedule

IBRD Disbursement Disbursement Profile /a Fiscal Year/ Semestral Cumulative Cumulative Year/ Cumulative Semester (US$ million) (%) Semester (%)

2000 Year I First (July-Dec. 1999) /b 13.3 13.3 8 First 0 Second (Jan-June, 2000) 3.3 16.6 10 Second 30

2001 Year 2 First (July-Dec. 2000) 9.1 25.7 16 First 38 Second (Jan-June, 2001) 12.4 38.1 24 Second 46

2002 Year 3 First (July-Dec. 2001) 19.4 57.5 36 First 54 Second (Jan-June, 2002) 18.9 76.4 48 Second 66

2003 Year 4 First (July-Dec. 2002) 16.3 92.7 58 First 74 Second (Jan-June, 2003) 14.7 107.4 67 Second 82

2004 Year 5 First (July-Dec. 2003) 12.9 120.3 75 First 90 Second (Jan-June, 2004) 12.9 133.2 83 Second 94

2005 Year 6 First (July-Dec. 2004) 8.6 141.8 89 First 98 Second (Jan-June, 2005) 7.8 149.6 93 Second 100

2006 Year 7 First (July-Dec. 2005) 6.2 155.8 97 First 100 Second (Jan-June, 2006) 4.2 160.0 100 Second 100

/a Disbursementprofile for agriculturalsector in Chinacompiled by OPRPG,June 30, 1995. /b Includingfront-end fee of US$0.6million and retroactivefinancing of US$12.7million to coverexpenditure incurredbetween January 10, 1999 andthe date of the signingof the Loan/CreditAgreements.

Completiondate: June 30, 2005 ClosingDate: June 30, 2006 99

Annex 7: ProjectProcessing Schedule

CHINA:Western Poverty Reduction Project

=~~~~

Time taken to prepare the project (months) 12 15 First Bankmission (identification) 10/04/97 10/04/97 Appraisalmission departure 01/07/99 01/07/99 Negotiations 04/12/99 04/12/99 Planned Date of Effectiveness 09/30/99

Prepared by: Provincial Governments of Qinghai and Gansu, and Inner Mongolia Autonomous Regional Government assisted by the Foreign Capital Project Management Center of the Leading Group for Poverty Alleviation and Development of the State Council.

Preparationassistance: Japanese PHRD Grant, FAO, Canadian and New Zealand Trust Funds.

Bankstaff who workedon the projectincluded:

Aklilu, Petros Rural Development, Task Team Leader Ashworth, Vincent Crops and Agricultural Services, Consultant Bidani, Benu Poverty Analysis & Monitoring & Evaluation Cheong, Chup Lim Water Conservation/Infrastructure, Consultant Cronberg, Christer Rural Enterprises, Consultant Crooks, Robert Peer Reviewer, Environment DeWitt, Nicolette Legal Advisor De Haan, Comelis Peer Reviewer, Livestock Durham, Sandra Financial Management and Control Goering, James Peer Reviewer, Rural Development-World Vision Han, Mengjie Health, Consultant Hu, Wenbin Education Kim, Hyung Min Disbursement Lin, Zongcheng Anthropologist Schumacher, Alex Environment, Consultant Sheehy, Dennis Livestock, Consultant Subbarao, Kalanidhi Peer Reviewer, Poverty Tay, Lang Seng Peer Reviewer, Irrigation and Infrastructure Wang, Xiaolan Financial Management and Control Zhou, Weiguo Cost Estimate, Procurement, Financial & Economic Analysis Zhu, Zhengxuan Organization and Management, Co-Task Team Leader 100

Annex 8: Documentsin the ProjectFile*

CHINA:Western Poverty Reduction Project

A. ProjectImplementation Plan Gansu: Project ImplementationPlan, May 1999 Inner Mongolia:Project ImplementationPlan, May 13999 Qinghai:Project ImplementationPlan, May 1999

B. Bank Staff Assessments IdentificationAssistance Mission: Back-to-Office Report, October 1997 TechnicalMission: Back-to-Office Report, February 1998 Project ConceptDocument (PCD), April 1998 Peer ReviewersComments on the PCD, April 1998 Minutesof Project ConceptDocument, April 1998 PreparationAssistance Mission: Back-to-Office Report, July 1998 Pre-appraisalMission: Back-to-Office Report, November 1998 ProcurementAssessment, November 1998

C. Other FeasibilityReports furnished by Gansu,Inner Mongolia,and Qinghai,September 1998 SocialAssessment Reports-Gansu (December 1998), Inner Mongolia (August 1998)and Qinghai (October 1998), EnvironmentalImpact Assessment Reports-Gansu, Inner Mongolia and Qinghai,November 1998 Land Acquisitionand CompensationPlan, Gansu and Inner Mongolia,December 1998 VoluntarySettlement Implementation Plan for Qinghai (includingResettlement Acton Plan), May 1999 Report on Land Availabilityin Qinghai,October 1998 Dam Safety Review,September 1998 Report on Livestock,October 1998 Reporton GroundwaterSimulation Model for BalongQinghai, October 1998 Assessmentof Land Suitabilityin Qaidanu Basin,August 1998 Evaluationof Land Suitabilityfor Irrigation,September 1998

*Includingelectronic files 101

Annex 9: SocialAspects

CHINA:Western Poverty Reduction Project

A. Overview

Introduction. The objective of the proposed Western Poverty Reduction Project (WPRP) is to alleviate chronic poverty for the absolute poor living in remote and inaccessible villages of Gansu Province, Qinghai Province and Inner Mongolia Autonomous Region within the People's Republic of China. The proposed project is the fourth in a series of World Bank-assisted poverty reduction projects in China that support the government's Eight-Seven Poverty Reduction.Plan (the 8-7 Plan). In all three of the Participating Provinces (PPs), project supported interventions would enable poor households to raise their incomes through increasing grain and livestock production sufficient to meet their basic food needs and in many cases also generating a marketable surplus to improve their living standards. This annex summarizes the most relevant findings and recommendations of the three provinces social assessments' (SA) prepared for the proposed project and outlines the World Bank's evaluation of the most substantial social issues.

Preparation of Social Assessments. Due to its nature as a targeted assistance activity, the proposed project required detailed social information for stakeholder and beneficiary identification, participation strategy and project design. Social Assessment (SA) documents were prepared for the three PPs by the Qinghai Plateau Geographic Research Institute, and the Agricultural and Animal Husbandry Engineering Design Institute of the Agricultural and Animal Husbandry College of Inner Mongolia. The SA documents for Gansu were prepared by local social scientists, coordinated by the Western Poverty Reduction Project Gansu Project Office. In addition, a Voluntary Settlement Implementation Plan (VSIP) was prepared for Qinghai which includes an Involuntary Resettlement Action Plan (RAP) as an annex, prepared in accordance with OD 4.30 Involuntary Resettlement. The SAs and the VSIP include Local Beneficiaries Action Plans. The SA documents and the VSIP were fumished to the Bank prior to appraisal and have been made available in the PPs to project-affected groups upon request. The SAs were carried out in close coordination with the environmental assessment work, under supervision of a World Bank social scientist taking into consideration OD 4.20, Indigenous People.

Lessons Learned. Since the project encompasses some multi-nationality districts, particularly in Qinghai, Bank guidelines and experience elsewhere in China, working in such with ethnically mixed areas were deemed relevant. The Gansu-Hexi Corridor Project, was reviewed as a useful model for lessons learned of voluntary resettlement from a degraded and desperately poor mountain area to a low-lying newly irrigated district with far greater subsistence and economic potential. Project start-up was not all smooth. One particular issue which arose was that, although a very thorough and adequate job of migrant selection and preparation was conducted, more attention could have been given to foreseeing the physical needs (primarily inadequacy of wate.- for desalinization) during project start-up. These experiences were taken into account in the design of the proposed project. Lessons learned from the Gansu Hexi Corridor Project, have been transferred to the proposed project.

Potential Adverse Impacts. The SAs have identified and analyzed the following potential adverse impacts of project activities on local society have: i) possible adverse environmental impacts; ii) land acquisition for road construction and enterprise establishment; iii) possibly inequitable distribution of benefits (including to women); and iv) issues related to ethnic minority rights. These findings are described below. In addition, the SAs recommended mitigation measures that have been included in project design and will be monitored during project implementation. Specific issues related to the voluntary settlement in Qinghai are described separately. The VSIP for Qinghai covers social, environmental and financial issues, presenting 102 the rationale for the voluntary settlement, selection criteria for migrants, procedures for migration, establishment of new sites and development of social services.

Use of Social Assessments in Design. The results of the SAs, obtained through beneficiary consultation and stakeholder analysis, were used to ensure that the project objectives and interventions are appropriate to the targeted beneficiaries and that the changes proposed in the agricultural, animal husbandry, and rural infrastructure components and Qinghai's voluntary settlement component, are acceptable to the affected population. The SAs also seek to ensure that changes brought about by the project are equitable and effective for all, irrespective of gender, ethnic, and social differences. Impacted stakeholders' views, have been analyzed and incorporated into the project design. Moreover the World Bank will pay great attention during project supervision to the proper implementation of these recommendations.

B. Social Assessment Preparation Process

Assessment of Methodologies. By and large, social assessments in China are carried out in a satisfactory manner. The administration of the social assessment methodologies, and particularly quantitative approaches are excellent. Qualitative approaches are less well developed, however, and even when implemented, often compromise respondent confidentiality. This is a common weakness in assessment techniques carried out in China but are usually offset in the initial data gathering stage through the use of structured interviews and focus groups. This is a common practice in poverty related projects in China that has proven a reliable methodology. This was also the methodology applied to the Project.

Social Surveys. The methodology of the SAs and their embedded poverty assessments were highly participatory in the form of interviews, focus groups and public meetings. In addition, the SA team in Qinghai conducted extensive socio-economic surveys in both the 'move-out' and move-in' areas. In the 'move-in' area, a 5 percent sample survey was conducted including herders, women, religious leaders, and local cadres. This survey also attempted to access minority nationality sentiment. Included in the sample were 8.4% of the area's Mongols, 9.9% of the Tibetans, 6.7% of the Hui, and 2.2% of the Han. A special survey was conducted for all 63 herder households directly affected by project land acquisition. Prior to finalization of project preparation, the SAs and the VSIP (for Qinghai) were discussed with selected households. Interviews were carried out by minority interviewers, where appropriate.

C. Selection and Characteristics of Project Areas

Inner Mongolia. 916 villages/gachas and 136 townships/sumus in 15 counties/banners were selected. All the counties/banners are nationally designated poverty counties, with over 31 percent poverty incidence. Priority has been given to townships with large numbers of ethnic minorities. The average annual per capita income of these poverty counties is Yuan 914, about- 57 percent of the provincial average. The annual per capita grain production in these poverty counties is 467 kg, about 70 percent of the provincial average. The project's direct beneficiaries are about 674,000 people in 167,100 households with an average per capita income during 1994-96 below Yuan 580 for farmninghouseholds and Yuan 950 for herders.

Gansu. 2,234 villages in 185 townships in 12 national poverty counties were selected as project areas, accounting for 23 percent of the 53 poverty counties in the province. About 29 percent of the population of these counties are classified as poverty households, about 1.5 times the provincial average (19 percent). The average annual per capita income of these poverty counties is Yuan 750 (in 1998 prices), about 68 percent of the provincial average. The annual per capita grain production in these poverty counties is 240 kg, about 59 percent of the provincial average. At the village level, the poverty rate of the project villages is 58 percent, with an average income 103 of Yuan 506 per year. Direct beneficiaries consist of about 929,600 people in 197,060 households with an average per capita income below Yuan 500.

Qinghai. There are 14 national poverty counties in Qinghai, of which six are located in the arid and environmentally degraded mountainous area to the east (Minhe, Xunhua, Hualong, Pingan, Huangyuan, and Datong). These counties have a total rural population of about 1.33 million people, about 114 people per square kilometers, over 18 times the average provincial rate. Minority nationalities account for 51 percent of the total population of these counties. The average annual per capita income of these poverty counties is Yuan 755, about 64 percent of the provincial average. The project is proposing to reclaim 19,200 hectares of wasteland in Xiangride Balong area of Dulan County, 450 kilometers from the six poverty counties listed above, for a voluntary settlement of about 57,800 poor from these counties. Dulan County has a total land area of 5.2 million hectares. The proposed project area is located in Balong and Zongjia townships, which have a total land area of 1,563,800 hectares. Three subsistence strategies characterize the area: the nomadic herders (60 percent), the sedentary herders (9 percent), and farmers (31 percent).

Ethnic Characteristics of Project Provinces. The total population in three project provinces comprises various ethnic population groups with Han Chinese in predominance. In Gansu, Han make up 88 percent while the minorities (largely Hui, Tibetan, Mongol, Kazak, and Yugu) accounts for 12 percent. In Inner Mongolia, the main groups are Han (79 percent), Mongol (16 percent), Manchu (2 percent), Hui (1 percent), and other minorities (2 percent including Dawer, Korean, Ewenke etc.). In Qinghai, the total population is made up of Han (57 percent), Tibetan (21 percent), Hui (14 percent), Tu (3 percent), Salar (2 percent), Mongol (2 percent) and other minorities (1 percent).

Ethnic Characteristics of Project Areas. The composition of beneficiaries in project areas varies among project provinces. In Gansu, Hui account for 7 percent while Han are 93 percent. In Inner Mongolia, about 14 percent of the beneficiaries are minorities and 86 percent Han. The ethnic composition of project beneficiaries in Gansu and Inner Mongolia is largely in proportion to the population composition in the project counties/banners. In Qinghai, the migrants from the six 'move-out' counties are composed of Han (42 percent), Tibetan (6 percent), Hui (36 percent), Salar (7 percent), and Tu (9 percent). The 'move-in' county (Dulan) is comprised of Han (53 percent), Tibetan (23 percent), Mongol (14 percent), Hui (7 percent), and other minorities (3 percent). The project-affected population in 'move-in' townships is Mongol (69 percent), Han (26 percent), and Hui (5 percent).

D. Stakeholder Analysis and Selection of Beneficiaries.

Selection of Beneficiaries. In Inner Mongolia, only nationally designated poverty townships with more than 30 percent of poor villages can participate. Participating villages must have at least half of their households to be a poverty household. In Gansu project Townships are required to have an average per capita income below Yuan 630 and per capita grain production below 300 kg. Project villages are required to have an average per capita income below Yuan 580 and per capita grain production below 300 kg.

Selection of the 'move-out' population in Qinghai was based on socio-economic surveys and beneficiary consultations. Designated poverty townships and villages were selected for resettlement. Household participation in the program is fully voluntary. The household surveys indicated that a vast majority (over 90 percent) would like to move. Those wanting to move stated reasons such as scarcity of arable land, poor soil, and inability to improve living conditions. Those unwilling to move expressed strong attachment to their native place, enjoy comparatively better living conditions, and are afraid of an unknown environment. Households' participation in 104 the project was elicited and mobilized through radio broadcasting, bulletins, village meetings, and direct door-to-door discussions. Interested households have submitted an application expressing their interest to participate. All together about 170,000 people have expressed their interest in resettling. A preliminary selection has been made with an ethnic composition of Han (42 percent), Hui (36 percent), Tibetan (6 percent), Salar (7 percent), and Tu (9 percent). Final selection of households will be published for ptilic viewing.

Areas participating in the project were selected as follows: townships with more than 30 percent of population below poverty line; villages with more than 50 percent of population below poverty line, with a minimum of 40 percent for minority villages included among project villages. The most important household selection criteria are: * maximum annual per capita net income of Yuan 300 per year and grain production of 140 kg; * lack of potable water; and * strong aspiration for migration.

E. Project Impact and Mitigation Measures

The SAs identified and analyzed potential adverse social impacts of project activities on local society including: i) possible adverse environmental impacts; ii) land acquisition; iii) possibly inequitable distribution of benefits (including to women); and iv) issues related to ethnic minority rights. These issues and concerns are elaborated below.

Possible Adverse Environment Impacts

No major concerns regarding environmental issues were identified in the SAs in Gansu and Inner Mongolia because the bulk of the proposed project activities are expected to produce positive environmental impacts. In Qinghai, the main concern identified in the SA is related to the fragile ecology in the 'move-in' area. Much of the proposed irrigation land in the 'move-in' area consists of wind-formed hummocks susceptible to wind erosion and the leveling and development of this land presents well-recognized environmental risks. The protection strategy to be adopted during the initial years of the project will be to level alternating strips of land, leaving the intervening unleveled strips to act as wind-breaks, and to avoid undertaking land leveling activities during the windy spring months. The land development process also involves extensive planting of trees in shelterbelts that will provide protection against wind erosion in the longer term. The agricultural development adjacent to the project area has overcome such environmental risks. The environmental issues are analyzed in detail in Annex 10.

Land Acquisition and Project Affected People

In Inner Mongolia and Gansu, very limited amounts of land may be acquired for the purpose of road construction and establishment of rural enterprises. Land Acquisition and Compensation Plans, covering an assessment of the surveyed landLs,compensation procedures, amounts and methods of compensation, and management, supervision and grievance mechanisms have been adopted. In Qinghai, there are about 4,000 farmers and herders in the 'move-in' area that would be affected by the proposed project. Resettlement Action Plan has been prepared for Qinghai.

Land Areas to be Acquired. In Inner Mongolia, the construction of a Class III, 45 kilometer road would require 40 hectares of Gobi grassland. The road would connect one sumu (township) and 583 grazing sites and would benefit 27,000 people of which 9,400 are designated poverty households. The land is owned by three gachas and is not inhabited or used for grazing. In Gansu, road construction and rural enterprise establishment would require land acquisition from local communities. The total amount of land acquisition for roads is 85 hectares, of which 20 hectares 105 are dry farmland and the rest wasteland. Land acquisition will take place in 29 project townships in seven counties. It will affect 9,552 people in 1,931 farm households with lease agreements. No land acquisition would result in relocation of existing houses. Resettlement Action Plans have been prepared for both provinces.

In Qinghai, the proposed project would affect 201 households with 1,504 herders in Zongjia Township and 399 households with 2,496 farmers and herders in Balong Township. The primary affected group consists of 63 households with 352 herders with access to 1,333 hectares of leased, poor quality spring and winter pasture stretching along the northern boundary of the project area. Another group of 289 households with 2,411 herders uses the project south-north corridor as a passageway. The third group of 248 households with 1,237 members farms amidst a dilapidated irrigation scheme.

Compensation Principles and Rules. The project will compensate people affected by land acquisition according to the Bank's OD 4.30 Involuntary Resettlement and national and provincial rules and regulations. In Inner Mongolia, the farmland compensation rate is Yuan 100 per mu and for grassland, Yuan 60 per mu. The total compensation in Inner Mongolia is estimated to Yuan 120,000. The gachas intend to use the proceeds for grassland rehabilitation in the surrounding overgrazed pastures. In Gansu, specific land acquisition must be examined and verified by the land management authorities, prior to approval. Compensation for farmland consists of land compensation, land settlement subsidy, and crop compensation.

In Qinghai, the proposed project has developed three compensation packages in consultation with the affected farmers and herders, village leaders and minority religious representatives. For the 63 herder households leasing spring and winter pasture, compensation is made by trading in the poor quality grazing land for irrigated forage land lands in accordance with a formula based on their existing and new lands' productivity. In addition, each herder household will be allocated, free of charge, 0.04 hectares of land for housing and will enjoy the same preferential policies established for the resettlement migrants (access to settlement subsidies, water supply, electricity, schools, and health care clinics). For the 289 households using the land as public passage way to grazing lands north of the project area, three fenced-off transit corridors will be established to allow the herders to move their animals across the proposed project area to graze in the natural pasture. In addition, the herders would also acquire 0.04 hectare of land for housing and 0.31 ha for farming (cultivation- 0.22, pasture- 0.04, forestry-0.05), on the same terms and conditions as the resettlement migrants. The 248 households living in the old irrigation area would be provided improved irrigation services supplied by the project. Those farmers with less than 0.31 ha of farmland (the standard set for migrant settlers) will be receiving new land up to this standard.

Management and Monitoring of Land Acquisition and Compensation. Provincial PMOs take the overall responsibility for land acquisition and assign expert groups to carry out external monitoring for land acquisition. County PMOs would plan and manage land acquisition and- ensure that procedures are followed. County level project monitoring plans include monitoring for land acquisition. Land management authorities and the Grassland Monitoring Station inInner Mongolia would supervise and monitor the process of land acquisition. The Land Management Bureaus approve land acquisition and change of land uses. 106

Ensuring Equitable Distribution of Benefits

The SAs identified some concerns that the benefits of the proposed activities may be inequitably distributed among ethnic nationalities, women and project affected households, including a concern, that non-poor households would benefit. Transparent selection rules have been designed and implementation of these rules will be closely supervised.

Participation of Women. The SAs paid special attention to integrate women's concerns into project design with specific project components benefiting women. The current situation of women in the PPs was analyzed in detail. Extensive interviews and group sessions were organized to listen to women and their suggestions in terms of the project design. This process was found especially useful to ensure the participation of those ethnic minority women, such as Tibetan and Mongol herders, who are not used to ge1ting involved in public activities. The main issues raised by women included: i) lack of employment opportunities for cash income; ii) lack of working capital for small scale household based planting, breeding and processing activities; iii) lack of access to adequate women's health care; and iv) restricted entry of girls into schools. The proposed project has been designed to accommodate these concerns. Specifically designed programs and activities include: i) women's' education (literacy and technical skills); ii) improved health care (provision of midwives and training in women's health issues); iii) emphasis on attracting girls to schools; iv) provision of special loans for women for cash- generating activities (sewing, noodle processing, embroidery, knitting, etc.).

Protection of Minority Rights

The SAs analyzed the characteristics of the major ethnic groups in the proposed project areas and established operational strategies in the project design for the minorities' development along with project implementation. Although Han people are the majority in most of the project areas, accounting for 86 percent of the total in the Inner AMongoliaproject area and 94 percent in the Gansu project area, it was considered important to, integrate minority nationality concerns in project design and implementation in order to ensure that the minority nationalities receive culturally compatible economic and social benefits from the development intervention. InInner Mongolia, minority nationalities account for about 14 percent of the total population in the project areas, (about 90 percent of them Mongols). The SA analyzed in detail the Mongol social organization, economic foundation, production patterns, household livelihoods, education, customs and religious beliefs. Consultations with herders and farmers suggested that household development models based on mixed farming, and animal husbandry and grassland rehabilitation particularly designed for minority herders should be piloted. In Gansu, the percentage of minority nationalities in the project areas is comparatively lower (5.5 percent), predominantly Hui nationality in Zhangjiachuan County. Nonetheless, the county PMO in Gansu selected six typical villages from three townships in the county to make detailed village investigation and villager consultation, so as to design appropriate sub-projects to be undertaken compatibly within the Hui socio-cultural context. In Qinghai, the SA conducted one village case study in each of the six 'move-out' counties to develop model settlement villages. This was done in close cooperation with community leaders and religious elders. 107

F. Specific Issues Related to Qinghai

Qinghai is the only PP where voluntary settlement is proposed. This has focused scrutiny on the project component. The SA for Qinghai raises the possibility for conflicts between farmers and herders, and ethnic conflicts in the 'move-in' area as a potential risk. Public disorder was also raised as a concern based on previous experience with settlers. Additional concerns are: i) dilution of the Tibetan presence in the 'move-in' area as part of a historical strategy of ethnic supplanting of minority nationalities; ii) perceived weakening of Tibetan culture as seen in the claimed replacement of the herding lifestyle with that of farming; iii) claimed difficulty of ascertaining the "real" desires of the Tibetan herders in the project 'move-in' area regarding these changes; and iv) the possibility of the use of prison labor in project construction. Careful review of each of these issues reveals that most of them were thoroughly discussed, researched, and provided for in project design in general and the Local Beneficiary Participation Plan, the Voluntary Settlement Implementation Plan, and the Resettlement Action Plan specifically.

Special Concerns

Dilution of Tibetan Presence in the 'Move-in' Area. Historically, and particularly in the 20kh century, there has been significant demographic shifts of the Han population southward and westward. These past twenty years, moreover, have seen long-distance migration of Chinese from all nationalities to the coast and other areas of the country in pursuit of better economic lives. In the Tibet Autonomous Region, and in other minority nationality areas, there has been an influx of migrants from other parts of the country such as Sichuan Province. This has sparked discussion of the long-term effects on the ethnic or nationality character of these areas. However, the proposed project entails no such long distance migratory pattern directly or indirectly. All 57,775 settlers are migrating within the same province (about 450 km on average).

The area the migrants will leave behind, the 'move-out' area, although populated now by mostly Han and Hui, holds special significance for Tibetans both home and abroad. The current Dalai Lama and the former Panchen Lama had their ancestral homes in two of the 'move-out' counties (Ping'an and Xunhua, respectively). Thus it is perhaps not surprising that although Tibetans in these counties now comprise close to 9.4 percent of the local population, they are applying to emigrate at a significantly lower rate than other nationalities, possibly reflecting the spiritual hold the home counties have on them. To partially compensate for the reduced applicant pool, Tibetan applications are approved at a rate higher than that of other nationalities (40 percent versus 34 percent), but this still leaves the Tibetan out-migration somewhat lower than their share of the local population. Such a migratory pattern will leave these six 'move-out' counties slightly more Tibetan at 9.5 percent (see Table 9-1). Also, various investments and interventions (soil conservation, cultural and educational programs, and culturally appropriate health care access) will be made in the 'move-out' counties in order to improve the social and environmental conditions for the remaining population. 108

Table 9-1: Populationby Nationalityin 'Move-Out'Area

6 Movi g-out Counties (Before Settlement) JNEmber 1 1,333,4841 649,6791124,842 4,969 398,6331 75,101| 78,5791 1,681 %of Total l 48.7%1 9.4% 0.4%1 29.9%1 5.6%J 5.9%1 0.1% Selected Settlers _Number 1 57,7751 24,4571 3,466i 20,568 3,853 5,431 _ % of Total I 42.3%1 6.0% 35.6% 6.7% 9.4% 6 Moving-out Counties (After Settlement) Number 1 1,275,7091 625,222 121,376 4,969 378,065 71,248 73,148 1,681 1%of Total - 49.0%1 9.5%1 0.4%1 29.6%1 5.6%1 5.7%1 0.1%

The immediate 'move-in' area, currently has no Tibetans at all. The project as currently planned will thus introduce a Tibetan presence into the area as they will make up nearly 6 percent of the population. The two 'move-in' townships of Zongjia and Balong which encompass the immediate project area, yield a pre-project Mongol majority (about 55 percent) for the area's population of 5,736. Among these are 276 Tibetan herders who bring their animals to pasture in the Kunlun Mountains 120 kilometers south of the project area during the summer and autumn, while wintering in an area 60 kilometers south of the project area. Given such distances, their pastoral lifestyle is not anticipated to be substantially affected by the proposed project. After the settlement is completed, Han and Hui would be the predominant nationalities in the project townships, followed by smaller numbers of the Tu, Salar, Tibetans, and Mongols. At the county level, the settlement impact will be much reduced. The percentage of Han will fall from about 53 percent to 48 percent, that of the Tibetans from about 23 percent to 14 percent and that of the Mongols from about 14 percent to about 7 percent, while the proportional shares of the other minorities, especially the Hui (from about 7 percent to 22 percent), will increase (see Table 9-2).

Table 9-2: Population by Nationkalityin 'Move-In' Area

-POPULATION BEFORE SETTLEMENT Qin ghai Provirtce _ Numnber 4,956,000 2,836,011 1,034,300 83,000 719,600850,100 6,200 186,800 % of Total 657.8% 20.90% 1.7% 14.52% 1.7 3.8% 0.2% Pro ject Specific Area_ Number 4,000 1,052 1152 2,763 ,800 75 °%ofTotal 26.3% _ 69.1% 4.62% 1 Pro ject Townships _ ,956,736 2,8611,000 T 06 83,137 719 27 _% of Total 36.8% 4.8% 5497% 3.2.% . 0.5% Dulan County _ Number 52,669 27,977 11,952 7,401 3,80 4 787 511 241 % of Total 47.% 22.7% 14.1% 7.22°1 4.2% 5.4 052% POPULATION AFI`ER SETTLEMENT Qbinghai Province II I I I INumber 4,956,00t 2,_:8_36,000|1,034,300| 83,000| 719,600| 86,200| 186,800| 10,100 |% of Total 57.10io] 2CI.9%r 1.7%1 14.5%| 1.7%l 3.8%l 0.20% Project Specific Area_ EE l INumber 61,775| 25,5091 3,4661 2,763| 20,7531 3,8531 5,431 - 1%of Total 41.3u/%1_ 6;.6% 4.5%1 33.6%1 6.2%| 8.8%| Project Townships __

INumber | 63,511r 26,56 3,7421 3,1371 20,7531 3,853 5,431r 27 |%of Total ll 41.8% .9% 4.9%1 32.7%1 6.1%| 8.6%|o 0.0% Dulan County ll l l l {Number | 110,444| 52,434t 15,418| 7,401| 24,368| 4,640| 5,9421 2411 1 %ofTotal | 47.5% 14.0% 6.7%1 22.1%| 4.2%1 5.4%1 0.2%1 109

The question of "dilution" or enhancement of ethnic/nationality presence can be answered in a number of ways, depending on the scope of the territorial unit considered. Overall, it appears that the project will not significantly impact China's spatial distribution of nationalities. The numbers involved are small; totaling the 57,775 migrants plus the 4,000 people in the immediate project area yields a number which comprises only about 1.2 percent of Qinghai's population of about 5 million.

Significance of Project Areas as of Central Concerns to Tibetans. Of all the Tibetan autonomous districts or Tibetan-populated districts in China, the three project-affected prefectures (Haixi Mongol and Tibetan Autonomous Prefecture, Haidong Administrative District, and Xining Municipality District) are the least Tibetan. All have small Tibetan population (11 percent, 10 percent, and 4 percent, respectively). In demographic terms, the project-affected districts are all relatively marginal to the Tibetan presence in China. Geographically, the three districts are in the northernmost fringe, in terms of distance from the heartland of Tibet.

Administratively, the three project-affected districts are the only districts in Qinghai province not designated solely Tibetan Autonomous districts; all the other five are Tibetan Autonomous Prefectures. Furthermore, according to the Alliance for Research in Tibet, they are the only Tibetan-related areas in all of China not recognized as a Tibetan Autonomous region, prefecture, or county. Historically, the three project-affected districts have been beyond Tibetan control since American colonial times.

Loss of Tibetan Pastoralism and Credibility of Reports of Tibetans Welcoming the Project. The question has been raised as to whether or not the project endangers the Tibetan herding subsistence life style. Actually, there are no Tibetans at all in the immediate project area and only 276 Tibetans who herd in the general project area. Most herding people in the project area are Mongols, while most of the Han and Hui in the area are farmers or participate in both herding and farming activities.

It should be noted that pastoralism as a way of life is already changing; few households are now completely dependent on herding. A change towards a semi-herding, semi-farming life style has already begun (as it also has in Inner Mongolia) with some herders building permanent brick and wood or clay and wood homes. Only Gouli Township in the 'move-in' area is a completely herding area; all the other five townships are characterized by a mixture of farming and herding. Special care will be given to ensuring that pastoralists will be able to maintain their preferred methods of livelihood. For those 352 Mongol herders whose grazing lands are directly impacted by the project, an exchange of "land for land" will replace overgrazed, degraded pastureland with newly re-seeded irrigated land. Allocation of land will be undertaken in such a manner as to enable pastoralists to continue to use this land for herding purposes. For those 2,411 herders whose traditional grazing areas are affected by the project settlement, three separate grazing corridors will be set up to enable their herds to pass through the settlement area.

The SA found that the herders, welcome the project, with about 10 percent expressing their reservations. As most data collection was conducted through interview questionnaires and focus groups (with researchers helping the mostly illiterate herders to fill in the forms), complete anonymity of responses was not always possible, thus raising the question of adequate respondent confidentiality. Nevertheless, there is reason to believe that the reported welcoming of the project by the majority of herdsmen is an accurate reflection of their desires and interests mainly because the herdsmen in the SA report their concern that due to population pressure over the past few years, the pastureland in the area has been severely degraded and that the area can no longer support them as it once did. The herdsmen have expressed the need to identify new economic options, including agricultural ones to bolster their livelihood. Half of the surveyed herders requested land allocation so as to pursue this possibility. The project's irrigation work is likewise 110 welcomed for providing a reliable water source not only for the herds but also for agriculture as they pursue their semi-herding, semi-farming lifestyle. The views of Tibetans in the six eastern 'move-out' counties were also solicited and there is no reason to doubt the sincerity of those who volunteered to 'move-out'. All migrants also have the option of returning to their original homes during the first two years, without financial or any other penalty, if they so decide.

Possibility of Use of Prison Labor in Project Construction. There are a number of prison farms in Qinghai. The closest one to the project area is the X(iangride State Prison Farm, located about 48 kilometers north of the area. This irrigated state farm is drawing water from the Xiangride River, which is also the primary source of irrigation water for the settlement area. The Bank Team has concluded that the proposed project would not benefit any prison farm directly nor will any prison farms or affiliated organizations participate in any activities related to the proposed project. The State Council's Leading Group for Poverty Reduction and Development and the Provincial Government of Qinghai have also confirmed this in writing.

Additional Actions to Reduce Possible Conflicts

The SA conducted in the 'move-in' area determined that the majority of the surveyed existing population accepts the migrants (29 percent of the herders said that the migrants were "welcome"; 62 percent did "not object" to the migrants). Nine percent of the surveyed existing population in the 'move-in' area expressed their unwillingness to welcome the migrants mainly due to concerns about public disorder and fears of grassland overgrazing. Previous migrations of settlers into the area during the 1980s led to cattle thefts, so both local cadre and herders expressed concern about a reoccurrence of such incidents and asked that better public safety measures be undertaken. The SAs address the concerns of potential conflicts with the influx of migrants. Another issue identified in the SA is potential conflicts over water. Many herders expressed their fear that they would be excluded from adequate water supply or from access to the new irrigation works. They were anxious to be reassured that the demands for water from the new settlements would not put them at a disadvantage. Cadres also feared that so many poor people from the mountainous areas would want to come to the new area that they would not wait for approval and would come illegally, thus overwhelming the local area and its infrastructure.

Social Stability. A number of possible mitigation strategies were identified in the SA, including having villagers from the 'move-out' area move as a group to the new village site, thus maintaining both formal and informal social control mechanisms. At the herders request, local police sub-stations and people's courts will be set up in the settlement area to ensure overall public safety and training of new and existing law enforcement officers will be conducted. To lessen the potential of inter-ethnic conflicts, the two Muslim groups, the Hui and the Salar, would be settled in adjoining areas, while the culturally similar Tu and Mongols would be located in similarly contiguous settlements. In addition, the Local Beneficiary Participation Action Plan promotes active participation in project implementation by the 'move-in' and 'move-out' peopL- and a series of suggestions made by the local herders were incorporated into the Plan. References are made to include community and religious elders in the leadership groups to help secure community-wide buy-in and ensure that the project implementation is compatible with the cultural preferences of the nationalities concerned. The project design included suggestions made by local authorities including: i) clear demarcation of the boundary between the proposed project area and neighboring grazing areas; ii) adoption of measures to protect adjacent grazing lands and desert vegetation; iii) provision of adequate support to the migrants when they arrive; and iv) ensuring that the irrigation development is in place prior to migration.

Pilot Program. A pilot program on 300 hectares for 200 households would be undertaken in advance of full-scale project implementation to test the proposed interventions at the field level with the participation of project beneficiaries. This would create a demonstration area which Ill could be later used as a training site for project participants and provide an opportunity to show typical types of irrigation layout, village infrastructure, schools, clinics, agricultural (planting of green manure) and land management practices which would be promoted by the project. The pilot program would also include participatory evaluation techniques to provide input from a representative range of parties in the project area. The pilot would also ensure that the minority needs are fully met. The pilot program would provide an opportunity to develop "lessons learned" which could be applied to the design and implementation of the full-scale phase of activities which would start after full evaluation of the pilot program by independent reviewers. It would be carried out in such a way that it does not interfere with current land use patterns of the herders or farmers living in the area, nor will it result in loss of their shelter, productive assets, or access to pastures. Irrigation will be provided through use of ground water and diesel engine driven pumps, so as to avoid the necessity of developing major irrigation works during the pilot phase. Expenditures for this activity have been approved for retroactive financing under the project. The PMO would be responsible for the implementation and supervision of the pilot program.

Monitoring of the Pilot Program. The pilot program is designed to avoid conflicts and enable adjustments in the project design before the bulk of the migrants arrive. The pilot settlement will be closely monitored to measure the effectiveness and adoption rates of the various program components and also to identify emerging issues, if any, in the relationship between migrants and receiving area population. Full scale settlement would only begin after the pilot program is completed and evaluated. In addition to the pilot itself, a 5 percent population sample tracking procedure has been developed for semi-annual progress reports to the PMOs and the Provincial Leading Group.

Indicators to determine the success of the pilot program are both economic and social. Economic Indicators include: (i) adaptation rates are high; (ii) incomes and yields increase; (iii) both farmers and pastoralists continue to earn their livelihoods without hindering the other's effort; and (iv) villages receive adequate water, electricity, and other infra-structural supports. Social Indicators would include: (i) the project is carried out in a socioculturally appropriate manner; (ii) beneficiaries' satisfaction rates are high; (iii) there is not only an absence of conflict but mutual inter-ethnic respect in religious and other spheres; (iv) villages have access to adequate and culturally appropriate social services (schools, health care, credit, cultural enrichment); (v) project beneficiaries from the move-out areas report the selection process ethnically equitable and efficient; and (vi) local governments and cadre from both the move-out and move-in areas collaborate to smooth the transition for host area population, migrants, and future migrants. The pilot program monitoring process will result in a report

G. Institutional Capacity

Project Provinces Capacity for Mitigation of Social Impacts and Protection of Minorit Rights. The PPs will closely follow the implementation of the SA suggestions and monitor impact and effectiveness of actions. Especially the pilot program will be closely monitored and thoroughly evaluated by independent reviewers before full-scale resettling will commence.

The People's Republic of China officially describes itself as a multi-nationality country. Its constitution delegates protected rights and powers of self-government to its 56 officially recognized nationalities. Nationwide, the non-Han 8 percent of the population listed as minorities enjoys special protections and status, including various degrees of administrative self-rule in a network of autonomous regions, prefectures and counties. Qinghai encompasses such minority jurisdictions in both the 'move-in' and 'move-out' areas (e.g. Haixi Mongolian and Tibetan Autonomous Prefecture; Hualong Hui Autonomous County). On the ground, autonomous status has meant greater legal protection for minority culture and languages and greater representation of minorities in local cadre/official posts. Such affirmative action policies apply to a wide range 112 of social policies, not least of which is exemption from stringent Chinese family planning policies.

In Qinghai minority jurisdictions, schools have been established to preserve and promote minority cultures. The proposed project design includes the establishment of 11 language minority schools (out of 21 planned primary or K-9 schools) in the 'move-in' area and a number of education programs in the 'move-out' area. In addition, culturally appropriate health care facilities are being established both, in the 'move-in' and 'move-out' countries. Furthermore, the Local Beneficiaries Action Plan has a list of explicit preferential suggestions for women and minorities, including special credit lines, educational preferences, special training opportunities, and provisions for continuation of religious and cultural customs.

On the ground, although minority policies have varied significantly over the decades of 20' century Chinese history, during past two reform decades, autonomous status has meant greater legal protection for minority culture and languages and greater representation of minorities in local cadre/official posts. Such affirmative action policies apply to a wide range of social policies, not least of which is exemption from stringent Chinese family planning policies. Overall, there is clear and strong evidence that the Chinese commitment to poverty reduction is serious and long lasting.

H. Grievance Mechanism

Consultation and Grievance Mechanism. Project beneficiaries and project affected people in all PPs will have access to a consultation and grievance mechanism to address their concerns in regard to the proposed project. The mechanism and procedures have been described to the project beneficiaries and stakeholders. It is believed that this wvillbe particularly useful in the case of land acquisition if farmers feel unsatisfied with compensation implementation or if conflicts arise between farmers and herders. Complaints and grievances can also be raised and reported to the higher level PMOs, the land management authorities, and the extemal, independent monitoring groups. Beneficiaries and stakeholders also have riglhts to raise any questions directly to local government and provincial authorities. The right to make lawsuits has been made explicit to the farmers and herders. All complaints, both oral and written, will be recorded including resolution and the time of resolution. External monitoring reports will include a summary of grievances received, if any, and the manner in which they were resolved. 113

Annex 10: EnvironmentalAspects

CHINA:Western Poverty Reduction Project

A. Overview

Introduction. This annex summarizes the main relevant findings and recommendations of three province specific environmental impact assessments (EIAs) prepared for the proposed project as well as results of additional environmental and technical evaluations undertaken as part of project feasibility assessment, and outlines the World Bank's evaluations of the main, substantive environmental issues.

Strategic Context. The proposed project supports several of the World Bank's key Country Assistance Strategy objectives including: targeted poverty alleviation investments, support for primary education with a particular focus on girls and minority nationalities, support for environmentally sustainable agricultural and livestock development and enhancing access to basic services for poor people. Access to health, education and other rural services would also be improved for the poor households. Although the proposed project will be implemented in three PPs, this report focuses mainly on proposed developments in Qinghai Province, which will involve new land reclamation and development in areas not presently utilized for crop production and, as such, offer the greatest potential for environmental impact. Developments in the other two PPs are substantially confined to modifications and improvements to existing land uses and mainly represent intensification rather than major changes in land use.

B. Environmental Review Process

Applicable Environmental Procedures. Consistent with the "Environmental Protection Law" of the People's Republic of China and Operational Directive (OD) 4.01, "Environmental Assessment" of the World Bank, each of the three components has been the subject of an Environmental Impact Assessment (EIA). The EIAs were prepared by the Environmental Protection Research Institute of Gansu Province, the Environmental Sciences Institute of Inner Mongolia Autonomous Region, and the Qinghai Institute of Environmental Science. All three institutes hold the "Environmental Impact Assessment Certification of Class A," issued by the State Environmental Protection Agency and the scope of the EIAs was substantially in compliance with that required for the World Bank's environmental Category A, including public consultation. However, the bulk of the environmental analysis associated with the proposed project was of an essentially technical nature and designed to ensure that all relevant environmental considerations were taken into account in project design. The EIA documents identify the key environmental risks, and mitigation and environmental monitoring measures with their associated costs, and specify implementation responsibilities. English language translations of these reports have been placed in the Public Information Center at the World Bank.

Environmental Category. The proposed project was reviewed and placed in World Bank environmental screening category "B" in accordance with the provisions of Operational Directive 4.01 and taking account of past practice with classification of a large number of similar integrated agricultural development projects financed by the World Bank in China over the last 10 years. Review of the scope and proposed activities of the project by Chinese and World Bank environmental staff indicated that the potential impacts were not unusual or unprecedented in terms of other large scale rural developments being undertaken in China, whether financed solely by the Government or co-financed by the World Bank and other donors. These activities generally have limited risk of adverse effects that could be addressed through design, mitigation 114 and/or monitoring measures5. The land reclamation and improvement activities supported under the project are of a kind that have been widely and successfully applied in other arid and semi- arid development projects in both the Project Provinces/Regions and other Provinces/Regions in central and western China. Previous interventions of this type by the World Bank have included the Tarim Basin I and II Projects. In addition, the proposed activities would be implemented in a large number of dispersed sites with most of the anticipated negative environmental impacts being of local nature. The mitigation measures and rnethods to be utilized are well known and established in China. It should be noted that during project preparation the overall scale of the proposed project, particularly the scale of developments in Qinghai Province, was substantially reduced with an incremental decrease in potential environmental impacts.

Preparation and Review Process. Project preparation has included the participation of a qualified, experienced international environmental consultant as an integral member of the interdisciplinary project team used by the World Bank. This specialist participated in the preparation, pre-appraisal, and appraisal of the project. The work of this individual was complemented by World Bank environmental staff who reviewed the environmental documentation prepared for the project. The environmental design of the project represents the combined work of these different experts. The EIA documents were submitted to Chinese authorities and the World Bank prior to appraisal and have been reviewed and approved by the State Environmental Protection Agency of China. They have been made available in the PPs to government agencies and project affected persons upon request. The EIA work was carried out in close coordination with the Social Assessments and the preparation of the Voluntary Resettlement Implementation Plan (VSIP) and the Involuntary Resettlement Action Plan (RAP) for the Qinghai sub-project.

Consultation Process. Public consultation activities were also undertaken as part preparation of all three EIAs. In Qinghai, these consultations were undertaken in both the 'move out' and 'move in' areas in conjunction with the social surveys carried out as background to the settlement program, while in Gansu and Inner Mongolia, specific surveys were carried out to gauge project awareness and opinions on various aspects including environmental issues. The surveys were quite large; 1,800 questionnaires were distributed in Qinghai 'move out' areas and 120 in the 'move in' area, 5,000 questionnaires were distributed in Gansu supplemented by face to face interviews, and 4,650 questionnaires were distributed in Inner Mongolia. In Qinghai, environmental issues were only mentioned by respondents in the 'move in' area and the main issue of note was universal concern (i.e. expressed by 100% of respondents) that the new migrants would cut vegetation and destroy the ecological environment. In Gansu, there was a high level of awareness of the project and about 80%1/0of respondents considered that the project would have beneficial or neutral environmental effects. In Inner Mongolia, there was a high level of awareness of the proposed project and the great majority (93%) of respondents felt that the environmental effects of the project would be neutral or beneficial.

Approach to the Analysis. The environmental analyses and related project documents have been reviewed from two main points of view: i) to determine whether there were any particular environmental aspects within the proposed project areas which would justify modifications or adaptations to the proposed development plans to enhance environmental sustainability; and ii) to define the environmental management and monitoring procedures which should be implemented as part of the implementation plan. The main environmental risks associated with the project relate to the fact that the poverty alleviation benefits arise from land use intensification in arid and semi-arid agro-ecological zones that are intrinsically sensitive from an environmental point of view. It should be noted that no pesticides are planned to be procured with the proceeds from this

5 TheProject includes the construction of the KeriDam. This dam is a regulatingreservoir located in extremelysparsely vegetated and unutilized desert. As such,the environmental implications of its constructionwere determined to beinsufficient to justify assigning the Project to a CategoryA. 115

loan/credit. Although the project is primarily a poverty alleviation project, it also provides some opportunities to produce environmental benefits through reduced land use pressures particularly in Qinghai and Gansu and promotion of more environmentally sustainable agricultural practices in Gansu.

Potential Environmental Risks. More particularly, the main environmental risks include: i) possible acceleration of desertification; ii) possible agro-chemical pollution due to excessive application of fertilizers and pesticides; iii) solid waste generation from small rural enterprises and plastic film used in cropping; and iv) possible impact on wild vegetation and animals due to direct displacement or indirect effects of proposed land development activities. A review of these potential impacts and proposed design, mitigation and monitoring measures are described below. It should be noted that Chinese government agencies and research institutions have established a substantial track record of environmentally sustainable development of similar areas in other Provinces/Regions under very similar biophysical and geographical circumstances. The World Bank is investing in similar, although much larger scale, land development programs in very similar enviroDments in Gansu Province (Gansu-Hexi Corridor Project) and in Xinjiang Uighur Autonomous Region (Tarim Basin I and II Projects) and project preparation and implementation experience in these projects has provided a substantial basis upon which to evaluate the feasibility of the proposed developments.

C. Project Setting, Development Alternatives and Proposed Interventions

Project Setting. The target population in the three PPs live in fragile agricultural and pastoral lands. In Qinghai, crop yields are low due to poor soil quality, low rainfall, recurrent drought, and undeveloped farming techniques. Increasing cultivation and grazing pressures in mountainous areas are having devastating ecological impacts and there appear to be limited and possibly even no environmentally sustainable development options available in those areas. In Gansu and Inner Mongolia crop yields are low largely due to a comparatively harsh agro-ecological environment which is characterized by poor soils, steep topography, recurrent drought, short growing period, and significant grassland degradation combined with the lack of access of the targeted poor population to productivity enhancing technologies and capital for planting higher value crops or forages. As in Qinghai, this situation is already resulting in severe environmental degradation and, without intervention; this trend will continue and worsen. However, and in contrast to Qinghai, there appear to be some options in both Gansu and Inner Mongolia for introducing productivity enhancing and environmentally sustainable crop and livestock development packages essentially in situ. There are proven technologies widely and successfully adopted in the proposed project areas and their further application is basically limited by shortages of investment funds of the type that the project will provide. The project will not solve all the problems in the three PPs, but it will make a substantial contribution to alleviating them.

Development Alternatives. All of the participating PPs and Counties suffer from severe poverty: which is leading to serious environmental problems. In this context, the "no development" option is not feasible within the government's poverty alleviation or environmental objectives. Within that context, development options in the participating PPs are severely constrained by environmental and technical factors. In Qinghai, the resource base of the move out Counties has been severely depleted due to intense population pressure and the associated unsustainable cultivation and over grazing that has been necessary for local people to survive. Neither the Government nor the World Bank has been able to identify agricultural technologies that can provide a solution given current population pressures and the only feasible altemative is to provide opportunities for at least some of the population to voluntarily move to other areas which offer the promise of a more sustainable future. 116

The benefits of this strategy can be extended further by rehabilitating the worst of the vacated land and treating (e.g. through terracing) the remainder for redistribution amongst those families remaining so that they too will have a better chance of improving their situation and the condition of the local environment (as will happen through an existing government program). The situation in the two other PPs is not quite as severe insofar as there are known strategies for improving the sustainability of rural development "in situ". In these cases, development options are constrained by the need to reduce risk. Preference under the project has been given to supporting known technologies, which have already been shown to be effective and acceptable to farmers. The menu of possibilities is very limited. To the extent that other development options are being pursued through the project (e.g. sustainable rangeland management in Inner Mongolia), this is being done through demonstration/pilot programs which will provide a better opportunity to evaluate the effectiveness of these approaches and their acceptability.

Given the general development framework, which arose from consideration of the above constraints, further work was done on consideration of detailed design options. From an environmental point of view, probably the most significant of this work related to water management and overall development options for the Xiangride-Balong irrigation development in Qinghai. A wide range of conjunctive water use options are available and mathematical modeling was used to evaluate five of these and select an approach which provided an adequate balance between engineering, environmental and other factors. Monitoring will be undertaken during project implementation to determine the need, if any, for revision of the proposed development strategy.

Proposed Intervention Strategy. An important aim of the proposed strategy is to progressively broaden the focus of the farming systems from exclusively food self-sufficiency, as it is at present, to one which continues to stress food security but also offers increased opportunities for generating cash incomes. As productivity and incomes increase, opportunities should be provided to progressively retire from cultivation steeper land and other areas subject to severe wind erosion, which are not suitable for sustainable crop production. It is intended that these lands would be sown to forage legumes and grasses, or protective trees. The consequent improvement in soil water conservation and progressive reduction of soil erosion (wind, sheet, gully) would, over time, also contribute towards raising sustainable crop and forage production. A priority objective of the development packages is to improve the nutrition and productivity of livestock which are essential, long-standing traditional elements of local farming systems in all three areas and cannot be ignored in any integrated development approach. Livestock development would encourage the cut-and-carry feeding of penned ruminant (cattle, sheep, goats) and improved non- ruminant (pigs, poultry, rabbits) integrated crop/livestock production systems, thereby progressively reducing or eliminating large and small ruminant grazing on natural pastures, which is one of the major sources of the land degradation problems being experienced throughout the PPs. D. Proposed Project Supported Activities

Agro-ecological and social conditions in each of the three PPs are quite different and, hence, the details of the proposed investments intended to achieve the above strategic objectives differ substantially between Provinces although the physical interventions fall within five general classes as follows: (i) Land and Household Development (all PP). The investment activities include intensification of crop production through the use of proven improved and environmentally-friendly technologies to raise crop yields, introduction ;and/or expansion of cash crop production (vegetables, tree crops, fruits), conversion of livestock management from free-range to confined- feeding production systems for sheep, goats, cattle, pigs, rabbits and poultry, improvement in the quality of farm level technical services through skills upgrading of county, township and village 117 technicians, and support for adaptive research into and on-farm demonstration of improved and environmentally sound farming systems; (ii) Irrigation and Land Improvement (all PP). In Qinghai, construction of a new regulating reservoir and renovation of an existing reservoir plus construction of irrigation and drainage systems to service approximately 19,000 hectares of land. In Gansu and Inner Mongolia, construction of dispersed, small scale irrigation works covering a total of 34,050 hectares plus terracing of 26,500 hectares of sloping land and installation of soil and water conservation works; (iii) Rural Infrastructure (all PP). Construction of 216 kilometers of Class III and IV rural roads (i.e. minor access roads) to improve access to 289 remote administrative villages; construction of drinking water supply facilities to serve 370,000 people in 760 administrative villages; and extension of electric power lines to 410 administrative villages; (iv) Rural Enterprise Development (all PP). Provision of credit to underwrite establishment of small scale, non-state owned, household-based rural enterprises for construction of food and fiber processing facilities based on appraisal criteria; and (v) Voluntary Settlement (Qinghai only). Voluntary settlement in a new agricultural development area for an estimated 57,800 absolute poor currently living in degraded mountainous areas of eastern Qinghai and 4,000 local herders and farmers living around the settlement area.

E. Common Indirect Environmental Impacts

There are several indirect potential environmental impacts that are common to the various project activities in all three PPs. These include: (i) Use of Plastic Mulch. Increased use of plastic mulch is becoming a problem throughout China, notwithstanding its demonstrated agricultural benefits. In one of the PPs (Gansu) it is proposed to establish a small plastic recycling plant of a type being promoted throughout the country. Whether this proves to be financially viable (something which will need to be firmly established before financing is approved) remains to be seen. Neither Qinghai nor Inner Mongolia is likely to be able to generate the quantities of plastic that would render such a plant viable and hence, no such developments have been proposed in these areas. On the other hand, informal, inter-Provincial plastic recycling has started to develop in China and this may offer an alternative. In the longer term, the problem is only likely to be dealt with on a national basis through the promulgation and enforcement of national design standards for plastic film which include, inter alia, the requirement for film to be degradable following exposure to UV light. Such film is already available in China although it costs more than conventional film and its use is not mandatory. In the interim, this form of pollution has to be viewed as an unavoidable aesthetic impact of the proposed project; (ii) Small Scale Agro-Enterprise. Some small scale agro-enterprises could generate liquid wastes containing organic material that are subject to regulation under national water pollution control laws. This is most substantially an issue with proposed developments in Gansu Province rather than in Qinghai and Inner Mongolia. In the cases that have tentatively been identified, the control technologies are well known, simple to design, construct and operate, and subject to regulatory procedures administered by Provincial and County Environmental Protection Bureaux (EPBs). Administrative steps will be included in sub-loan processing procedures in all three PPs to ensure that loans will not be issued without environmental clearances being received from relevant EPBs; and (iii) Increased Use of Agricultural Inputs. Increased on-farm productivity will be achieved, at least in part by increased use of inputs, including fertilizers and pesticides. Regardingfertilizer usage, there are parts of China, particularly in lake catchments in the highly developed coastal zones of the country, where nutrient enrichment (eutrophication) problems are being experienced and research indicates that, while most of the problem is due to emissions of nitrates in human 118 and industrial wastes, some (typically on the order of 20-25 percent of total nitrogen emissions) is due to non-point emissions from farms. Such problems are not being experienced in central and western China for the good reason that farmers in these regions tend, for financial reasons, to significantly under-utilize fertilizer. This is likely to continue to be the case in the project areas. In addition, the farming practices being promoted under the project are intended, among other things, to reduce soil loss, conserve water, and increase organic matter retention. All other things being equal, all of these strategies contribute to reduction in nutrient losses from fields. Regarding the use of pesticides, World Bank funds will not be used for procurement of such chemicals although it is likely that farmers engaging in cash crop and orchard production will procure them using their own funds. The project will rely on continuation of existing pest management programs by Provincial and County Plant Protection Departments. These include farner education, a highly organized pest forecasting system that is designed to discourage calendar spraying, restrictions on the types of chemicals which can be sold and mandatory requirements for the display of occupational health and safety warnings and recommended application levels and procedures on containers.

F. Qinghai

Proposed Activities in Qinghai

Activities in Qinghai, unlike the developments in the other two PPs, are based totally on the development of irrigated agricultural production to accommodate voluntary resettlers. The project area has an arid to semi-arid climate and crop production is not possible in the absence of irrigation (either natural or artificial). Activities will involve development of an irrigation system to service an area of about 19,000 hectares which will be located in six separate development areas spread over a total area of approximately 250,000 hectares. Thus, the development area comprises about 8 percent of the "project area" (the geographical area whose boundaries circumscribe all the individual development areas).

Water Resources Development. Irrigation infrastructure would be developed through two separate systems servicing five irrigation districts located in close proximity to each other: (i) Xiangride-Balong Irrigation Sub-District. The first system would be completely new and will comprise a diversion weir on the Xiangride River, a 28 kilometer conveyance canal leading to a 23 million cubic meter-regulating reservoi?. A 59 kilometer long canal would connect Keri Dam to the Balong Irrigation Sub-district (14,700 hectares) and part of the Qingshuihe Irrigation Sub-district (1,660 hectares). The canals will all traverse and be located within extremely sparsely vegetated desert land (vegetation cover generally less than 5 percent) and will be concrete lined to reduce seepage losses; and (ii) Hatu Irrigation Sub-District. The second system will involve rehabilitation and increasing the delivery capacity of an existing irrigation system through improvements to an existing intake- structure, raising the wall height of an existing regulating dam (Hatu Dam) by one meter, and treating the base of the reservoir to reduce leakage. This system would service a total of about 2,800 hectares in three irrigation sub-districts (Hatu - 760 hectares, Yikegaoli - 740 hectares and Xiaoxiatan - 1,130 hectares).

Activities will also include financing of installation of appropriate drainage systems and on-farm irrigation works in all the serviced areas.

6 This is an off-riverreservoir of a typeoften referredto as "turkey'snest dam."That is a reservoirwhich is surrounded on three or four sides by constructedwalls. Keri Dam will have a front wallheight/crest length of 40meters/390meters and sidewall heights/crest lengths of 21 meters/135meters and 26 meters/480meters. The reservoirwill inundate approximately127 hectaresof unuseddesert land and have an averagedepth of about 15 meters. 119

Crop Production. Crop production investments in Qinghai would include the initial land development (clearing, leveling, and cultivation) for crops (starting with green manure planting), forage and tree planting on 19,200 hectares to which water will be supplied through the irrigation component. Project funds will be provided to cover investments for the first crops to be sown by the migrants (seed, fertilizer, plastic mulch, pesticides, etc.), development of seed production and multiplication areas (679 hectares in total), development of two technical service centers, and installation of applied research and demonstration plots for farmer training. The main crops to be grown include grain (wheat and barley), potatoes, rape, beans and vegetables (onions, garlic, cabbages). Forage crops (vetch, alfalfa, clover, and oats) would be grown as an integral part of crop rotation as is the widespread practice in arid and semi-arid development projects in China7. About 15 percent of the land development area will be used for establishment of forested windbreaks, shelterbelts, fuel wood forests and production forests as is conventional practice in such developments.

Livestock Development. Forage production will be derived from the areas to be newly irrigated, with winter forage consisting of ensilaged greenfeed and byproducts, ammoniated straws and stalks which are widely used in China and supplemented, in some cases, by purchased or home- produced concentrates such as oilseed cake, wheat bran and barley. Livestock production will be based entirely on the stall-fed, cut-and-carry feeding system with no utilization of grassland areas. Production systems will include pig breeding and fattening, sheep and cashmere goat breeding and fattening and/or household chicken and egg production. The majority of households are expected to raise only a small number of animals, largely for household consumption but with one or two surplus animals sold for cash income, although a minority may choose to specialize in one type of animal production system. Small forage areas would be planted on each farm while specialized forage production farms (about 2,880 hectares in total) would be designated within the settlement area. Each household would be allocated land (about 0.256 hectares) including the forage production farms. Forage would include hardy varieties such as milk vetch, alfalfa, and wild rye grass. A condition of household-based livestock financing would be a proven self- sufficiency in forage or grain. Livestock numbers would be based on feed balance calculations. Grazing on the natural grasslands surrounding the settlement area would be prohibited. Local Grassland Monitoring Stations supervised by Project Management Offices (PMOs), would spot check the accuracy of feed balances and compliance with confined feeding.

Low-cost, adobe and plastic cover livestock sheds, ammoniation pits, capital stock purchase, forage establishment costs would be financed, and for specialized households, one third of the cost of initial feed purchases. A number of households would specialize in forage processing with small low-cost straw and chaff cutters.

Rural Roads Development. Construction of 116 kilometers of all-weather, Class IV gravel roads on a 4x4 kilometer grid would be financed, which would follow branch canals in all irrigation areas plus 426 kilometers of farm roads. These roads will be no risk of increasing access to areas- outside the project.

Rural Enterprises. This component will underwrite development of extremely small scale agro- processing enterprises including; a feed concentrate mill, a flour mill, a fodder chopping mill, a vegetable oil extraction mill and noodle manufacturing plant.

Rehabilitation of 'Move-Out' Areas. As part of existing land rehabilitation programs in the "move out" areas, approximately 85 percent of vacated land will be planted to trees and grasses and retired from cultivation. The remaining 15 percent is of lesser slope and will be terraced and

7 Farmers in the arid regions frequently plant a crop such as clover as a green manure crop simultaneously with a grain crop or, if the season is long enough, as a second crop for forage production or as green manure. 120 shared amongst families remaining in the area to increase per capita farm size and improve the possibilities for developing more environmentally sustainable farming systems.

Existing Environmental Conditions

The key environmental concern in Qinghai is the potential impact of the land use change to be promoted in the Xiangriade-Balong areas (hereafter referred to as the Qinghai-project area) of Dulan County where the resettlers are to be established. As noted above the environmental effects in the 'move-out' counties will be neutral to beneficial due to reduced pressure from agriculture and grazing on marginal lands combined with actions to rehabilitate these areas.

Location and Landforms. The Qinghai-project area is located in the southeastern corner of the Qaidam Basin, which extends diagonally southeast to northwest through the northern half of Qinghai Province. The mountains bordering the Basin to the northeast, southeast and southwest reach heights of up to 5,000 meters above sea level (a.s.l.) while the lowest elevations in the center of the basin are in the order of 2,500 meters a.s.l. The general slope of the Basin is from southeast to northwest.

The main landforms comprise the denudational mountains and hills which form the boundaries of the basin (and are also the only location of reliable precipitation), denudational and depositional gobi on the piedmont plains at the base of the mountains (the lower parts of these are the only areas where agriculture other than itinerant livestock grazing is possible) followed, in the lower reaches of the Basin, by sand dunes and salty marshes. Numerous salt lakes including some of the largest in China dot the lower parts of the Basin. The gobi piedmont areas are subject to on-going depositional action and reworking as a result of flooding from the rivers that emerge from the mountains. Closer to the foot of the mountains, they are covered with gravel and vegetation is virtually non-existent. As distance from the mountains increases the slopes flatten, the surface materials become finer and the groundwater table is shallower (3-25 meters). These areas offer agricultural development potential provided that supplementary water can be provided.

Immediately to the north of the Qinghai-project area is located an extensive saline swampy area or "soak", covering about 35 square kilometers which follows the course of the Qaidam River and exists due to the intersection of the underlying water table and the ground surface. The water is saline due to the evaporative concentration of salts dissolved in the groundwater. The ground surface in this area is braided, due to periodic flood flows in and overflows from the river and vegetation cover is variable from quite dense in the gullies to non-existent on the ridges. This type of swampy ecosystem continues intermittently along the course of the Qaidam River system for some 180 kilometers to the northwest until the r iver terminates in the Huaerxun Lake system, one of the largest systems of salt lakes in China.

Climate. The Qaidam Basin is totally cut off from maritime air masses and is thus the driest part of Qinghai and amongst the driest areas in all of China. Average annual rainfall in the northwestern parts of the Basin is in the order of 20-50 millimeters but this increases progressively towards the project area reaching 50-100 millimeters immediately west of Balong and 100-200 millimeters east of Balong. Rainfall in all areas is erratic and unreliable. Precipitation in the surrounding mountains can be as high as 700 millimeters, much of which falls as snow. Runoff and infiltration of this precipitation is essentially the sole source of water for all life within the Basin. Average temperatures are cool (around 4°C) with substantial seasonal variation (maximum temperatures can reach approximately 330C and the minima can reach 26°C). There are about 220 frost-free days per year and windspeeds, particularly during March-May, can be high. 121

Surface Water Resources. The surface water systems comprise a network of ephemeral and semi-permanent "rivers" which debouch from the mountains forming the southern rim of the Basin, flow across the sloping piedmont areas and discharge into the trough of the Basin which, as previously mentioned, flows north of the Qinghai-project area. The overall system is referred to as the Qaidam family of Rivers. Upstream of the Qinghai-project area, the main stem of the river is referred to as the Xiangride River which rises outside and to the south of the Qaidam Basin and then flows through a notch in the southern rim of the Basin about 50 kilometers southeast of the project area. The Xiangride River flows into a soak to the east of the Qinghai- project area. Drainage from this soak represents the start of the Qaidam River. The Qaidam River issues from this area and flows northwest past and to the north of the Balong Irrigation Sub- District. It is joined by numerous ephemeral tributaries also flowing out of the southern rim of the Basin, several of which (including the Xianomuhong River, the Hatu River and the Yikegao River) are already used as irrigation water sources. The Xiangride/Qaidam River system flows more or less permanently (at least in the vicinity of the Qinghai-project area) due to its upstream length (some 250 kilometers), and the fact that its headwaters flow through two large lakes with a combined capacity of about 8 billion cubic meters (Bcm).

These are three sources of water for the project area. Their identities and average yields (in million cubic meters (Mcm) per year) are: * Flow in Xiangride River - 3 87 Mcm per year * Flow in six small rivers flowing into the Balong area - 162 Mcm * Groundwater in the Balong area - 100 Mcm Thus total availability is about 650 Mcm per year.

Flora and Fauna. Qinghai in general and the Qaidam Basin in particular, contain a comparatively low diversity and abundance of both flora and fauna due, in part, to the extremity of the environment and partly to the limited number of biogeographical units which occur in the region. The project area supports two main associations of native flora:

(i) Xerophytes and super-xerophytes of the piedmont and gobi areas, predominantly Haloxylon ammodendron, Ephedra przewalskii, Ceratoides latens, Reauria soongorca, Calligonum mongolicum, Artemisia desertorum and Populus przewalskii. All are widespread throughout similar areas in the region; and

(ii) Halophytes (salt tolerant species) which occur in the saline swampy areas to the north of the project area. Main species include Nitraria spp., Tamarixchinensis, Phragmites communis, Achnatherum spp.

There are no conservation reserves in or near the area nor was a need for such reserves identified in the Biodiversity Review of China (WWF, 1996). In addition to the native species, a wide- variety of introduced tree species and agricultural crop species are found in existing oases scattered around the area.

Wildlife resources occurring in and adjacent to the Qinghai-project area are moderately diverse, particularly the bird species, although not abundant due to the exceptionally low productivity of the natural land systems. The main mammals identified in the area includeProcapra gutturosa, Lepus oiostolus, and Vulpes corsac. Birds are. more diverse, with some eight species being identified in the area. Again, the habitat types represented in the area are widely distributed in central western and central northern Qinghai. 122

Environmental Management Issues

There are two categories of environmental management issues associated with the proposed development. First, the environment poses certain risks to the success of the proposed development. Second, the development itself poses certain risks to the local environment. The first category of risks arises from the extremely harsh nature of the environment in which the development is to be undertaken. The main issues are:

(i) Wind Erosion. Soils in most of the land to be developed are light textured and have been wind-blown into hummocks which will require leveling for irrigation development, raising the risk of increased wind erosion. This is a common problem and the standard safeguards are to level the land in alternating strips and to cease land development works during the windy months. These safeguards provide effective short-term control. Longer term protection is provided through the extensive shelter belts of trees which are planted as soon as irrigation water becomes available (and before crop production commences) and are an invariable characteristic of oasis developments throughout China; (ii) Soil Salinity. Soil investigations in the Balong irrigation area indicate the presence of saline/sodic soils in some areas which, if incorrectly managed, can threaten crop yields due to toxicity and/or impeded drainage. The safeguard required is to install necessary drainage, to ensure that land leveling is undertaken accurately, to over-irrigate during the first years of development to leach salts out of the surface layers and to closely manage subsequent irrigation to ensure that the salt front is kept well below the root zone and/or conjunctively use ground and surface water to control water table depth (as is proposed in parts of both the Balong and the Qingshuihe irrigation districts). Again, these measures are well known by irrigation and extension services in China and the project includes provision for further soil surveys as part of final detailed design of drainage systems and for training and demonstration of correct irrigation techniques for the new farmers. Nevertheless, this would have to be rated as a potentially significant risk, particularly in view of the fact that the majority of new migrants will not have had significant prior irrigation experience. Continuous monitoring of crop yields and soil and groundwater salinity levels is an essential safeguard for assessing the situation and identifying the need to take action through supplementary training of farmers and information dissemination. The risk of soil salinization and sodification is a related but well known risk that will also require constant monitoring; and (iii) Water Supplies and Flood Risks. The irrigaition system will be totally dependent on the availability of water to meet crop demands. The total gross annual irrigation demand is estimated to be about 300 Mcm, which is less than half the resources available in a 75 % rain year. Water balance and reservoir operation studies incorporating groundwater simulations were carried out to establish the adequacy of supplies from the Xiiangride River and the Balong groundwater. Operational studies carried out for the optimum development option (Keri Dam plus 67 wells) using continuous flow records covering the period 1962-69, which w.as the driest period during- the 40 years (1956-96) for which records are available, showed that the regulated flow in the Xiangride River plus conjunctive use of Balong groundwater satisfied demands in all years including a surplus for uses downstream of Keri Dam equivalent to 115 Mcm per year. Similar modeling showed that flows in the Hatu and Qingshuihe Rivers are sufficient to meet projected demands in 75 percent of rain years, which is consistent with current irrigation standards in China. In common with all irrigation areas in China, standard operating procedures are invoked to equitably ration water in drought years. Thus, there is an acceptable reliability of supply. As will be mentioned further below, the abstraction of water for the project area will come at the possible expense of water supplies to the saline swampy areas to the north of Balong and this may have indirect environmental consequences. Regarding flood risks, all rivers flowing out of the mountains are subject to periodic very high flows. These have been taken into account in project design with all development areas being provided with flood protection bunds. 123

In conclusion, the environment in which the proposed development is to be undertaken poses certain well-defined risks that accompany all similar developments in the arid and semi arid zones of China. The degree to which these risks are managed depends on the effectiveness of the key environmental monitoring programs that are included in the project design, and the effectiveness of the training and development programs that are implemented to educate and assist the project participants. Additional possibilities are provided through adoption of new technologies such as reduced tillage for improved wind erosion protection and moisture conservation which has been accepted as feasible in the area.

The second category of environmental issues relates to the possibility that the Qinghai activities could adversely affect environmental conditions in and around the site. The main risk in this regard arises from the fact that the project would involve a significant increase in the area of intensive agricultural development in the project area, which raises the potential for both direct and indirect environmental effects.

The major form of direct environmental impact having potential significance arises from conversion of approximately 19,000 hectares of lower piedmont land to irrigated agricultural land. In its present state, these lands have very limited value as pasture, with only about 1,300 hectares being suitable for low grade (that is, unimproved) pasturage, virtually all of which is located in areas of shallow groundwater along the northern boundaries of the project area. The majority of the project development area supports very sparse populations (ground cover is less than 5 percent) of xerophytic plants that are widely distributed across similar terrain around the Qaidam Basin. The direct ecological impact of the project, in terms of displacement of existing natural ecosystems, is thus considered to be insignificant. It should be noted that the proposed developments will provide certain ecological benefits insofar as floral biodiversity and the structural complexity of the local ecological system will be substantially increased, albeit artificially, and there will also be local micro-climatic benefits (moderation of seasonal temperature variations and increased humidity) which have been well-documented in China.

As previously mentioned, the proposed new regulating reservoir will occupy an area of 128 hectares of piedmont land which is at a slightly higher elevation than the irrigation areas and thus is more gravelly and sparsely vegetated. This reservoir is also located "off-river" and hence will not interrupt flow in the Xiangride/Qaidam river system. The environmental consequences of this development are limited.

There will be direct social impacts on both herder families (63 households or 352 persons in all) who use pasturage along the northern boundary of the project area as well as other herders (289 households or 2,411 persons) who drive their flocks through the area as they move them between summer and winter pastures north and south of the project area. Compensation arrangements have been proposed and, to the extent that these are freely agreeable and acceptable to the affected, persons, it is difficult to envisage any adverse environmental impact potentials arising.

The main indirect potential environmental impacts arise from the application of approximately 300 Mcm of water to the new irrigation areas, much of which presently flows into the saline swampy areas north of Balong. Modeling undertaken as part of project design indicated that, in the long term (5 to 10 years) the average groundwater level in the swampy areas could be reduced by up to 0.5 meters immediately north of the Balong irrigation district. It should be noted that groundwater levels in the area fluctuate substantially throughout the year depending on the flow in the Xiangride/Qaidamu system and their tributaries (the fluctuation is reportedly in the range 0.1 to 0.5 meters). If there is an adverse effect, it is likely to occur on the margins of the area where vegetation is probably surviving close to the limits of its tolerance and has much less capacity to accommodate variations in the current pattern of water availability. The likelihood of 124 this happening is conjectural at the present time due to the inherent inaccuracies of groundwater modeling.

It will be necessary to monitor groundwater levels and vegetation distribution to the north of the Qinghai-project area to assess the degree to which these effects materialize and determine the need for any corrective action. It should be noted that the impact is substantially determined by three factors: i) the balance between the use of surface and groundwater in the Balong irrigation sub-district; ii) the degree to which groundwater is exploited in the Qinshuihe Irrigation sub- district (to the west of Balong); and iii) the locations of the groundwater wells. The surface/groundwater usage balance selected for the preliminary design is intended to strike a balance between environmental, practical and financial factors. There are other feasible design and management options which can be adopted if necessary and would either leave groundwater levels in the swampy areas unchanged or actually increase them in some parts (for example, if no groundwater pumping was included in the project, water tables throughout the Balong area will progressively rise, extending the swampy area substantially but destroying the irrigation area).

G. Gansu

Proposed Gansu Activities

The Gansu activities will be implemented in 2,234 villages in 12 counties in southeastern Gansu Province. All land covered by the sub-project is already being used for agricultural production. The purpose of sub-project investments is to promote adoption of improved, more environmentally sustainable agricultural crops and technologies and to add value to agricultural outputs close to the point of production to increase income retention in the sub-project areas. Physical interventions include land development, small-scale irrigation, improved crop production, livestock development, forestry development, rural enterprise development and provision of rural infrastructure.

Land Development. The Gansu activities will support land terracing over an area of about 21,000 hectares of moderately sloping and presently degraded land.

Small Scale Irrigation. The activities will support installation of micro-irrigation systems, purchase of sprinkler sets, lining of existing irrigation canals and improved rainwater collection cisterns. The areas concerned are small and widely dispersed throughout the Gansu counties.

Crop Production. The activities in Gansu will support improved planting offood crops over an area of about 31,000 hectares. Interventions include provision of plastic mulch for maize in 11 counties and plastic mulch for wheat in four counties, provision of virus-free potato seed in four counties, promotion of wheat/corn inter-planting in two counties and promotion of high production broad bean planting in four counties. Activities will also support production of vegetable cash crops (9,000 hectares), fiber hemp production (1,040 hectares), and introduction of four species of Chinese medicinal plants and tubers (7,780 hectares). The production of Chinese medicinal herbs would focus on higher altitude areas of Gansu where poverty is endemic.

Livestock. Activities will support pig and cattle fattening, sheep production by individual households to supplement farm crop income, mixed livestock production systems using remaining natural grassland areas, commercial scale breeding pig and cattle fattening production operations, soft-shelled turtle production, and fish production. Most feed for ruminant animals and pig production and turtles and fish production would be obtained from crop byproducts and manufactured feed sources and have little additional impact on natural grassland ecosystems. Fodder legumes would be fed directly to livestock or processed into higher quality feeds. By 125 products of maize and broad bean production to be supported under the project will provide animal feed and ammoniated grain straw is useful as livestock feed. A large amount of surplus crop byproducts and the technology to improve feed quality are available. Based on 1999 forage balance determinations the project would have, at full production, a 42 percent forage surplus over incremental livestock demand.

A total of 11 household livestock production models have been developed to suit different conditions in different counties. The majority incorporates multiple species (that is, sheep and cattle or sheep, cattle, pigs, etc.) and most rely totally on crop byproducts and/or prepared feed for inputs. However, grazing on natural grasslands is planned in three counties in Gansu, representing a total of 8,050 hectares or 19 percent of the total natural grassland area in these three counties. Most of these areas are in good to excellent condition due to a rotational grazing system that is already being followed. However, some other areas have been degraded by use of the land for cropping. About 2,300 hectares of this land would be rehabilitated under the project by planting grasses, alfalfas and clovers which will then be used as livestock forage. The sustainability of all production systems has already been established through preparation of county level feed balances. Additional livestock and grassland management plans must be prepared by each participating village as a condition of participation in Gansu activities.

Forestry. The activities will support planting of about 15,500 hectares of forests and perennial tree crops including apple and pear, prickly ash, walnut, Japanese larch, mulberry, olives and tea. These species will be planted on areas of steeper land that will be retired from cropping as part of the Gansu activities.

Rural Enterprise Development. The activities will support investments in a wide range of enterprise developments that have in common involvement of value-added processing of products generated from the project areas and addition of this value as close to the point of production as possible to maximize income retention in the project areas. There are two categories of investment: i) large scale development (investment greater than US$30,000); and, ii) small household enterprises. The Provincial Government has identified several candidates within the first category, including fruit storage, pig breeding farms, basket-making factories, medicinal plant processing facilities alfalfa processing and carpet weaving factories. With the exception of the pig breeding farms, none of these are of significant environmental consequence. Nevertheless, before any are accepted into the project they have to undergo defined technical and financial assessments including evaluation by Provincial and County level (EPBs) for compliance with any relevant emission standards and, as required, supervision of commissioning and operation of pollution control facilities.

Rural Infrastructure. The activities will support construction of 174 kilometers of all weather gravel roads in 10 counties linking existing villages with county road networks. Three small bridges will also be constructed. Improved water supplies will be provided to about 260,0OG people in all participating counties. Supply systems will include springs, existing surface sources and rainwater runoff.

Existing Environmental Conditions

The 12 participating project counties in Gansu, are all located in the hilly parts of southeastern Gansu Province with elevation varying from about 550 meters to more than 4,000 meters. Environmental conditions are extremely variable due, primarily, to topographical variations. The climate ranges from sub-tropical humid in the southern parts of the project area to temperate humid in the north. Average annual temperatures range from about 6°C (Minxian County) to 150 C (Wenxian County) and average annual rainfall ranges from about 440 millimeters (Wenxian County) to 800 millimeters (Kangxian County) although the majority of counties fall in the lower 126 end of the range. Growing seasons range from about-138 days in Qingshui County to 262 days in Wenxian with the majority experiencing growing seasons less than 160 days.

Five of the participating counties are located in the hills of the Loess Plateau and all but one of these (Minxian County) suffer from soil erosion which ranges from bad to very severe. Two of these counties (Xihe and Wudu) are more than 60 percent affected by severe erosion. There is virtually no natural vegetation left in the project area, most having been long converted to agricultural use and much of that having been since turned into wasteland through unsustainable usage. There are two high altitude Nature Reserves in the region (in Wenxian and Wudu Counties) but they are not located in the project areas.

Environmental Management Issues

The objective of the Gansu activities is primarily to reduce poverty through a variety of measures, some of which include the promotion of agricultural activities which are more environmentally sustainable than those currently being followed in the region. There is little doubt that the primary environmental risk associated with the project's development strategy arises from the proposed livestock development program. During the project design-phase, considerable work was done to prepare livestock development models including feed balances to ensure that the feed and fodder resources necessary to support the proposed livestock production systems will actually be available. Analyses done so far suggest that the resources arising from project investments alone are likely to significantly exceed feed demands. Nevertheless, these balances are based on County level data and provisions have been included in the project design for preparation, review and clearance of township level balances prior to disburseimentsbeing authorized for those townships as set out in the project documents. These safeguards were progressively developed through the project design phase and final environmental clearance depended on the inclusion of final arrangements as outlined and relevant conditionalities in the proposed legal agreement.

H. Inner Mongolia

Proposed Inner Mongolia Activities

The proposed activities in Inner Mongolia comprise investments in 136 townships within 15 counties/Banners covering a total project region that is about 138,000 square kilometers in area. Investments will be made in crop production (improved production of existing crops and introduction of certain new crops), livestock production, tree planting/forestry development, small-scale irrigation, land terracing/soil conservation, agro-enterprises and rural infrastructure.

Crop Production. The activities will finance improved production of a wide variety of existing crops, particularly maize and potatoes, which account for about 75 percent of the total area to be planted, but including also subsidiary crops such as buckwheat, peas and hybrid sunflowers. The total area over which such developments will be supported is 87,000 hectares spread across all participating counties/Banners. As in Gansu, all of these areas are already being used for agricultural production so that the main effect of the project is to intensify an existing land use rather than to introduce a major change in use. Also as in Gansu, the Inner Mongolia activities would supply households with the first year's requirements of quality inputs (seeds, plastic mulch, etc.).

Livestock Component. The project will support development of a wide variety of integrated household production systems (11 different models are possible), all but one of which are based on cut-and-carry plus natural pasture grazing approaches. The livestock to be supported under the different models include sheep, goats, cattle, rabbits, chickens, and pigs and it is estimated that about 139,000 households will participate in the program. 127

The integrated household pasture and fodder production system to be supported under the project is referred to as caokulun, a sustainable livestock management system that has been developed to assist in reversal of the rangeland degradation occurring in Inner Mongolia in recent decades. A caokulun is one hectare of land protected by a wall and a tree-shrub shelterbelt in which irrigated fodder and grain crops are produced to benefit two households. Associated with the caokulun is an area of natural pasture assigned to each household (on a 30-year leasehold basis) on which household livestock are permitted to graze, and which will fall under the prevailing grassland management regulations. The livestock feed produced by the caokulun is used to supplement natural pasture forage obtained by grazing during summer and autumn and to provide a full feed ration to household livestock during the winter and spring, thus relieving grazing pressure at critical times to improve ecological condition of natural pasture ecosystems.

The project also includes a component that will demonstrate, on a pilot scale, another improved integrated farming system. The pilot will be undertaken on an area of about 1,370 hectares of natural grassland that is presently being over-grazed. It will demonstrate an integrated farming system comprising mixed crop and livestock farming, utilizing crop by products as the main source of ruminant forage for cattle and sheep, and recycling animal manure to the croplands. Finally, it is proposed that approximately 2,000 hectares of degraded grassland be rehabilitated through planting of improved pastures and reserved for forage harvesting (that is, not for grazing) while an additional 2,200 hectares will be fenced off for and managed under a regulated grazing regime. These activities are designed to complement the Inner Mongolian Government's existing reseeding and grassland rehabilitation programs.

Tree Planting. The activities will support planting of trees or protective bushes on an area of about 15,000 hectares of severely degraded land. Species to be planted include trees such as larch and wild apricot.

Small Scale Irrigation. The activities will support development of about 11,500 tube wells/dug wells to irrigate an area of about 22,000 hectares (that is, about 0.5 hectares/well). Canal lining would be provided on existing canals servicing an area of about 3,760 hectares. Finally, rainwater-based harvesting systems would be installed, covering about 3,300 micro-irrigation systems which are used for supplementary watering of rain fed crops.

Land Terracing/Soil Conservation. An area of 8,500 hectares of presently cultivated land with slopes less than 15° will be terraced while an additional 5,900 hectares of wasteland and degraded land along roadways and on ridges will be planted with trees for wind erosion control.

Agro-enterprises. The activities will provide financing for farmers to procure small feed chopping and threshing machines and small tractor sets. The scheme is voluntary. For the purposes of project planning it has been assumed that demand for about 200 sets of each might be realized.

Rural Infrastructure. Forty-five kilometers of existing rural roads will be rehabilitated to Class III standard. One hundred and eighteen drinking water supply systems will be installed (dug wells and/or cisterns), and 860 kilometers of low voltage power lines (10 kV and 0.4 kV) will be installed. 128

Existing Environmental Conditions

The proposed project area is located in the central part of Inner Mongolia Autonomous Region, south of the Mongolian Plateau. It is totally within the Mongolian Steppe biological unit although over the 1,000 kilometers of project area, four sub-zones are defined based on topographical factors. The climate throughout the area is semi-arid, with average annual rainfall ranging from 325 millimeters to 390 millimeters, average annual temperatures ranging from 1.20 C to 5.6° C, and average windspeeds being moderate to high (ranging from 3.2 to 3.8 meters/second). Water resources (mainly groundwater) in the region are generally adequate although unexploited (it is estimated that only about 25 percent of the sustainably usable groundwater supply in the area is presently tapped) and soils are generally well-structured but nutrient deficient (predominantly brown chestnut soils). The dominant vegetation type is steppe grassland, which accounts for about 70 percent of the project region and, in many places, has been overgrazed and is degraded. Crops account for about 12 percent of the area (less than 10 percent of this is irrigated) and forests for about 9 percent.

The area contains some of the best grasslands in China and includes a wide variety of species including Stipa spp., Agropyron spp., Festuca spp., and Cleistogenes spp. which often can be highly productive, producing in excess of 2 tons/hectares of fresh grass when managed properly. Salt marshes occur in the lower lying parts, depressions without drainage and the dominant species in these areas are common ones such as Carex spp. and Phragmites spp. Terrestrial wildlife is generally of limited diversity and abundance and dominated by domesticated species (sheep, goats, horses, etc.). Native domesticates include the Mongolian Horse and Mongolian cattle. There are nine Nature Reserves in the region, two of which are State level, making them of national significance. Only one of these is located in the project region but it is not located in a participating Banner and will not be directly or indirectly affected by the project.

Environmental Management Issues

The primary environmental management issue in Inner Mongolia is the same as for the proposed Gansu Province developments; that is, promotion of increased livestock production for poverty alleviation runs the risk of exacerbating unsustainable use of rangelands. During the project design-phase, considerable work was done to prepare livestock development models including feed balances to ensure that the feed and fodder resources necessary to support the proposed livestock production systems will actually be available. Analyses done so far suggest that the resources arising from project investments alone are likely to significantly exceed feed demands. Nevertheless, these balances are based on County level data and provisions have been included in the project design for preparation, review and clearance of township level balances prior to disbursements being authorized for those townships. The details of the proposed safeguards are outlined in the Environmental Mitigation and Monitoring Plan. These safeguards were progressively developed through the project design phase and final environmental clearance- depended on the inclusion of final arrangements as outlined and relevant conditionalities in the proposed legal agreement. 129

I. Environmental Mitigation and Monitoring Plan and Additional Studies

Introduction

The environmental mitigation and monitoring strategy is based on lessons learned from many years of attempts to apply environmental management criteria to broad-scale rural development projects not only in China, but throughout the East Asian region. The approach taken also acknowledges that an integral project objective is the promotion and financing of environmentally sustainable and/or reduced environmental impact rural production methods and, as such, there is a commonality of intention between general project management and supervision activities (by all participants - World Bank and counterparts) intended to guarantee project success, and those intended to avoid and minimize adverse environmental effects.

In this context, the key criteria that have been followed in development of the project as a whole and its environmental management strategy have been as follows: (i) To the maximum extent possible, environmental objectives should be pursued through the design of the project itself rather than relying unduly on exparte supervisory processes; (ii) Given the potential associated costs, environmental monitoring and supervisory activities should be focused on those activities and processes which have the potential to register adverse environmental impacts that are significant within the environmental and developmental contexts in which the project is to be carried out. Furthermore, monitoring of the physico/chemical environment should only be undertaken where meaningful baseline data exist or can be cost- effectively developed upon which to base interpretation of results; (iii) The environmentally sustainable development technologies supported under the project should, to the extent possible, be selected from amongst those technologies that are already being applied within the project area and are well understood and accepted by project participants. To the extent that new technologies or modifications to existing technologies are to be promoted, it should be through research and/or demonstration and/or pilot activities; (iv) The technological approaches and environmental performance standards required should be fully consistent with existing government programs and standards; and (v) As much reliance as possible should be placed on the use of existing government regulatory, management and control agencies in preference to constructing "shadow" management and control systems which are unique to the project.

Mitigation Strategies and the Environmental Monitoring Program. An Environmental Mitigation Strategy is provided as Table 10-1, which sets out the main, potentially adverse environmental issues, in declining order of significance, that have been identified through a variety of sources including the provincial project EIAs, supplementary technical reports by relevant Chinese consultants and World Bank advisers, and field based review of these elements of the proposed project. It indicates in which PPs the issue arises to a significant degree, the nature of the mitigation measures incorporated into project design and the assessed need for any supplementary monitoring and control activities. This is complemented by a Project Monitoring Program, which is provided in Table 10-2. Each village in the Project area would prepare and monitor an annual livestock and grassland management plan to ensure acceptable livestock feed balances and environmentally sustainable use of grassland. In addition, a Dam Safety Panel would conduct periodic reviews of the adequacy of the design and construction of the project dams during construction and at the start of operations. 130

Table 10-1: EnvironmentalMitigation Strategies

, , ,A,s~~~~~~~~~~~~~~~~~~...... ::,:: . :: ...... ::

O. ~4$4~... . ~NoeA

1. Livestock 1. I Feed Availability All 1. Adoptionof production systems alreadybeing used in or near the project area Yes 2. Developmentof a variety of production systemsto allow participantsto choose based on their own judgement 3. Preparationof County level feed balances to establishthat, in principle,the proposed systems are sustainable 4. Developmentand agreementof "Five Point ManagementPlan" (see Note B) 5. Reinforcementof the agreed plan through inclusionof legal conditionof the Credit Agreement 1.2 Animal Wastes Gansu 1. EPB clearancefor larger scale production units Yes 2. Proven design adopted for small scale units 2. Irrigation/WaterResources 2.1 Water Availability Qinghai 1. Design based on availablehistoric data Yes 2.2 Downstream Impacts(surface) Qinghai No specific mitigations included Yes 2.3 DownstreamImpacts (groundwater) Qinghai 1. Modelingof project developmentoptions Yes 2. Selectionof preferred optionwhich provides a balance between various development/environment criteria 2.3 Downstream Impacts(ecological) Qinghai ProposedManagement Plan for the Balong Soak Yes 2.4 Direct ecological Effects Qinghai 1. Location of proposed developmentarea No 2.5 Construction impacts 1 Project focus on small scale irrigationwnArks Yes 2. Inclusionof standard constructionphase environmentalsafeguards for small scale civil works 3. Rural Enterprises All (but Gansu 1. Focus investmentson small scale,household developmentswhich generally have little or no Yes in particular) environmentalimpact potential; 2. Provisionof administrativestep in ABC loan processingprocedures which prevents approval of loans Yes for developmentswhich require EPB approvalbut do not have it. 4. Land Reclamation Qinghai 1. Soil surveysto identify saline/sodiclands Yes 2. Reduction in scope of project to reduce areas of saline/sodiclands 3. Specificationof leaching and water management proceduresto maintain salt front below root zone based on existingproven procedures 2. Strip leveling of hummockedland to reduce susceptibilityto wind erosion 3. Restrictionof constructionactivities during the windy period (see Note C) 4. Inclusion of major shelterbeltplanting program in the project for longer term control

5. Crop Production 5.1 PlasticMulch All No special mitigationsother than in Gansu where a plastic recycling plant has been proposed as a No possible candidatefor financingthrough the credit scheme

5.2 Agro-chemicalusagc All I. Bank will not finance procurementof pesticides/herbicides Yes 131

2. Programnincludes funds for extension which contains standard packages on agro-chemical. management 3. Program includes provisionfor adaptive research to optimize input usage and promote manuring technologies 4. Program includes extensive soil conservationcomponents which have the indirect effect of helping reduce off-farm, non-pointemissions of all sorts 6. Rural Infrastructure 6.1 Roads and Bridges All 1. Focus on very small scale developments(generally Class III and IV gravel roads) No 2. Preferenceto upgradingof existingroads 3. Focus on farm access roads, inter-villageroads and connector roads to higher levels within the road hierarchy so as to avoid risk of increasingaccess to remote areas 6.2 Water Supply 1. Project to finance only very small scale household based systems No 2. System designs include basic safeguardsfor public health (for example, provisionof lips on water supply wells) 6.3 Rural Electricity Project to financeonly local distributors No 7. Forestry 7.1 Development 1. Selectionof species/techniqueswhich have already been proven in the project areas No 2. Plantings include economic forests plus fuelwood areas based on quantitativedemand estimates to ensure that participantshave an economic stake in protectingtree plantings 7.2 Management 1. Participantshave a financial stake in the managementof economic forests Yes 2. Public education (in the case of Qinghai- this is part of the normal extension program delivered to migrants in desert reclamation areas) 8. Other 8. 1 Impact on Nature Reserves and other All 1. All project developmentsspecifically excluded from such areas No conservationareas 2. Participatingconsultants requested to confirmthat the design guideline was complied with

NOTES: A. The monitoringwork for which a need is identifiedin this column is additional work or increases in already proposed programnsthat is required to provide a satisfactorylevel of assessmentof actual environmentalimpacts. B. The Five Point Plan which was agreed to as part of project preparationis as follows: * Natural pasture land use will conformto existing grassland managementand monitoringlaws and regulations; * Ruminant livestockfeeding will be based foremost on availablecrop byproducts and forages rather than on grazing natural pasture in order to minimize pressure on fragile grasslands; * Projectparticipants who lease crop and/or grasslands will receive long-term(30+ years) land leases under which they will have the rights but also the responsibilitiesof land use; * Livestockdensities will be based solely on realisticallyprepared and annually updated feed balances for green forage, crop byproducts,and feed grains; seasonal highs and lows in spring and fall; and types of livestockto be raised; and, * The project village uni! and its local governmentwill be the basic unit to develop, implementand update grass and cropland managementplans C. These managementXovisions are followed as standard operating procedure in desert reclamationprocedures. 132

Table 10-2: Proposed Environmental Monitoring Program

Livestock Feed Availability All PPs 1. Check village feed balances PMOs On demand CC 2. Monitorarea and status of grassland arcas surroundinga selection of Assigned Annual basis PB developmentareas in all PPs organizations

(Note B).______AnimalWastes All PPs 1. Reviewof random sampleof ABC loan approvalsto ensure that the Bank During supervision BB necessary EPB clearances were secured Irrigation/WaterResources Water Availability Qinghai 1. Continuationof existingflow monitoring Provincial Water Continuous CC 2. Continuationof recordingof existingbulk water supply deliveries ex- Resources Bureaus. Hatu and commencementof such recordings ex-Keri Downstream Impacts (surface) Qinghai Monitor applicationsof inputs plus periodicmonitoring of surface water County EMS (Note C) Twice/year PB quality in a representativesample of locationsupstream, middle and downstreamof swampy area north of Balong Downstream Impacts Qinghai Monitor pumping/applicationrates in tubewell supplied areas, Assigned Annual PB (groundwater) groundwaterlevels in tubewells and groundwaterlevels in monitoring organizations(Note D) wells in the swampy area adjacent to Balong. Downstream Impacts Qinghai Review Statusof Management Plan for BalongSoak. Contracted institution Annual starting PB (ecological) Annual surveys of land cover and wildlife numbers. Year 3 Construction Impacts All Randomfield inspectionsof civil works Bank Bi-annual during BB ______i iti l_l___yeaye rsitiars__ Rural Enterprises All (but Randomchecking of loan approvalsto ensurethat no loans were PMOs/Bank missions Random CC/BB Gansu in approvedfor proposals which lack EPB clearances. particular) Land Reclamation Qinghai 1. Further detailed soil surveys during detailed design Design institute As required PB 2. Sampling and analysis of soil and drainage water salinity/sodicity Design institute Monthly PB during leaching 3. Monitoringof drainage discharges (quantityand quality) from the Design institute Monthly PB developmentareas 4. Supervisionof constructionprocedures to ensure compliancewith AgricultureBureau Ad hoc CC specifications 5. Inspectionof shelterbeltplanting program Forestry Bureau Ad hoc CC Crop Production Agro-chemical usage All Review of procurementcontracts to ensure no pesticides/herbicidesare Bank As required BB financed Monitoringof progress of extension/trainingprograms Bank As required BB Monitoringof adaptiveresearch and soil conservationprograms Bank As required BB Monitoringinput levels in a random sampleof plots and comparison Contracted institution PB with relevant national standard (GB-4285-84) 133

NOTES: A. PB = project budget BB = Bank Budget (i.e. a responsibility to be accepted by the Bank and covered from its internal funds) CC = cost covered under general project monitoring budget. B. In Inner Mongolia, this work will be done by the Grasslands Management Bureau (GMB), an institution of the Regional Government, which already has this type of work within its purview. GMB will, however, require supplementary financing to cover direct costs associated with doing this work. The other two PPs are not as well equipped in respect of this type of monitoring and it is proposed that this work be done on contract by relevant universities or Institutes. C. EMS = Environmental Monitoring Station which is a unit of the Environmental Protection Bureau and usually the only institution at a county level capable of undertaking water chemistry analyses. D. The groundwater modeling work was done by Xinjiang Agricultural University, one of the leading groundwater modeling institutions in China and it is not unreasonable to assume that they may continue to be engaged on this work. However, they would have to apply for appointment on a competitive basis in accordance with relevant Bank procedures. 134

Costs of the Environmental Monitoring Program. As can be seen in Table 10-2, supplementary budgets will be required to finance ithe several activities over and above the project monitoring and reporting activities which are covered in other budget categories. The cost provisions made to cover these supplementary activities are provided in Table 10-3.

Table 10-3: Estimated Cost for Proposed Environmental Monitoring Program -MON1TeR1NG! hWI - -'COSTVROVr

Livestock Monitorarea and status of grasslandareas 1,200 surroundinga selection of developmentareas in all PPs Irrigation/WaterResources Monitorapplications of inputsplus periodic 400 monitoringof surfacewater quality in a representativesample of locationsupstream, middle and downstreamof swampyarea north of Balong Monitorpumping/application rates in tubewell 1,200 suppliedareas, groundwater levels in tubewells and groundwaterlevels in monitoringwells in the swampyarea adjacent to Balong Annualsurveys of landcover and wildlife numbers 300 1.Further detailed soil surveys during detailed 300 design; 2. Samplingand analysis of soiland drainage water 300 salinity/sodicityduring leaching; 3. Monitoringof drainagedischarges (quantity and 200 quality)from the development areas

TOTAL 3,900

The overwhelming majority of these expenditures (Yuan 3,900,000) will take place in Qinghai Province.

Additional Activities

To increase the contribution to the environmental sustainability of the proposed programs certain additional studies, development and training activities are proposed, which would require additional funding as follows:

Management Plan for the Balong Soak. The project already involves quite extensive monitoring of vegetation cover and groundwater of the Balong Soak to gauge potential impacts. This interpretation of monitoring results would be based on the assumption that maintenance of existing groundwater and vegetation is the optimum outcome, based on the "do no harm" philosophy. This assumption is not based on a detailed understanding of the dynamics of the ecosystem which, if better understood, may provide a basis for coordinated management of the irrigation area and the adjacent swampy areas to the benefit of both. The proposed work would begin with basic research into ecological relationships, flora and fauna surveys, identification of pressures on the area other than potentially hydrological issues (the main pressure is believed to be uncontrolled grazing by itinerant herders) which would lead to development of management possibilities and options. A consultation process would then be commenced with a view to achieving some consensus on how utilization practices might be altered, leading to production of an agreed management plan. The plan could form the basis for bilateral support for implementation. 135

This type of study would be well within the capacity of one of the leading ecological research institutes of the Chinese Academy of Sclences and the estimated cost is about Yuan 1.5 million which would be spent over three to four years.

Pest Management in Minority Areas. There already exists a comprehensive program of pesticide regulation, education and monitoring carried out through local Plant Protection Bureaux and Plant Protection Stations. Previous experience suggests, however, that much more could be done to improve the effectiveness of these programs in general and amongst minority communities in particular. The proposed activity would underwrite a series of activities to strengthen existing programs in the project areas through development of training materials, training of trainers and labeling in local languages. Again, this work could substantially be undertaken by local experts working with indigenous trainers, possibly with some assistance from a foreign pesticide management expert as adviser. Estimated cost is in the order of Yuan 1 million.

Training in Regional Land Use Planning. Urban land use planning in China is quite well developedalbeit unimaginative.Regional land use planning for rural areas is almosttotally non- existentand scarcely even exists as a discipline. The proposed program,to be focused in Gansu (due to the particular suitability of the issues they are facing to being addressed through regional planning approaches), would involve provision of training, carrying out of a regional planning case study, development and piloting of public consultation procedures to be carried out as part of the planning process. This would have to be done by foreign consultants. The estimated cost is in the orderof Yuan 1.5million

Supervision Needs. A substantial number of the field monitoring tasks required as part of the environmental monitoring program can be carried out by non-specialist staff provided that an environmental specialist participates in project launch workshops to agree with counterparts on formats for periodic progress reports to ensure a clear accounting of environmental issues. This specialist should probably return at the end of the second year when, based on previous experience, there should have been sufficient progress in the field to be able to actually inspect operations performance, particularly on construction-related issues. Specialist technical advice will also be required during initial missions to agree with consultants in Qinghai on the location of groundwater monitoring wells and monitoring/reporting protocols. Thereafter however, it should be possible to monitor progress in this regard through desk analysis of progress reports. An environmental specialist should participate in supervision missions at least once each year. 136

Annex 11: Statement of Loans and Credits CHINA: Western Poverty Reduction Project As of May 17, 1999

Difference Between expected Original Amount in US$ Millions and actual Fiscal disbursements a/ Project ID Year Borrower Purpose Cancel- Un- IBRD IDA lations disbursed Orig Frm Rev'd

Number of Closed Projects: 106

Active Pro ects CN-PE-3653 1999 PEOPLE'S REPUBLIC OF CHINA CONTAINER TRANSPORT 71.00 0.00 0.00 71.00 0.00 0.00 CN-PE-36953 1999 PRC HEALTH IX 10.00 50.00 0.00 60.00 0.00 0.00 CN-PE-41890 1999 PEOPLE'S REPUBLIC OF CHINA LIAONING URB TRANSP 150.00 0.00 0.00 150.00 0.00 0.00 CN-PE-49665 1999 PRC ANNING VALLEY AG.DEV 90.00 30.00 0.00 119.00 3.60 0.00 CN-PE-50036 1999 PEOPLE'S REPUBLIC OF CHINA ANHUI PROVINCIAL HWY 200.00 0.00 0.00 200.00 9.94 0.00 CN-PE-51856 1999 MINISTRY OF FINANCE ACCNTG REFORM & DEV 27.40 5.61 0.00 32.96 2.04 0.00 CN-PE-63123 1999 PRC YANGTZE FLOOD EMERGY 40.00 40.00 0.00 79.16 3.34 0.00 CN-PE-3539 1998 PRC SUST COAST RES DEV 100.00 0.00 0.00 96.00 .17 0.10 CN-PE-3566 1998 BASIC HEALTH 0.00 85.00 0.00 81.04 -. 20 0.00 CN-PE-35698 1998 PRC HUNAN POWER DEVELOP. 300.00 0.00 0.00 300.00 12.00 .00 CN-PE-3591 1998 PRC STATE FARM2S COMMERCI 150.00 0.00 0.00 102.59 -30.24 0.00 CN-PE-3606 1998 GOC ENERGY CONSERVATION 63.00 0.00 0.00 63.00 3.18 0.00 CN-PE-3614 1998 PEOPLE'S REPUBLIC OF CHINA GUANGZ. CITY CRT.TRP 200.00 0.00 0.00 192.00 7.12 0.00 CN-PE-3619 1998 MINISTRY OF FINANCE 2ND INLAND WATERWAYS 123.00 0.00 0.00 123.00 20.00 0.00 CN-PE-36414 1998 CHINA GUANGXI URBAN ENV. 72.00 20.00 0.00 88.81 -1.03 0.10 CN-PE-36949 1998 PEOPLES REPUBLIC OF CHINAA NAT.HWY 3-HUBEI 250.00 0.00 0.00 234.96 6.63 0.00 CN-PE-40185 1998 PRC SHANDONG ENVIRONMENT 95.00 0.00 0.00 92.00 23.80 0 .00 CN-PE-45788 1998 PEOPLE'S REPUBLIC OF CHINA TRI-PROVINCIAL HWY 230.00 0.00 0.00 230.00 43.54 0.10 CN-PE-46563 1998 PRC TARIM BASIN II 90.00 60.00 0.00 140.69 6.80 0.00 CN-PE-46952 1998 PRC FOREST. DEV. POOR AR 100.00 100.00 0.00 200.68 9.17 18.34 CN-PE-49700 1998 PRC IAIL-2 300.00 0.00 0.00 280.00 7.30 0.30 CN-PE-51736 1998 GOC E. CHINA/JIANGSU PWR 250.00 0.00 0.00 250.00 102.64 0.00 CN-PE-56491 1998 PRC HEBEI EARTHQUAKE 0.00 28.40 0.00 5.66 -13.33 O.C0 CN-PE-34081 1997 PRC XIAOLANGDI MULTI. II 430.00 0.00 0.00 323.36 92.85 0.00 CN-PE-3590 1997 PRC QINBA MTS. POVTY RED 30.00 150.00 0.00 148.20 27.18 0.00' CN-PE-3635 1997 PRC VOC. ED. REFORM PROJ 10.00 20.00 0.00 11.98 -1.22 0.00 CN-PE-3637 1997 PRC NATL RUR WATER III 0.00 70.00 0.00 62.02 11.81 0.00 CN-PE-36405 1997 PRC WANJIAZHAI WATER TRA 400.00 0.00 0.00 320.09 13.42 0.00 CN-PE-3643 1997 PRC XINJIANG HIGHWAYS II 300.00 0.00 0.00 230.72 62.72 0.0C CN-PE-3650 1997 GOC TUOKETUO POWER/INNER 400.00 0.00 0.00 391.00 144.78 3.51 CN-PE-3654 1997 PRC HUNAN/GUANG HWY2-NH2 400.00 0.00 0.00 321.24 81.22 0.00 CN-PE-36952 1997 PRC BASIC ED. IV 0.00 85.00 0.00 41.32 -24.85 0.00 CN-PE-38988 1997 PRC HEILONGJIANG ADP 120.00 0.00 0.00 93.02 10.71 0.00 CN-PE-44485 1997 SHANGHAI WAIGAOQIAO 400.00 0.00 0.00 382.70 35.94 0.00 CN-PE-34618 1996 PRC LABOR MARKET DEV. 10.00 20.00 0.00 23.99 22.66 0.00 CN-PE-3507 1996 GOC ERTAN HYDRO II 400.00 0.00 0.00 23.48 -21.26 0.00 CN-PE-3563 1996 PRC ANIMAL FEED 150.00 0.00 0.00 138.78 113.97 22.13 CN-PE-3569 1996 P.R.C. SHANGHAI-ZHEJIANG HI 260.00 0.00 7.75 106.81 32.91 23.21 CN-PE-3589 1996 PRC DISEASE PREVENTION 0.00 100.00 0.00 68.97 59.63 0.00 CN-PE-3594 1996 PRC GANSU HEXI CORRIDOR 60.00 90.00 0.00 125.49 27.79 1.00 CN-PE-3599 1996 YUNNAN PROV. GOV. YUNNAN ENVIRONMENT 125.00 25.00 0.00 139.11 20.95 0.00 CN-PE-3602 1996 PRC HUBEI URBAN ENV. PRO 125.00 25.00 0.00 123.98 78.98 0.00 CN-PE-3638 1996 PRC SEEDS SECTOR COMMER. 80.00 20.00 0.00 73.11 22.34 0.00 CN-PE-3646 1996 PRC CHONGQING IND POL CT 170.00 0.00 77.99 91.22 106.20 0.00 CN-PE-3648 1996 SHANGHAI MUN. GOVT SECOND SHANGHAI SEWE 250.00 0.00 0.00 148.15 72.54 0.00 CN-PE-3649 1996 CHINA SANXI POVERTY ALLEV 0.00 100.00 0.00 34.51 -5.88 0.00 CN-PE-3652 1996 PRC 2ND SHAANXI PROV HWY 210.00 0.00 0.00 139.29 56-.&6a 0.00 CN-PE-36950 1996 PRC BASIC ED. POOR III 0.00 100.00 0.00 9.05 -21.13 0.00 CN-PE-40513 1996 PRC 2ND HENAN PROV HWY 210.00 0.00 0.00 175.14 55.81 0.00 CN-PE-3493 1995 PRC INLAND WATERWAYS 420.00 0.00 0.00 270.62 -10.79 0.00 CN-PE-3571 1995 PRC RAILWAYS VII 420.00 0.00 29.00 344.67 209.66 1.51 CN-PE-3585 1995 GOC SHENYANG IND. REFORM 175.00 0.00 0.00 97.54 52.68 0.00 CN-PE-3596 1995 PRC YANGTZE BASIN WATER lD0.00 110.00 0.00 27.75 -16.90 0.00 CN-PE-3598 1995 LIAONING ENVIRONMENT 110.00 0.00 0.00 64.96 54.86 0.00 CN-PE-3600 1995 PRC TECHNOLOGY DEVELOPME 230.00 0.00 0.00 123.15 45.73 0.C0 CN-PE-3603 1995 PRC ENT. HOUSING SOC. SE 275.00 75.00 20.00 196.98 156.61 9.26 CN-PE-36041 1995 MOF FISCAL & TAX REF. & 50.00 25.00 0.00 66.29 41.78 0.00 CN-PE-3612 1995 PRC XINJIANG HIGHWAY I 150.00 0.00 0.00 41.95 33.63 0.00 CN-PE-3634 1995 PRC MATERNAL CHILD HEALT 0.00 90.00 0.00 18.40 9.92 0.00 CN-PE-3636 1995 PRC BASIC EDUC IN POOR & 0.00 100.00 0.00 3.38 -. 63 0.00 CN-PE-3639 1995 PRC SOUTHWEST POV. REDUC 95.00 200.00 0.00 135.27 28.38 0.00 CN-PE-3642 1995 ZHEJIANG POWER DEVT 400.00 0.00 0.00 129.95 -44.08 0.00 CN-PE-3647 1995 PRC ECONOMIC LAW REFORM 0.00 10.00 0.00 5.51 5.72 0.00 CN-PE-36947 1995 GOC SICHUAN TRANSMISSION 270.00 0.00 0.00 132.50 120.94 84.43 137

Difference Between expected Original Amount in US$ Millions and actual Fiscal disbursements a/ Project ID Year Borrower Purpose Cancel- Un- IBRD IDA lations disbursed Orig Frm Rev'd CN-PE-37156 1995 PRC IODINE DEF. DISORDER 7.00 20.00 7.00 1.77 11.44 -. 32 CN-PE-3502 1994 MOH RUR HEALTH MANPOWER 0.00 110.00 0.00 27.28 21.76 0.00 CN-PE-3504 1994 PRC HEBEI/HENAN NATIONAL 380.00 0.00 0.00 39.36 11.01 0.00 CN-PE-3540 1994 PRC LOESS PLATEAU 0.00 150.00 0.00 27.92 -21.54 0.00 CN-PE-3557 1994 PRC FOREST RESOURCE DEV 0.00 200.00 0.00 40.42 14.68 -30.70 CN-PE-3562 1994 PRC XIAOLANGDI 460.00 0.00 0.00 .48 .48 0.00 MULTI PURPOSE CN-PE-3586 1994 PRC SHANGHAI ENVIRONMENT 160.00 0.00 0.00 58.50 53.61 0.00 CN-PE-3593 1994 PRC SONGLIAO PLAIN ADP 0.00 205.00 0.00 24.64 -11.85 0.00 CN-PE-3595 1994 PRC RED SOILS II DEVELOP 0.00 150.00 0.00 40.31 17.62 0.00 CN-PE-3609 1994 GOC SICHUAN GAS DEV 8 255.00 0.00 0.00 121.72 72.23 0.00 CONSERVATION CN-PE-3622 1994 SEANGHAIMUNICIPAL GOVT SHANGHAI MTP II 150.00 0.00 0.00 7.36 7.37 0.00 CN-PE-3626 1994 GOC FUJIAN PROV HIGHWAY 140.00 0.00 0.00 58.63 39.29 11.14 CN-PE-3633 1994 GOVERNMENTOF PRC TELECOMMUNICATIONS 250.00 0.00 30.00 14.55 44.55 _- -3.05 CN-PE-3641 1994 PRC YANGZHOUTHERMAL POW 350.00 0.00 0.00 73.39 58.82 0.00 CN-PE-3644 1994 PRC XIAOLANGDI 0.00 110.00 0.00 25.53 10.12 0.00 RESETTLEMENT CN-PE-3473 1993 P.R.C. ZHEJIANG MULTICITIES 0.00 110.00 0.00 19.77 19.19 0.00 CN-PE-3512 1993 GOVTOF PEOPLES REP. OF C SHANGHAI PORT REST. 150.00 0.00 25.74 3.88 29.27 3.20 CN-PE-3518 1993 PRC GUANGDONGPROV. 240.00 0.00 0.00 2.08 .06 0.00 TRANSPORT CN-PE-3533 1993 TIANJIN IND. II 150.00 0.00 40.35 13.59 44.93 9.60 CN-PE-3559 1993 PRC AGRIC. SUPPORT SERVI 0.00 115.00 0.00 10.40 -1.16 0.00 CN-PE-3561 1993 PRC SICHUAN ADP 0.00 147.00 0.00 5.48 0.00 0.00 CN-PE-3567 1993 PRC EFFECTIVE TEACHING S 0.00 100.00 0.00 6.25 4.64 4.70 CN-PE-3570 1993 PRC RAILWAY VI 420.00 0.00 0.00 93.77 93.12 0.00 CN-PE-3580 1993 PRC SO.JIANGSU ENVIRON. 250.00 0.00 0.00 4.15 .48 0.00 PROTECT. CN-PE-3581 1993 PRC HENAN PROV. 120.00 0.00 0.00 8.92 8.93 0.00 TRANSPORT CN-PE-3592 1993 PRC REF. INST'L.& PREINV 0.00 50.00 0.00 18.45 18.48 0.00 CN-PE-3597 1993 PRC TAIHU BASIN FLOOD CO 100.00 100.00 0.00 47.43 43.08 0.00 CN-PE-3616 1993 PRC TIANHUANGPING HYDRO 300.00 0.00 0.00 65.95 50.41 0.00 CN-PE-3623 1993 PRC FINANCIAL SECTOR T.A 0.00 60.00 0.00 18.53 16.33 0.00 CN-PE-3627 1993 PRC GRAIN DISTRIBUTION P 325.00 165.00 0.00 316.84 314.96 94.30 CN-PE-3632 1993 ROC ENVIRONMENT TECH ASS 0.00 50.00 0.00 10.70 11.06 .22 CN-PE-3503 1992 ZOUXIAN THERMAL POWE 310.00 0.00 0.00 7.63 7.64 0.00 CN-PE-3534 1992 PRC ZHEJIANG PROV TRANSP 220.00 0.00 0.00 19.91 19.92 14.52 CN-PE-3564 1992 BEIJING MUNICIPALITY BEIJING ENVIRONMENT 45.00 80.00 0.00 25.28 20.02 12.22 CN-PE-3565 1992 SHANGHAI METRO TRANS 0.00 60.00 1.69 1.32 -. 08 0.00 CN-PE-3568 1992 R.O.C. TIANJIN URB DEV & EN 0.00 100.00 0.00 15.49 12.42 -5.32 CN-PE-3624 1992 MIN. OF PUBL.HEALTH INFECTIOUS DISEASES 0.00 129.60 0.00 31.45 26.83 26.79

Total 14,878.40 4,045.61 239.52 9,907.33 3,050.80 299.73

Active Projects Closed Projects Total Total Disbursed (IBRD and IDA): 8,723.74 12,531.17 21,254.91 of which has been repaid: 73.75 2,539.74 2,613.49 Total now held by IBRD and IDA: 18,610.68 9,624.01 28,234.69 Amount sold : 0.00 0.00 0.00 Of which repaid : 0.00 0.00 0.00 Total Undisbursed : 9,907.33 20.27 9,927.60

a. Intended disbursements to date minus actual disbursements to date as projected at appraisal.

Note: Disbursement data is updated at the end of the first week of the month and is currently as of 30- Apr-99. 138

China STATEMENT O]F IFC's Committed and Disbursed Portfolio As of 31-Mar-99 (In US Dollar Millions)

Committed Disbursed IFC IFC FY Approval Company Loan Equity Quasi Partic Loan Equi Quasi Partic ty 1987/92/94 China Bicycles 8.50 3.39 0.00 0.00 8.50 3.39 0.00 0.00 1993 Shenzhen PCCP 3.76 .99 0.00 0.00 3.76 .99 0.00 0.00 1993 Yantai Cement 13.51 1.95 0.00 7.22 13.51 1.95 0.00 7.22 1994 China Walden JV 0.00 6.00 0.00 0.00 0.00 3.53 0.00 0.00 1994 China Walden Mgt 0.00 .01 0.00 0.00 0.00 .01 0.00 0.00 1994 Dalian Glass 20.50 2.40 0.00 40.50 20.50 2.40 0.00 40.-50 1994 Dynamic Fund 0.00 12.35 0.00 0.00 0.00 10.08 0.00 0.00 1994/97 PTPLeshan 11.20 1.00 0.00 14.00 11.20 1.00 0.00 14.00 1995 Dupont Suzhou 23.36 4.15 0.00 46.80 23.36 4.15 0.00 46.80 1995 Newbridge Inv. 0.00 7.19 0.00 0.00 0.00 3.89 0.00 0.00 1995 Suzhou PVC 22.00 2.48 0.00 22.20 22.00 2.48 0.00 22.20 1996 Beijing Hormel 4.64 .50 0.00 4.95 4.64 .50 0.00 4.95 1996 Jingyang 40.00 0.00 0.00 100.00 40.00 0.00 0.00 100.00 1996 Nanjing Kumho 15.03 3.81 0.00 42.73 12.66 3.81 0.00 35.98 1996 Tianjin Kumho 11.17 0.00 0.00 33.00 0.00 0.00 0.00 0.00 1996 Weihai Weidongri 3.83 0.00 0.00 0.00 3.83 0.00 0.00 0.00 1997 Ningbo 0.00 2.00 0.00 0.00 0.00 2.00 0.00 0.00 1997 Orient Finance 13.33 0.00 0.00 16.67 13.33 0.00 0.00 16.67 1997 PTP Hubei 12.63 0.00 0.00 25.38 10.96 0.00 0.00 22.04 1997 Rabobank PTPC 2.40 0.00 0.00 2.40 0.00 0.00 0.00 0.00 1998 Caltex Ocean 21.00 0.00 0.00 45.00 16.23 0.00 0.00 34.77 1998 Chengxin-IBCA 0.00 .36 0.00 0.00 0.00 0.00 0.00 0.00 1998 Pacific Ports 0.00 3.64 0.00 0.00 0.00 3.64 0.00 0.00 1998 Leshan Scana 6.10 1.35 0.00 0.00 0.00 0.00 0.00 0.00 1998 Rabobank SHFC 2.75 0.00 0.00 2.75 2.25 0.00 0.00 2.25 1998 Zhen Jing 0.00 2.00 0.00 0.00 0.00 2.00 0.00 0.00

Total Portfolio: 235.71 55.57 0.00 403.60 206.73 45.82 0.00 347.38

Approvals Pending Commitment Loan Equity Quas Partic 1998 CHENGDU CHEMICAL 7.40 3.20 0.00 8.60 1997 CHINEFARGE 12.80 0.00 0.00 20.00 1998 EURECA 0.00 3.00 0.00 0.00 1998 JLANGSUCOLINE 6.50 0.00 0.00 0.00 1997 NISSAN/DONGFENG 20.20 0.00 0.00 27.00 1998 ORIENT FIN A INC 3.33 0.00 0.00 0.00 1997 PTP HOLDINGS 0.00 1.50 0.00 0.00 1998 SHANGHAI KRUPP 30.00 0.00 0.00 78.44 1997 SMC 14.00 0.00 0.00 14.00 1996 TIANJIN 9.10 0.00 0.00 9.10 1998 WUHAN CIG 0.00 1.50 0.00 0.00 1998 WUHAN PORT 5.00 0.00 0.00 5.00 1998 XIB 50.00 20.00 0.00 0.00 1998 ZHEJIANG COLINE 6.50 0.00 0.00 0.00 1998 ZHEN JING 4.50 0.00 0.00 0.00

Total Pending Commitment: 169.33 29.20 0.00 162.14 139

Annex 12: Country at a Glance CHINA: Western Poverty Reduction Project

East Lower. POVERTY and SOCIAL Asia & middle- - ______China PacIfic Income Developmentdiamond* 1997 g; :t0;;:t;R 0 Poculaton.mid-vear (millions) 1.227.2 1:75f 2.2&3 Life expectancy GNPoWer caota (Aliasmethod, US$) 860 970 1.230 GNP(Aflas mehod. USS bi0cions) 1.055.4 1.700 2.803 Aore anualarotvh 19922.98

Pooulatiorr(%. l i : 0 0000000j00i;::;j:0 1.0 1.3 1.2 Laborforc#3(%): : ;: :: 0::: i:: 0 :! 1.1 t51 1.5 GNP | Gross : :g;forcek :;0: (::::giE::LaborE per* i primary Mest recent estinate latest vear avallable.199981 i capRa enrollment Povertv (% ef ocomationbelow natlb oemeblinal py tv7 Urbancooulatiion(%oftotal oooulelfon) 33 3 42 Lifeexoectancv at birthfvearsl 70 69 69 Infantmortality oe 1.000live blrthsl 32 37 36 Child malnutrition1% of chidte under-5$) 18 20 17 Access to safewater Acceisslosafe water (% oftbool ) 90 77: 78 illiteraev-%of DoCulatiRnae15+1 ::: : 19 is I8 Gross oninmarvenrgilment (%of school-aceDoovrlitoni 120 118 114 China Male: 121 120 116 Lower-middle-incomegroup iFe,nale ; 00; t120 119 114 KEY;ECONOMICRATIOS and LONG-TERMTRENDS ::;497?:: 1874 lii t1997 1998 Economicratios GDPIUSS billons) 172.3 268.2 902.0 956.1 Grossdomestic investment/GDP 28.5 38.1 38.2 39.0 Trade Exiodts of ooodsand serviceslGDP 4.8 13.6 23.0 21.7 T Grossdornestic savinos/GDP 29,0 36.2 42,7 43.6 T Grossnational savinalGDP 29.0 36.2 39,4 41.7 Currentaccount balance/GDP 0.3 0.1 2.9 2.7 Domesfic ntt InterestDavments/GDP .. 0,4 0.6 0.7: ains nvestment Total debt/GDP 1.t3.2 16.3 18,7 Savings Total debt servce/exDorts .. 9.6 aE 9.9 Presentvalue bf debtlGDPI. .. 14.9 Presentvalue of deWexoorts .. . 863.9.. Indebtedness 19778 71997t98898 198 1999.3 (averaceannual ormffwt GDP: 9.8 10.3 .$8 7.8 7,0 China GNP mocaotta 8.5 8.8 74 6.5 6.1 Lower-middlc-incomegroup Exoortsof coodsand services 18.4 14,6 23.1 4.6 : 4.2:

STRUCTUREof the ECONOMY 1977 1987 1997 1998 Growthrates of output and investment(%) t% of GDP) Aariculture 29.4 26.8 18.7 17.6 T Industrv 47.1 43.9 49.2 49.3 2 Manufacturina 31.1 34.4 37.3 37.0 _ Services 23.4 29.3 32.1 33.1 T Privateconsumotion 63.4 51.3 45.7 43.7 93 94 95 96 97 98 General oavemmentconsumotion 7.5 12.5 11.6 12.8 - GDI -GDP lmoortsofooodsandservices 4.2 13.5 18.5 17.1

1977-87 1988-98 1997 1998 Growthrates of exports and imports (%) (averaae annualarowth) Aoriculture 6.4 4.4 3.5 3.5 40 lndustrv 10.9 14.1 10.8 8.8 30 Manufacturino 11.4 13.4 9.9 7.8 ,0A Services 12.7 8.6 8.3 8.5 10 Privateconsumrotion 9.7 8.6 6.2 5.1 Generaloovemment consumotion 9.0 9.9 8.2 7.8 o Gross domesticinvestment 10.7 11.7 5.6 10.0 93 94 95 96 97 98 lmDortsof aoodsand services 19.2 12.5 12.7 1.5 - Exports -*Imports Grossnational Droduct 10.0 10.0 8.5 7.5

Note: 1998 data are preliminaryestimates. This table was oroducedfrom the DeveloomentEconomics central database. The diamondsshow four kev indicatorsin the countrv (in bold) comoaredwith its income-orouoaveraae. If data are missina. the diamond will be incomolete. 140

China

PRICESand GOVERNMENTFINANCE 1977 1987 1997 1998 Inflation(%) Domestic prices (% change) 30T Consumerprices .. 7.3 7.9 -0.8 2 lmplicitGDP deflator 1.6 5.1 1.2 -1.8

Government finance | _ _ _ _ _ (% of GDP,includes current grants) 93 94 95 Current revenue .. 21.5 12.0 13.1 -101 Currentbudget balance .. 4.2 0.6 -2.2 - GDP deflator o CPI Overallsurplus/deficit .. -2.1 -1.5 -4.0

TRADE (US$millions) 1977 1987 1997 1998 Exportand import levels (USSmillions)

Total exports (fob) .. 39,437 182670 183,610 000 Food .. 4,781 11,054 9,960 Fuel .. 4,544 6,987 2,914 15000 ML- Manufactures .. 26,206 158,767 161,935 Total imports (cif) .. 43,216 142,361 140,226 100o000 Food .. 3,055 6,308 6,119 5,D Fuel and energy 539 10,306 9,275 Capital goods .. 21,110 57,930 58,515 o Export price index (1995=100) .. 75 95 91 Import price index (1995=100) .. 75 89 87 flExports HImports Terms of trade (1995=100) .. 99 107 105

BALANCE of PAYMENTS

(US$ millions) 1977 1987 1997 1998 rCurrent account balance to GOP ratio%)

Exports of goods and services 8,550 39,120 207,253 207,140 4 Imports of goods and services 8,148 38,880 166,755 163,266 { Resource balance 402 240 40,498 43,874 2 r

Net income 82 -164 -15,900 -19,842 o E.r iI Net currenttransfers 0 249 1,832 2,163 82 9469 96 97 98 Current account balance 484 325 26,430 26,195 I2 Financing items (net) -390 4,468 9,427 -22 662 Changes in net reserves -94 -4,783 -35,857 -3,533 -4

Memo: Reserves including gold (US$ millions) .. .. 143,363 Conversion rate (DEC, /ocaV/US$) 1.9 4.5 8.3 8.3

EXTERNAL DEBT and RESOURCE FLOWS 1977 1987 1996 1997 (US$ millions) Composition of total debt, 1997 (USS millions) Total debt outstanding and disbursed .. 35,340 128,817 146,697 IBRD .. 1,427 7,616 8,239 IDA .. 1,330 7,579 7,830 A: 8,239 oG 31,494 B: 7,B30 Total debt service .. 3,852 15,756 18,445 D: 2,904 IBRD .. 208 840 858 IDA 12 73 81 20,782 E

Composition of net resource flows Official grants 209 249 228 Official creditors .. 626 4,401 4,315 Prvate creditors .. 5,462 6,454 8,134 Foreign direct investment .. 2,314 40,180 44,236 Portfolio equity .. 0 3,466 8,457 F: 75,478

World Bank program Commitments .. 1,306 1,900 2,425 A - IBRO E - Bilateral Disbursements .. 702 2,097 2,275 B- IDA D - Other multilateral F - Private Principal repayments .. 97 364 377 C - IMF G - Short-term Net fows .. 605 1,734 1,898 Interest payments .. 124 549 562 Nettransfers .. 482 1,185 1,335

Note: This table was produced from the Development Economics central database. 4/27/99 IBRD 29993

CHINA WESTERNPOVERTY REDUCTIONPROJECT QINGHAIPROVINCE QILAN

MOVEOLT AREASMAGAC MOVEIN AREAS DLNH

PROVINCEROADS ANN S./ T 1 I NATIONALROADS 4 5 L (@ PROVINCECAPITAL ACCH NATIONALCAPITAL 4 COUNTYSXAN) BOUNDARIES -1 PREFECTUREBOUJNDARIES I7 PROVINCEEOUNDARIES

PRO,JECICOUN,rIES: M(A~SXiNN ANHLT DATONG -IAOIOI

HNOVJN~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~IN

PEDEALIONG ,\

MON NOlAN~~ DULA

8~~, -- "<"~ A"'-"--- 0 0 D~~~.ig~~UO,~T?MW~~ddO~0Z XlA/TI

'-I'PP0PI'll FEDERATION~~ ~~~ ~~ ~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~A'IO IBRD29994 ,__96, I.-98t 0° 102- 104° 106 1 410 ; Pt v __ _ S - , - _ _- -MONGOLIA - N ~~~~~~~~~~CHINA vs K -*-*% %._ %,--'j o ~5 WESTERNPOVERTY REDUCTION PROJECT GANSUPROVINCE 425 < PROJECTCOUNTIES PROJ(CTCOUl,I ES: o.o7Jf : (Ad iood5 f A NJA?~~~~~~~~~~~~~~~~~~~~~~ bodes t~~~~~~~~~~~~~~~~~~~~HANC;.ACHUAN OIN AN ,4va. \ . PROVINCEROADS

-P. - > V ' .lROADS t,, ,s,_NATIONAL OANU :tt tLL)ttl:& ~~~~~~~~~~PROVINCECAPITAL WUSrIAN COUNTY(XIANI BOUNDARIES LIN XIAJ\

PREFECTUREBOUNDARIES 'INE PROVINCEBOUNDARIES AOIA

/DUNI-IIJAhGICJ , YuGU\ .. . ox.:I ;,. . . -- INTERNATIONALBOUNDARIES WONXSIAN DANGOLANC

%~~~, ' ,.LNS /t>95

AKESAI USUBID. Lrx 9 N vL & MBINQIN ,5 t3 t__4.oy~A,,,9' < 3 , tNLTCANG

-- Hv ^>? 3, ... '\~~~~~~~~~~~~~~~~~~~~~~~~~~~~CW , G . ..,. .... ,,l...... ,,,0 ; I I55 105I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~IN ,1C

_3D0MSTOROtL10 130 ^ t ,{ \! JL#INAt, ToTrochur,T. 110'

6 Thobcodzi,do,6iltlnnt A efornol:Ms¢- X005. .o.oE,..tloo 0sto TOO ?50g 6 RUSSIANI I B"P.b,;:..!rba U.SY/U2fOhG 3YUAN XIFLNC' SING

, ,._ FERDUEsRATION' < AI ~?a+

MONOORA f 2N<'f-">gh Xk < t\X-' Zz>*ai ;0m ,_ FEDERATIONOOD "on ,RP IthI I>~'so~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~ ~ ~~~~~~~~~~~~~~~~~~~~~~~G S } ~~~~~~t - hengd0t12 0vT0tI8 i. * NorioroICo~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~XJf,~A,NI IJ HUININt*UA$4. 7 =1P.DRjs' ..( 04~~~~~~~-6 NV~~lu H

0 .5DB 5$ P1- 12 29995 120 IBRD 112 116' 08, 1 1W FEDERATION RUSSIAN hilool CHINA MONGOLIA gel? REDUCTIONPROJECT WESTERN POVERTY 52' REGION NEIMONGOL AUTONOMOUS RUISSIAN PROJECTCCUNTfES/BANNERS FEDERATION!i PROVINCEROADS NATIONALROADS RJ PROVINCECAPITAL CHENRA'Eki (XIAN)BOUNDARIES COUNTY YAKESH(l lv BOUNDARIES AP PREFECTURE H PROVINCEBOUNDARIES INTERNATIONALBOUNDARIES

'ERHU s, PROJEC'fCOUNTIES/BANNERS: 4 AO,: [AIPUSI DA'FRANMAOMINGAI\ DUCLUN SIZIWANG KESI- IKETENG You =F ICE CI-IAHAlFR-YCUYlZHC)NG LINXI GUYANC SALIN-YOU WLJCHUAIII -% WENGNJTE (,HAHA'FR-YC)UYlHCU WILA 44' IONG SHANGDUHUADE, KE'ERQ(N-Y0UYlI rJLOMETERS ...... ghk Z i-6&iNtJOLl-

XIWUZHUMUQIN ABAGA SUNIFIZUO 11 MONGOLIA ERLIANHAOTE 44- XILiMq N 'I K'NG SUNli-,-

EXNA gn VVULA WUL4TEHOU ZH T Sb-y.,,g WUYUAN HANGJIN-HOU,-' 9 T. Ji,.I,.,, ALASHAN-YOU DEN KO CIAN WULATE- HAER YOUYICAN HAN AtAiHAN-Zuo XINGHE DQIGSHENGIDAI EN 40' WUHAI YUIN ETLJOKE T. UIHE NESFO T. Yli SHANXI

12.4' 120' II& T12- MAY1999