Document of The World Bank Report No: ICRR00001574 IMPLEMENTATION COMPLETION AND RESULTS REPORT (Cr. 37490) ON A CREDIT IN THE AMOUNT OF SDR 82.20 MILLION (USD 112.56 MILLION EQUIVALENT) TO THE REPUBLIC OF INDIA FOR A CHHATTISGARH DISTRICT RURAL POVERTY PROJECT April 21, 2011 Sustainable Development Sector Unit Agriculture and Rural Development India Country Management Unit South Asia Region INDIA Chhattisgarh District Rural Poverty Project IMPLEMENTATION COMPLETION AND RESULTS REPORT APRIL 21, 2011 CONTENTS Data Sheet ……………………………………………………………………………………………………….i-vii A. Basic Information 4 B. Key Dates 4 C. Ratings Summary 4 D. Sector and Theme Codes 5 E. Bank Staff 5 F. Results Framework Analysis 6 G. Ratings of Project Performance in ISRs 9 H. Restructuring (if any) 9 I. Disbursement Profile 9 1. Project Context, Development Objectives and Design 10 2. Key Factors Affecting Implementation and Outcomes 12 3. Assessment of Outcomes 18 4. Assessment of Risk to Development Outcome 24 5. Assessment of Bank and Borrower Performance 25 6. Lessons Learned 27 7. Comments on Issues Raised by Borrower, Implementing Agencies and Partners 28 Annex 1. Project Costs and Financing 30 Annex 2. Outputs by Component 31 Annex 3. Economic and Financial Analysis 33 Annex 4. Bank Lending and Implementation Support/Supervision Processes 40 Annex 5. Impact Evaluation Results 42 Annex 6. Summary of Feedback from Beneficiary Workshops 47 Annex 7. Summary of the Borrower’s Completion Report and Comments on the Draft ICRR 50 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders 62 Annex 9. List of Supporting Documents 63 CURRENCY EQUIVALENTS (Exchange Rate Effective March 31, 2010) Currency Unit = Indian Rupee (INR) INR 1.00 = USD .0222 USD 1.00 = INR 45.0 FISCAL YEAR (July 1 – June 30) ABBREVIATIONS AND ACRONYMS BPL Below Poverty Line MIS Management Information System BRGF Backward Region Grant Fund MP Madhya Pradesh CAS Country Assistance Strategy MTR Mid-term Review CDD Community Driven Development NGOs Non-Government Organizations CEO Chief Executing Officers NREGA National Rural Employment CIG Common Interest Group Guarantee Scheme DCA Development Credit Agreement NRLM National Rural Livelihoods Mission EA Environmental Assessment NTFP Non Timber Forest Produce EFA Economic and Financial Analysis O&M Operation and Maintenance EPVG Extremely Poor and Vulnerable Groups PAD Project Appraisal Document ERR Economic Rate of Return PDO Project Development Objective DCA Development Credit Agreement PFT Project Facilitation Team DPIP District Poverty Initiatives Project POP Poorest of the Poor EMF Environmental Management Framework PPR Project Procurement Report FMR Financial Management Report SC Scheduled Caste GoCG Government of Chhattisgarh SGSY Swarnajayanti Gram Swarozgar GP Gram Panchayat Yojana GS Gram Sabha SHGs Self Help Groups HRD Human Resources Development SPMU State Project Management Unit IE Project Impact Evaluation SRI System of Rice Intensification INR Indian Rupees ST Schedule Tribe KPI Key Performance Indicators TTL Task Team Leader LF Logframe M&E Monitoring and Evaluation Vice President: Isabel Guerrero Country Director: Roberto Zagha Sector Manager: Simeon K. Ehui Project Team Leader: Samik Sundar Das ICRR Team Leader: Jim Hancock, FAO A. Basic Information Chhattisgarh District Country: India Project Name: Rural Poverty Reduction Project Project ID: P076467 L/C/TF Number(s): IDA-37490 ICR Date: 04/25/2011 ICR Type: Core ICR GOVERNMENT OF Lending Instrument: SIL Borrower: INDIA Original Total XDR 82.2M Disbursed Amount: XDR 36.6M Commitment: Revised Amount: XDR 36.6M Environmental Category: B Implementing Agencies: Office of Project Director, Chhattisgarh District Poverty Reduction Project Cofinanciers and Other External Partners: B. Key Dates Revised / Actual Process Date Process Original Date Date(s) Concept Review: 05/16/2002 Effectiveness: 11/13/2003 11/13/2003 Appraisal: 12/18/2002 Restructuring(s): Approval: 04/24/2003 Mid-term Review: 02/12/2007 07/02/2007 Closing: 03/31/2009 03/31/2010 C. Ratings Summary C.1 Performance Rating by ICR Outcomes: Moderately Unsatisfactory Risk to Development Outcome: Substantial Bank Performance: Moderately Satisfactory Borrower Performance: Moderately Satisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Bank Ratings Borrower Ratings Quality at Entry: Moderately Satisfactory Government: Moderately Satisfactory Implementing Quality of Supervision: Moderately Satisfactory Moderately Satisfactory Agency/Agencies: Overall Bank Overall Borrower Moderately Satisfactory Moderately Satisfactory Performance: Performance: i C.3 Quality at Entry and Implementation Performance Indicators Implementation QAG Assessments Indicators Rating Performance (if any) Potential Problem Project Quality at Entry No Satisfactory at any time (Yes/No): (QEA): Problem Project at any Quality of Yes None time (Yes/No): Supervision (QSA): DO rating before Moderately Closing/Inactive status: Satisfactory D. Sector and Theme Codes Original Actual Sector Code (as % of total Bank financing) General agriculture, fishing and forestry sector 50 50 Other social services 20 20 Sub-national government administration 30 30 Theme Code (as % of total Bank financing) Indigenous peoples 11 11 Other social protection and risk management 22 22 Participation and civic engagement 22 22 Rural policies and institutions 23 23 Rural services and infrastructure 22 22 E. Bank Staff Positions At ICR At Approval Vice President: Isabel M. Guerrero Mieko Nishimizu Country Director: N. Roberto Zagha Michael F. Carter Sector Manager: Simeon Kacou Ehui Constance A. Bernard Project Team Leader: Samik Sundar Das Luis F. Constantino ICR Team Leader: Samik Sundar Das ICR Primary Author: Jim Hancock F. Results Framework Analysis Project Development Objectives (from Project Appraisal Document) To improve opportunities for the poor and vulnerable, especially women and tribals, to meet their own social and economic development objectives. To achieve this objetive, the project would: (i) create infrastructure and income opportunities for the rural poor; (ii) ii empower active groups of disadvantaged people; and (iii) support village governments in becoming more responsie and effective in assisting them. Revised Project Development Objectives (as approved by original approving authority) (a) PDO Indicator(s) Original Target Formally Actual Value Values (from Revised Achieved at Indicator Baseline Value approval Target Completion or documents) Values Target Years Indicator 1 : Increase in Income of project beneficiaries A range from Value Increase by various estimates is quantitative or Not estimated None given 50% in real of aound 30% Qualitative) terms at MTR increase on average Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments (incl. % The indicator was achieved achievement) Indicator 2 : Reduction in migration among beneficiaries Anecdotal information that Value there is reduced quantitative or None Given No target set N/A migration change in Qualitative) labour patters to more local self employment. Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments Preliminary result from Impact Evaluation indicated no major change in (incl. % migrations pattern. achievement) Indicator only partially achieved. Participation of disadvantaged people, particularly women and tribals, in Indicator 3 : subproject activities and decision-making (PAD text) (in village processes, logframe) Qualitative assessment shows women and tribal Value groups were quantitative or Low participation No value given n/a involved in the Qualitative) project identification, planning and implementation Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments Indicator difficult poorly defined and not quantified. (incl. % Achieved iii achievement) (b) Intermediate Outcome Indicator(s) Original Target Actual Value Formally Values (from Achieved at Indicator Baseline Value Revised approval Completion or Target Values documents) Target Years Indicator 1 : Number of sub-project submitted, funded and implemented (LF only) Value (quantitative NA 20,000 N/A 20,446 or Qualitative) Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments (incl. % Target Achieved achievement) Indicator 2 : Number of families benefited (PAD text, and LF) Value (quantitative NA 150,000 N/A Over 150,000 or Qualitative) Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments The result takes into account possible overlap between the infrastructure and CIG (incl. % activity beneficiaries. achievement) The indicator was achieved Average cost of subprojects compared to government standards (PAD text) Indicator 3 : Costs of sub-projects by category (LF) Infrastructure costs Not quantified. Value estimated as largely Assumed lower (quantitative N/A NA similar. CIG than government or Qualitative) subprojects cost standards slightly lower. Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments Based on anecdotal information. (incl. % Target was achieved, but just. achievement) Indicator 4 : Percentage of CIG members belonging to poor and vulnerable groups (only LF) Value Women 30% Women 23% (quantitative NA SC (14%) NA SC (14%) or Qualitative) ST (35%) ST (37%) Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2009 Comments Comparison had to be adjusted to individual districts proportions of SC/ST. (incl. % Result of detailed analysis showed slight positive targeting. Indicator achieved. achievement) Indicator 5 : Performance of para-professionals supporting CIGs and GPs (only in LF). Value (quantitative N/A None indicated NA None or Qualitative) Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 iv Comments (incl. % This activity was not pursued by the Project. achievement) Number of beneficiaries with access to banking system (PAD text); Indicator 6 : Percentage of villagers with Bank accounts (LF) Value Has increased. (quantitative 12% 20% n/a Details in impact or Qualitative) evaluation. Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments (incl. % This target could have been dropped. achievement) Community contributions in cash (PAD text) Indicator 7 : Community contributions in village funds (LF, inputs) Rs 17.83 crore 6.49 million USD Value (3.56 million USD) (but implicitly (quantitative 0 No revision Communities did based on 10% of or Qualitative) contribute 10% of subproject costs) subproject costs. Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments 55% of target was achieved. However if this had been based on the 10% of the (incl. % project cost that was contributed by beneficiaries, the target was achieved. achievement) Number of GS and GPs practicing participatory budgeting (PAD text) Indicator 8 : Percentage of GS and GPs practicing participatory budgeting (LF) Value (quantitative Not given No target set n/a NA or Qualitative) Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments Emphasis was not put to related activities or monitoring of the target. Nor was (incl. % indicator cancelled. achievement) Target was not achieved. Improved people#s perceptions of Gram Sabhas (GS) and Gram Panchayats Indicator 9 : (GPs) (PAD text) - Peoples, and in particularly the poor#s perception of GS and GPs (LF) Some increased perception in Value improved access to (quantitative Baseline No target set n/a GS and GPs, but or Qualitative) also possibly due to other programs. Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 Comments (incl. % Target no achieved. achievement) Increase in access to clean drinking water by the poorest habitations in each Indicator 10 : Panchayat. (PAD text); Percentage of poorest habitations with clean drinking water (LF) v Impact Evaluation Value notes no difference (quantitative No value No target set NA in access between or Qualitative) Project and control areas. Date achieved 12/30/2006 12/30/2006 02/12/2007 03/31/2010 This was condition for community infrastructure, but was not implemented, as Comments other government programmes implemented drinking water facilities in a wide (incl. % range of villages. However target was not modified. achievement) Target was not achieved. G. Ratings of Project Performance in ISRs Actual Date ISR No. DO IP Disbursements Archived (USD millions) 1 06/17/2003 Satisfactory Satisfactory 0.00 2 12/04/2003 Satisfactory Satisfactory 0.00 3 06/25/2004 Satisfactory Satisfactory 5.00 4 12/28/2004 Satisfactory Unsatisfactory 5.00 5 06/07/2005 Satisfactory Unsatisfactory 6.97 6 12/26/2005 Satisfactory Satisfactory 8.99 7 06/21/2006 Satisfactory Satisfactory 11.53 Moderately 8 12/29/2006 Moderately Satisfactory 22.68 Unsatisfactory 9 06/23/2007 Moderately Satisfactory Moderately Satisfactory 25.47 10 10/05/2007 Moderately Satisfactory Moderately Satisfactory 32.37 Moderately 11 06/22/2008 Moderately Satisfactory 45.74 Unsatisfactory Moderately 12 12/29/2008 Moderately Satisfactory 50.74 Unsatisfactory 13 03/04/2009 Moderately Satisfactory Moderately Satisfactory 52.61 14 08/17/2009 Moderately Satisfactory Moderately Satisfactory 55.51 Moderately 15 04/03/2010 Moderately Satisfactory 57.59 Unsatisfactory H. Restructuring (if any) Not Applicable vi I. Disbursement Profile vii 1. Project Context, Development Objectives and Design 1.1 Context at Appraisal 1.1.1 At the time of Project preparation, Chhattisgarh was a new State that had recently separated from the State of Madhya Pradesh (MP) in November 2000. Chhattisgarh had a higher incidence of poverty, scheduled tribes (STs) and scheduled castes (SCs) than did MP, itself one of the poorest states in India. As elsewhere in the country, the incidence of poverty in Chhattisgarh was greater in more remote rural, non-irrigated and forested areas, and among STs, SCs and the landless. Some 83% of Chhattisgarh’s population was engaged in agriculture and allied sectors (Census 2001), and 43% lived below the poverty line (BPL), of whom nearly three-fifths were STs and SCs. Insurgent movements were active, particularly in the heavily tribal Northern and Southern regions. Hence, strengthening the productive capacity, incomes and well-being of rural households, addressing sources of conflict, and creating and strengthening the institutional architecture necessary to administer the State were among the highest priorities of the new Government of Chhattisgarh (GoCG). 1.1.2 India had had a long history of implementing rural poverty reduction programs, but these seldom had the desired outcomes, often due to proliferation of uncoordinated interventions, poor targeting, weak engagement and ownership by intended beneficiaries, insufficient transparency, sustainability issues, and inefficient administration. The new GoCG wanted to improve this record and requested the World Bank’s support to develop a rural community-driven development (CDD) project using more innovative methods of addressing needs of the poor and strengthening local governments, modeled on a similar Bank-supported project which had recently been launched in MP.1 1.1.3 The conditions in Chhattisgarh and the request from GoCG met the core concerns of the World Bank Country Strategy for India (CAS R2002-0203), which highlighted the need to develop well-targeted programs to assist the poor through delivery mechanisms that would promote decentralization and strengthen government effectiveness. The CAS also emphasized support to states with high poverty, good potential for demonstration effects, and where state governments had shown commitment to addressing poverty and exclusion issues. Although still a very young state, Chhattisgarh had already launched several of such initiatives, including pensions for destitute women and a policy on women’s empowerment. 1.1.4 The resulting Chhattisgarh District Rural Poverty Project was intended to improve the incomes, well-being and empowerment of the rural poor and excluded people by helping them form common interest groups (CIGs) to further their own development. It would also strengthen the capacity of local Panchayat governments (GPs) to engage in participatory planning and efficient implementation of small pro-poor investments. Project facilitation teams (PFTs) would carry out detailed consultations and assessments as the first activity when starting in a new village, to identify the potential target population and understand their key issues. The Project would provide matching grants to both the CIGs and GPs. CIG members would participate in the identification, planning, implementation, operation and maintenance of sub-projects, and would manage their own Project bank accounts. GPs would use participatory methods to prepare Village Development Plans, and then would also manage Project funds for their implementation. 1 Madhya Pradesh District Poverty Initiatives Project (Cr. 3430-IN) and related projects in Andhra Pradesh (Cr 3232- IN), and Rajasthan (Cr.3339) 10 1.2 Original Project Development Objectives and Key Indicators 1.2.1 The Project Development Objective (PDO) in the Project Appraisal Document (PAD) was “to improve opportunities for the poor and vulnerable, especially women and tribals, to meet their own social and economic development objectives”. To achieve this objective, the Project would: (i) create infrastructure and income opportunities for the rural poor; (ii) empower active groups of disadvantaged people; and (iii) support village governments in becoming more responsive and effective in assisting them. 1.2.2 The three key PDO indicators from the logframe 2 were: (i) increase in income of project beneficiaries; (ii) reduction in migration among beneficiaries; and (iii) participation of disadvantaged people, particularly women and tribals, in sub-project activities and decision- making. A further 9 intermediate outcome indicators3 would measure: (i) share of CIG members belonging to poor and vulnerable groups; (ii) the number of families benefitted; (iii) average cost of sub-projects compared with government standards; (iv) share of CIG members belonging to poor and vulnerable groups; (v) number of beneficiaries with access to banking system and obtaining credit; (vi) community contributions in cash: (vii) number of Gram Sabhas (GSs) and GPs practicing participatory budgeting; (vii) improved perceptions of GSs and GPs; and (ix) increase in access to clean drinking water by poorest habitations in each Panchayat. 1.3 Revised PDO and Key Indicators N/A 1.4 Main Beneficiaries 1.4.1 The Project would provide direct benefits to about 150,000 rural families (750,000 people), from among the most socially and economically disadvantaged population in 2,000 villages. They would become direct beneficiaries through one, or both, of two modes: (i) self-selection by eligible people who would opt to form common interest groups (CIGs) to undertake collective, beneficiary led sub-projects. Eligibility criteria for participation in CIGs would include all SCs and STs; households with migrant wage earners; households without permanent dwellings or whose houses needed repair; women and female-headed households; landless and marginal farmers; and the lowest 70% of households ranked in order of poverty by the GSs in a participatory manner; and/or (ii) by being among those households impacted by infrastructure sub-projects undertaken by the GPs as part of Panchayat development plans that would be prepared under the Project. 1.4.2 As the Project included components on strengthening GPs, GSs and implementing agencies, by implication indirect benefits would also accrue to local governments, public officials and private NGOs trained in pro-poor development methods, and to the State government more generally to the extent that its rural development program effectiveness was improved. 2 There are some slight differences in the phrasing and ordering of indicators between the main text and the logframe of the PAD. See also Section 3.2. 3 Logframe also listed an indicator on use of para-professionals, but this is not mentioned elsewhere in PAD or in any implementation documents; this ICRR therefore treats the reference as a typo error in the logframe. 11 1.5 Original Components: The Project had two components: 1.5.1 Component (1): Institutional and Human Capacity Building through: (a) formation and strengthening of organizations to build the capacity of NGOs and Project Facilitation Teams (PFTs) in supporting CIGs; (b) communications and information; (c) human resource development for project stakeholders, (d) strengthening of Panchayats; (e) monitoring and learning and special studies; and (f) Project administration. 1.5.1 Component (2): Community Investments including: (a) matching grants for community investment sub-projects to finance collective income-generating activities; (b) matching grants for Panchayat plans to finance investment sub-projects in village infrastructure; (c) village funds financed by community contributions and to support operation and maintenance of village infrastructure, and further village development beyond the life of the Project; and (d) an innovations fund to provide small grants for demonstration sub-projects. 1.6 Revised Components N/A 1.7 Other Significant Changes 1.7.1 The main modification occurred in mid-2005, when SDR 13.26 million (USD 20.06 million equivalent) was cancelled from this credit and transferred to finance the India Tsunami Emergency Reconstruction Project.4 Later, in 2009 the IDA Credit was extended by one year to March 31, 2010, and some minor changes were made in component allocations. At closing, there was a final cancellation of SDR 32.37 million (USD 49.05 million equivalent). 2. Key Factors Affecting Implementation and Outcomes 2.1 Project Preparation, Design and Quality at Entry 2.1.1 Strengths: The Project was closely aligned with the priorities of the central government of India at the time, which was partnering with the World Bank on a similar cohort of rural CDD projects in other states (see footnote 1); with those of the new GoCG; and with the World Bank’s CAS for India. The goal of reducing rural poverty and exclusion has remained high on the agendas of all three throughout the life of the Project. The Project also had several important design features that enabled it to achieve a substantial part of its core objectives.  The design drew on both the World Bank’s global experience with rural CDD projects, and on specific design experience of a similar operation in MP, the State from which Chhattisgarh had recently separated but with which it nonetheless had a number of characteristics in common.  As part of project preparation, implementation arrangements were pilot tested by 120 CIGs and for PFTs in the district of Bilaspur.  The design was kept relatively simple both in terms of number of components and implementation responsibilities. Design of the Chhattisgarh Project focused heavily on the first stages of CDD at the community and local government level. This was appropriate and reflected the relative newness of the State, its more limited institutional capacity, and the 4 Cr. 4054-IN of May 12, 2005. 12 special challenges of building ownership and fostering cohesion between citizens and local governments in a conflict-affected setting.  Emphasizing the role of local (as opposed to the state) government structures was appropriate as these district and block entities had been carried over from MP largely intact, and were relatively more stable compared to the new GoCG state-level structures. Capacity building at local government and community levels was also consistent with GoCG’s stated decentralization objectives.  Putting funds for CIG sub-projects under the direct control of beneficiaries to finance activities that they themselves had selected as priorities was appropriate for promoting empowerment, transparency and building ownership.  The risk of weak government institutions and NGOs, and the inexperience of future beneficiaries in playing an active role in implementing development programs were correctly identified. Provision for extensive capacity building aimed to address this. Shortcomings: Project design also had several weaknesses, related mainly to its scale.  Although the risk of weak institutional capacity had been identified, the challenges were even greater than recognized. As a new State, Chhattisgarh was still in the process of creating and consolidating core institutions and the availability of suitable staff in government and NGOs – both to participate in training and engage in implementation, especially for technical aspects – was overestimated. Similarly the need for more in-depth community mobilization and strengthening was underestimated, especially when targeting tribal communities and the extreme poor. Finally, assumptions about the extent to which GPs could be transformed into more participatory entities through training alone were not realistic; to fully achieve expected outcomes would have required complementary policy and institutional arrangements to empower these institutions – an agenda that went beyond the scope of the Project.  The extent to which political instability would affect implementation was not fully appreciated at the design stage. Elections and changes of administration slowed the project, both at its launch and several times throughout the project period, and mounting insurgency limited the extent to which the Project could function in some areas.  Sub-project costs were over-estimated. The INR 30,000 per household ceiling was high considering the extent of poverty and financially overburdened the beneficiaries (CIG members were expected to contribute 5% to sub-project costs and 10% to village funds). The costs of the sub-projects exceeded those in the parallel project in MP.5  Key Project indicators lacked baseline figures, some of which might have been estimated from the parallel MP project and Bilaspur pilot experiences, and later adjusted after carrying out a full Project baseline survey. Some targets were adopted without sufficient consideration as to how feasible they would be to monitor (e.g. reducing migration). 2.1.2 The Project was reviewed by the Bank’s Quality Assurance Group (QAG), and received a Satisfactory rating for quality at entry, although the review did flag some concerns about its overall size. On balance, the Project responded to important, shared strategic development priorities; the basic Community Driven Development (CDD) design was appropriate; and the objective of enhancing incomes and improving the well-being of 150,000 poor and excluded rural households, and increasing their level of participation in Project activities, would prove feasible. However, this would be achieved more slowly than expected and with lower inputs per beneficiary in terms of both USD invested and technical assistance and training provided (albeit 5 This was partly due to feedback to the design team from MP counterparts that unit allocations in their project were proving to be low. 13 at a pace, and at levels, probably more commensurate with conditions in the State over the life of the Project). 2.2 Implementation 2.2.1 Implementation began more slowly than expected. The Project launch was delayed by about a year due to local and national elections, the former resulting in a change of political party and therefore of many state officials who had been engaged in preparation (on the positive side, the first Project Director needed to re-establish key relationships and understanding about Project objectives and how it would operate). There were also delays in filling critical vacancies. Unlike many CDD operations that operate largely outside the mainstream of government structures, in this case it was planned to fill only about 20% of positions through recruitment from Non- governmental Organizations (NGOs) and other private sources, and to redeploy the majority (80%) from other government positions. This was necessary because of the critical shortage of qualified individuals in the State; it would also help strengthen the State’s institutional capacity by allowing staff to gain experience working on the Project. Nonetheless, it still proved difficult to find adequate numbers of qualified individuals, and training institutions which were meant to be engaged to build capacity were themselves in the process of being established, especially the State Institute for Rural Development (SIRD). 2.2.2 When implementation continued to lag after the June 2004 Project launch, IDA declared Implementation Progress (IP) unsatisfactory and spent the following 18 months or so making frequent (often monthly) supervision inputs to turn the situation around. This met with some success, as CIG formation, Panchayat planning and sub-project implementation picked up on the ground. Disbursements also began to pick up and by the time of the mid-term review (MTR) in mid-2007, information available from the baseline survey, MIS and beneficiary interviews led the MTR team to positive conclusions about the likelihood of achieving the PDO. Some USD 20 million equivalent had been canceled and reallocated to fund part of IDA’s Tsunami response, and the MTR considered that the remaining credit could be fully utilized and most key indicators achieved, with some extension of the Closing Date to compensate for the slow Project start up. 2.2.3 Factors that challenged implementation throughout the life of the project included: delays associated with further local and national elections; gradual mounting of insurgency activities in parts of Chhattisgarh; turnover of staff deputed from other government agencies, as incentives to encourage them to stay and perform (e.g. longer contracts and allowances) were not applied; and turnover of Project Directors. The last also reduced continuity and contributed to rigidities in the application of project criteria and procedures that was sometimes inconsistent with the intended Project ethos which would have allowed more flexibility (e.g. the number of members per CIG). Elsewhere these factors would probably have constituted teething problems, but a new State like Chhattisgarh struggled with them for most of the Project life, and this is clearly reflected in IDA’s biannual Implementation Supervision Report (ISR) ratings and Aide-Memoires. 2.2.4 Despite these difficulties, the response by rural communities to a Project that allowed them to engage in decision-making and address critical income and other local needs, combined with growing ownership and support from some key senior officers and the efforts of some very dedicated though overstretched staff, made it possible to achieve a number of key targets. Towards the end of the Project, and with an agreed one-year extension of the Closing Date until the end of March 2010, GoCG and IDA put considerable efforts into ensuring completion of sub- projects and associated financial management records, ensuring sustainability of sub-project investments, and strengthening market linkages to consolidate beneficiary income gains (e.g. support in forming federations of CIGs; piloting adoption of System Rice Intensification, or SRI, 14 which is now being incorporated in the State’s own extension system; piloting village level institutions, etc.). As the extended Closing Date approached, since undisbursed credit funds were available and a number of these initiatives were still in early stages of implementation, GoCG and IDA teams discussed the possibility of a further extension but decided instead to plan for a transition to the GoI’s National Rural Livelihood Mission (NRLM) that was under preparation at 6 the time, with a focus on the poorest states, such as Chhattisgarh. The World Bank’s National Rural Livelihood Project (NRLP) that supports NRLM, currently under preparation and scheduled for Board presentation in July 2011, does include Chhattisgarh as one of its target states. 2.3 Monitoring and Evaluation Design, Implementation and Utilization 2.3.1 The M&E design described in the PAD was comprehensive, covering basic monitoring, baseline and end-of-project evaluations, performance tracking, institutional tracking, community self-assessment, and learning processes, but overall implementation was fragmented and systems came late into place. A Baseline study that was contracted out was of acceptable quality, although it was not completed until just before the MTR. Moreover, due to lack of understanding and capacity issues the findings from the Baseline study could not be institutionalized into M&E system, thus performance tracking on results indicator was hardly evident. Annual reporting was often late. The ICRR team saw little evidence of concerted performance tracking. Basic monitoring and data collection were generally in place locally in paper form (and using basic spreadsheet software), but an attempt to establish a computerized web-based management information system was long and drawn out, and ICRR did not provide reliable outputs by ICRR mission. A community institution grading system was piloted and generated some useful learning. Overall, the low number of qualified State level specialists, and their high turnover made it difficult to develop a cadre of Project officers who would appreciate the value of M&E as a management tool for decision-making, although some of the Project Directors appreciated this in the later stages of the Project. 2.3.2 The end-of-project Impact Evaluation (IE) was initiated in May 2010 by an experienced survey team, but it faced considerable difficulty in matching survey households with those surveyed at baseline. The results therefore compared some impact indicators in control and project households for a limited sample group. The rigorousness of the analysis of attribution thus had limitations (summary in Annex 5). For purposes of this ICRR, the mission utilized a range of sources, including interviews at State and field level, stakeholder workshops (Annex 6) and individual household interviews, an economic and financial assessment (EFA) based on a survey of a representative sample of 150 randomly selected CIGs (Annex 3), and review of several thematic studies (Annex 9). 2.4 Safeguard and Fiduciary Compliance 2.4.1 Procurement: Overall, procurement performance was Moderately Satisfactory. The Project had quite a good start because of the presence of a competent Procurement Officer trained in the World Bank Procurement Guidelines, who remained in place until February 2008.7 This Officer’s role was critical in ensuring that initiation and implementation of community infrastructure and CIG sub-projects were in compliance with the Bank guidelines and procedures. 6 NRLM is a reformulation and substantial improvement of the Government of India’s earlier SGSY (Swarnajayanti Gram Swarozgar Yojana), with a focus on rural income generation. . 7 In its Quality at Entry Assessment, the QAG panel rated preparation and readiness of fiduciary arrangements (both procurement and financial management) as Highly Satisfactory. 15 However, performance from early 2008 through Project completion was less satisfactory due to lack of adequate procurement capacity and continuity of key dedicated staff at State level. Some actions identified in supervision Aide Memoires were not followed up (e.g. the need to modify the Operational Manual to reflect the shopping procedure; AM of May-July 2008). Procurement issues in these final years had an impact on Project disbursements, and delayed actions relating to federating CIGs and strengthening market linkages. Project Procurement Reports (PPRs) confirmed the presence of assets purchased by CIGs, but identified a lack of documentation and implementation issues at CIG level, and some procurement that was not in compliance with the Operational Manual. However, follow-up was difficult due to procurement capacity constraints. 2.4.2 Financial Management: Financial management performance was Moderately Satisfactory. Unlike procurement, implementation of financial management procedures was weaker in the initial period because of capacity constraints, including prolonged vacancies in key staff positions; delays in installation of the computerized financial management system which resulted in manual record keeping and accounting; and delayed submission of Financial Management Reports, FMRs). Remedial measures were taken to improve this situation, including appointment of adequate staff, training of all Project finance staff, and installation of an off the shelf accounting package ‘Tally’ for computerizing project records. The Project also implemented sample auditing of completed sub-projects to ensure accountability. The FMRs were directly generated from Tally and submitted to IDA on time. Almost 50% of sub-projects were initiated during the last two years of the Project, in order to ensure that they were suitably closed in the books, the focus of the finance team shifted to ensuring that sub-projects were in fact completed, and completion certificates obtained and entered in accounting system. Due to these extensive efforts, the Project managed to close successfully and obtained clearances for almost 98% of all sub-projects. A very considerable share of SPMU staff effort went into this work during the final period of the Project. 2.4.3 Environmental Management: Overall, compliance with environmental safeguards was Moderately Satisfactory. The safeguard policies triggered by the project were OP 4.01 (Environmental Assessment), OP 4.09 (Pest Management) and OP 4.37 (Safety of Dams), although the very small scale of the sub-projects suggested that potential impacts would be relatively small. The Project had an Environmental Management Framework (EMF) that focused on preventing negative environmental impacts as a result of individual or cumulative effects of sub-projects. The EMF included procedures for environmental screening and assessment of individual sub-projects, training strategy, institutional arrangements and a supervision plan. About half of the sub-projects sampled in an external audit demonstrated implementation of required mitigation measures. The GoCG agency designated to support the State Project Management Unit (SPMU) on implementation of the EMF was not always consistent in service delivery, due largely to its own capacity constraints as a new state entity. The key staff position at the SPMU responsible for the EMF implementation was unoccupied for a significant period, but renewed efforts of the SPMU, especially from 2008, helped in revitalizing EMF implementation. 2.4.4 Social and Tribal Strategy: Compliance with social safeguards was Moderately Satisfactory. The Project placed a major emphasis on reaching out to STs, SCs and women, and had a sound Tribal Strategy that benefitted from the analysis and preparatory work for the MP project, and from the Bilaspur pilot in Chhattisgarh. ST participation in CIGs has been largely satisfactory due to the grassroots mobilization by the PFTs (also true for SCs and women, as discussed more fully in Section 3 on assessment of Project outcomes). However the sheer extent and coverage of the Project meant many actions under the Tribal Strategy were not carried out. This included a separate annual review of tribal CIGs, and regional learning workshops on a routine basis. A socioeconomic and legal study of tribal land in Chhattisgarh completed in 2006 16 provided GoCG with recommendations on steps to reduce land alienation and irregularities in land registration. Another study assessed strategies for working with the poor and the potential of a cluster approach. The ICRR was not able to determine what follow-up these studies may have had. 2.5 Post-completion Operation and Next Phase 2.5.1 By closing, most CIGs were running their own operations without expectation of further support from the Project. In a much smaller number of cases, CIGs were splitting up with little results and/or being run by only a few individuals or families (see Annex 3). Longer-term sustainability of the successful CIGs will depend on profitability and perceived benefits of continuing collective activities. Although not originally planned as part of the Project, some activity-based federations to support CIGs with marketing and input provision were being established in the year before completion. According to GoCG, some 10,000 agriculture-based CIGs are receiving assistance from the government, mainly in the form of extension services. Infrastructure sub-projects under the Panchayat plans were mainly implemented through regular government agencies, which will be responsible for their O&M. The 10% cash contributions from communities were placed in bank accounts (Apna Kosh) for future use; although originally intended to support O&M if needed, thinking evolved over the life of the Project and it was agreed instead that the funds would be held for communities as an initial savings pool to support future income generation activities. 2.5.2 The task team and GoCG counterparts considered the possibility of a second phase project, but GoI and IDA management decided instead that it would make more sense to provide follow-up through a multi-state project in support of the National Rural Livelihoods Mission, and this operation is under preparation at the time of this ICRR. The implementation experience from this project has several important lessons for the multi-state project in support of the National Rural Livelihoods Mission. Foremost among them are: 2.5.3 Dedicated support structure: The Government of GoCG had set up an independent society for implementing the project. Even though the society was conceptualized to have independent governance structure, the district level and block level facilitating teams were embedded in the Zilla Parisad (District Administration) structures with direct reporting relationships to CEO, Zilla Parisad and Block Development Offices. As a result, the professional teams deployed for implementation of the project were withdrawn for use in supporting the priorities fixed by district administration. Important lesson for the NRLM has been to develop a professional organization with multi-stakeholder governance structure with balanced representation to government, banks, private sector and civil society organizations. Direct reporting of district and block level implementation units to the state level support structure will ensure unity of purpose and scalar principle so essential for mission organization. 2.5.4 Identification of the poor and institution building: The Project had invested in elaborate participatory identification of the poor which became basis for selection of project beneficiaries for the sub-projects. The project design did not provide building sustainable institutions of the poor as an important investment to consolidate the CIGs and facilitate linkages with mainstream sectors including financial sector and markets. However, impact studies have shown that the CIGs that were linked to the banks were able to renew and expand livelihood assets and explore market linkages. In the NRLM the focus is on building institutional platforms of the poor and systematic inclusion of the poor and the poorest in these institutions. 17 Livelihood interventions: As the sub-projects identification focused on the asset based livelihoods, interventions opportunities for improving access to entitlements, public services and safety nets so relevant for the poorest among the poor remained out of the radar. The quick results from pilot efforts to develop clusters around NTFP value-chains and sustainable agriculture practices have influenced the implementation of NREGA in the State. This also has been included in the NRLM as an important design feature. 3. Assessment of Outcomes 3.1 Relevance of Objectives, Design and Implementation 3.1.1 The Project objectives have been of high strategic relevance, from preparation through completion, and they have remained consistent with the priorities of GoI, GoCG and IDA over that period. Chhattisgarh has experienced growth rates averaging 10% in recent years, driven mainly by exports of hydroelectricity and minerals, but poverty and exclusion are still pervasive in many rural communities, and insurgency has been on a gradual rise in parts of the State. Improving incomes and well-being on a sufficiently large scale and in a transparent, inclusive and sustainable manner has been – and remains – essential, from both a development and conflict management perspective. 3.1.2 The CDD approach adopted at the Project design stage, which addressed both the livelihoods and governance dimensions of these challenges, was generally appropriate and remains relevant. This was underscored by both communities and local officials. Specifically, they noted that the Project was unique in Chhattisgarh in terms of its focus on inclusion, incorporating livelihoods and working at scale across the State. However, the design team did underestimate the extent of the capacity and institutional challenges that the new State (and hence the Project) would face, and this resulted in an overly ambitious design of the Project. Most core outcome indicators were fully or partially achieved (see below), but at a slower pace and with less investment per capita than initially planned (although still sufficient to achieve the Project’s main income and empowerment objectives). As the first effort of its kind in a new State, the Project could have focused entirely on capacity building and organizational efforts at the community and GS level, and this is where its achievements have been greatest. However, it also aimed to achieve some strengthening of local governments (GPs), mainly through training. This did not bring about much change, although the GP sub-projects are clearly improving infrastructure access for beneficiaries. 3.1.3 Project implementation has not been unilinear, but rather a record of ‘two steps forward, one back’. The process of consolidating new State institutions, and political developments resulted in an uneven, but generally improving, quality and pace of implementation. While the majority of core indicators have been met (with a one-year extension of the Closing Date), many investments took place in the final two years of implementation and this complicates the assessment of prospects for sustainability, which depends heavily on actions to be taken in the post-project period. 3.2 Achievement of PDO 3.2.1 The PDO was to improve opportunities for the poor and the economically vulnerable in the State, especially women and tribal people, to meet their own social and economic development objectives. This outcome was to be achieved by (i) creating infrastructure and income opportunities for the rural poor; (ii) empowering active groups of disadvantaged people; 18 and (iii) supporting village governments in becoming more responsive and effective in assisting them. 3.2.2 Results were to be measured by changes in incomes, migration and participation of disadvantaged people, as well as by other intermediate outcome indicators in the PAD logframe. ICRR findings are summarized below. PDO Indicators  Increase in beneficiary incomes (Achieved): Project achievements were substantial. In village interviews and beneficiary workshops, the increase in incomes of CIG members was invariably cited as the most important Project achievement. CIGs and PFTs reported increases of “2-3 times”. The EFA, based on a sample of 150 CIGs, found an average increase of about 54% per household in real terms compared with pre-project income levels. The lower – though still very robust -- EFA estimate takes into account a share of CIG activities that did not produce benefits. The IE, based on a more limited survey sample, found 20% increase in income over baseline, compared to less than 5% in controls.  Reduction in migration (Partially achieved): The target was partially achieved. In CIG interviews and beneficiary workshops it was reported that in a large number of cases the CIG activities had reduced the need for members and their employees to work away from their own villages and to seek off-farm or agricultural wage labor. In many cases non-farm CIG activities replaced work previously undertaken in nearby cities, and changed beneficiaries from unemployment/wage workers to self-employed. The IE also showed greater non-farm and self-employment in Project villages compared to control villages. However, it was unable to determine overall changes in migration per se. In retrospect, the feasibility of tracking this indicator should have been given more attention during design and at the MTR.  Participation of the most disadvantaged people, particularly women and tribals, in sub- project activities and decision-making (Achieved): Communities were quite involved in selecting, planning, and verifying through GS village meetings, their own CIG sub-projects. These investments have, to a large extent, improved their livelihoods, indicating they selected appropriate activities. A range of stakeholders noted that this opportunity to participate was a unique experience in the State. The IE focus group responses of CIGs noted that the project had strengthened social and political empowerment processes. Some CIG women leaders have contested, and won, Panchayat elections. With regard to non-farm activities, even CIG members who were only partially active were aware of Project activities and purpose, and felt they had a stake in the CIG. Members, including women, STs and SCs, noted that they had increased their interactions with GPs and outside agencies as a result of the CIG activities, and that project activities have made them more exposed to GS and GP functioning, their activities and responsibilities. The selection of Panchayat infrastructure sub-projects was more top-down, done largely by the GPs, although communities felt they were generally useful to them. Intermediate Outcome Indicators for Creating Infrastructure and Income Opportunities  Number of sub-projects submitted, funded and implemented (Achieved). The PAD estimated 20,000 CIG sub-projects and this have been exceeded. Despite the slow start-up of the Project, the CIGs implemented 20,446 sub-projects (with 100-200 still to be 19 finalized). A further, 3,314 community infrastructure sub-projects have been fully implemented by GPs.  Number of families benefited (Achieved). The PAD estimate of 150,000 families has been exceeded. CIG sub-projects benefitted 111,728 families, and Panchayat infrastructure sub- projects benefitted 104,000 families, for a total of 215,728. However, as there was some overlap (households benefitting both from CIG and GP sub-projects), and a small number of sub-projects did not provide benefits, this ICRR simply concludes (conservatively) that the number of households benefitted exceeded the target.  Average cost of sub-projects compared to government standards (Not achieved). The PAD assumed that infrastructure sub-project costs would be lower than government costs for similar investments, but ICRR team estimates found the two to be fairly similar. This is not surprising as the Panchayat infrastructure sub-projects were implemented largely by government agencies, applying their standard unit costs. On the other hand, CIG sub-project costs were reportedly lower than those of some other government programs aimed at strengthening rural incomes. Intermediate Outcome Indicators for Empowering Active Groups of Disadvantaged People  Percentage of CIG members belonging to poor and vulnerable groups (Achieved). With respect to inclusion, just over half of Project beneficiaries were STs (37%) and SCs (14%). Women constituted 25% of beneficiaries, and 20% of CIGs were women only. In the absence of data by block and village, a proxy analysis of the ratios of SC and ST membership with those of the districts as a whole was conducted. Overall the Project was able to target slightly more than originally intended by these estimates, although in a few districts with significant tribal populations the reach of the Project was limited by security considerations, as these were also strongholds of left wing extremism. Regarding poverty, the Project struggled initially with application of various criteria for identifying the poor, including the BPL designation (known for incomplete coverage and leakages). In response, it applied two rounds of more participatory wealth ranking. A very large proportion of active CIG members interviewed at ICRR stage had previously been wage and agricultural laborers, a strong feature of the baseline livelihoods of the poor and the poorest. Interviews with PFT members and other stakeholders indicated that in several districts inclusion of the poorest, however defined, was in the order of 75-80%. Blocks with relatively updated MIS data showed ranges from 30-60%. The average household net income at baseline in successful CIGs assessed under the EFA was less than USD 1 per day. Overall, the ICRR concludes that Project targeting was an improvement on a completely neutral approach; mostly poorer people have been members of the CIGs and there was no significant elite capture of CIGs. The IE focus group discussions noted that sometimes the poorest of the poor had more difficulties in accessing the sub-projects at implementation stage.  Number of beneficiaries with access to the banking system (Partially achieved). In ICRR field interviews, beneficiaries, especially women, highlighted a considerable increase in their confidence in dealing with banks. While the IE did not show a significant difference in the share of CIG beneficiaries with bank accounts compared to control areas, it did find that more CIG members with bank accounts succeeded in taking loans compared to the situation in control villages (48% v. 37%), whereas in control areas more still used money lenders. 20 The CIG members’ loans were for productive purposes primarily for agricultural investments and were double the amount of productive loans taken out in control areas. IE results indicate that ST and SC project households have higher savings compared with control households, and lower savings among other backward castes (OBCs). Overall, CIG households save more than non-CIG households. This saving discipline contributes to CIG households becoming more creditworthy. According to the IE, the total amount of loans taken by CIG households is 30% higher than the amount borrowed by non-CIG households. The duration of the loans in project households was reduced from 10 to 6 months, thus indicating quicker turn-around. Project households borrow 30% more from banks and 40% less from money-lenders, compared to control households.  Community contributions in cash (Achieved). CIG members had to contribute 5% in cash upfront towards sub-project costs, which they did. They then had to place 10% into a village fund (Apna Kosh) as a condition for the release of the second tranche of the sub-projects. The 10% was put into the Apna Kosh (village fund). Members also consistently raised the 10%, although it led to some implementation delays for poorer individuals. At completion, around USD 3.56 million or 10.1% of actual sub-project costs, has been placed in fixed term deposits.8 Intermediate Outcome Indicators for Supporting Village Governments in Becoming More Responsive and Effective in Assisting the Disadvantaged  Improved people’s perceptions of Gram Sabhas and Gram Panchayats (Partially achieved). Participation of beneficiaries in GSs, and their interactions with GPs, increased as a result of the Project. Beneficiaries emphasized that they were more engaged in GSs to discuss their own sub-projects, and this increased their confidence in dealing with other programs and projects. The IE noted little difference between Project and control villages on this point, possibly because other programs operating in Chhattisgarh now also have participatory requirements. As to the GPs, stakeholders commented that overall these bodies have changed very little. In retrospect, the PAD overestimated the impact the Project would have on GPs; to fully achieve expected outcomes would have required complementary policy, institutional and operational arrangements that went beyond the scope of the Project.  GSs and GPs practicing participatory budgeting (Not achieved). GSs and GPs received some orientation and training on Project processes, including the need for participation in Project resource allocation decisions, but there was little focus on participatory budgeting techniques more generally (i.e. going beyond Project resources), and supervision mission Aide-Memoires do not reveal any sustained attention to this issue.  Increase in access to clean drinking water by the poorest habitations (Partially achieved). This indicator was intended to track GPs’ responsiveness to an important need of the rural poor, directly through the Project or by convergence with other programs. After Project start- up, other programs did address water supply issues, and by completion the Project areas have relatively high densities of drinking water pumps (although the IE found little difference vis- à-vis control areas, presumably because such programs operated statewide). The Project did support a small number of hand pumps directly (217) in poorer hamlets away from the main roads. Although the indicator was fully achieved in one sense, the ICRR rating is only ‘partially achieved’ because of difficulties of attribution. 8 This is below the nominal estimate in the PAD because actual sub-project costs were lower than original forecasts; however, it is fully compliant with Apna Kosh contribution requirements. 21 3.3 Efficiency 3.3.1 As is common among demand-driven projects for which the precise investment mix cannot be determined ex ante, the PAD did not estimate an overall Project ERR. Rather, it compared sample sub-projects to similar investments implemented in India, which yielded more than 15% rate of return. For the ICRRR, incremental benefits from livelihood investments were assessed based on data from multiple sources, including the mission’s field visits. Four livelihood project types grouped under farm (agriculture and animal husbandry) and non-farm (traditional and trading/service) activities were evaluated for their impact on income and poverty. All Project costs were included in the analysis at constant 2010 prices. The results were that: (i) average annual incremental financial income for CIG activities ranged from INR 26,951 (non-farm) to INR 32,015 (animal husbandry); (ii) weighted average annual incremental financial income across CIG activities was INR 28,890 per CIG; (iii) average financial rate of return varied from 21.5% for dairy CIGs to 33.4% for piggery CIGs; (iv) ERR for the project investments (INR 304.0 crore, 2010 prices) was estimated at 15.2%; and (v) ERR for the project investments was 15.9%. For want of data, the analysis did not include returns on GP infrastructure sub-projects. 3.4 Justification of the Overall Outcome Rating Rating: Moderately Unsatisfactory 3.4.1 The Project brought reasonable income improvements to a larger beneficiary population than originally forecast, but in a not so efficient manner, and made moderate advances towards the goal of empowering disadvantaged people who were among Chhattisgarh’s poorer and most excluded rural population. Its achievements in strengthening local governments were meager, although this was less a key PDO target than an intermediate outcome objective. The overall outcome rating is based on the analysis of key PDO and intermediate outcome results and shortcomings, with the heaviest weight placed on the income related indicator. 3.5 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects and Social Development 3.5.1 CIG interviews and beneficiary workshops, backed by PFT staff observations in all Districts visited by the ICRR mission, provided consistent feedback that Project activities and incomes generated had yielded specific household benefits for rural poor beneficiaries.9 The most frequently mentioned benefits included being able to send children to private school, reduced indebtedness, and improved living standards (see Annex 6). The IE identified only minor differences in schooling, demographic, asset and health indicators compared with the control group, but noted that in the Project areas use of private medical services and household expenditures were greater overall, and perception of food consumption was slightly better, and proportional expenditure on food was lower. The IE indicated that food security had shown improvement, but this had also occurred in control areas. In beneficiary workshops, community 9 While it was clear that the Project had succeeded in reaching mainly poor households and avoiding elite capture, the ICRR team had more difficulty determining the extent to which the poorest of the poor (POP) had benefited. As this was not an explicit Project goal (documents generally referred only to ‘the poor’), data was not organized for this purpose. In more detailed interviews with a sample of CIGs, confirmed by the IE focus group findings, the team concluded that the Project substantially reached the poor but had more limited reach to the POP. 22 representatives also noted that an important Project achievement was the improved community spirit, unity and collaboration, increased self-confidence, dignity and respect shown to them. 3.5.2 The overall proportion of female CIG members was 25%10 and 20% of all CIGs were exclusively women. The ICRR mission’s observations and interviews with women CIG members, as well as their strong presence in beneficiary workshops, showed that they had experienced a range of benefits from the Project activities. Women and other stakeholders noted improved incomes, greater respect in their roles at work and in the home, increased confidence in working outside the home, in dealing with range of stakeholders, and discovering working through collective action. (b) Institutional Change and Strengthening 3.5.3 Institutional change and strengthening is largely described in Section 4. The Project supported the development of a very large number of CIGs across the State. Some 70% of these were still intact at the end of the Project and generating collective benefits, although a small proportion had become dominated by one or two people, or were centered around families. In the final 12-18 months (including a one-year extension of the Closing Date), several initiatives were launched to consolidate gains achieved and strengthen prospects for the economic and institutional sustainability of the CIGs, including: (i) facilitated formation of activity-based CIG federations for economies of scale and backward and forward market linkages; (ii) the formation of village-level institutions for saving and credit purposes; and (iii) pilot testing of a partnership with NABARD to facilitate transition of CIGs into Joint Liability Groups (JLG) to improve access to financial services. However the scaling up of these initiatives took longer than expected and at the time of the ICRR relatively little implementation of these enhancements had begun. GoCG showed strong verbal willingness to continue the Project processes and scale up the pilots through the proposed NRLM, for which a Bank project is scheduled for Board presentation in June 2011, and which includes Chhattisgarh. (c) Unintended Outcomes and Impacts 3.5.4 In its later stages, the Project supported several important unplanned innovations:  As mentioned above, to strengthen sustainability and access to markets and credit, the Project promoted the development of federations of CIGs in pro-poor sectors with economic potential. Eight activity-based federations were in their early stages of formation by end of the Project, covering about 900 CIGs (in piggery, non-timber forest produce, dairy, fisheries, agriculture, and tailoring). Some village level (mainly thrift-oriented) federations were also being tried, but faced challenges in combining interests of diverse CIGs and existing self-help groups (SHGs). The activity federations were beginning to address some key marketing issues and provide channels for other public programs. These pilots resulted in the drafting of operational manuals that are currently under review by GoCG for their mainstreaming.  The Project promoted and piloted adoption of the System of Rice Intensification (SRI) with CIG beneficiaries. Since 2008, it supported training of farmers, in conjunction with the Indira Gandhi Agriculture University, developing farmers as master trainers to scale up the innovation. This experience demonstrated the feasibility of reduced fertilizer, water pesticide and labor requirements while increasing yields, which in turn has led to its enthusiastic 10 Although no quantitative target was set at project design, from the time of the MTR supervision missions began monitoring against a notional goal of about 30% and reflected this in aide-memoires. 23 adoption by neighboring farmers. The area of adoption at end of Project covered about 750 ha, and was expected to double in the coming wet season. GoCG was keen to promote this and was therefore mainstreaming SRI in its extension system, with targets and incentive packages, and a possible tenfold increase in coverage in coming seasons.  The Project also piloted the adoption of Sustainable Agricultural Practices among women farmers in tribal villages, using technical assistance from an experienced institution based in Andhra Pradesh. Discussions are underway to scale up the pilot though government channels.  The Project piloted a Non Timber Forest Produce (NTFP) Federation, run exclusively by women collectors to support female entrepreneurship in a sector that is traditionally the domain of tribal women. This Federation is also trying out an operational mechanism to include SHGs as members (an approach that allows for self-targeting). Discussions are underway to ensure financial support to this Federation through other government schemes.  The Project has supported the training and job placement for rural youths (2,284) from poor (BPL) households.  Finally, the Project has helped to implement several excellent examples of convergence among different programs (see Annex 3). 3.6 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 3.6.1 A summary of stakeholder workshops is provided in Annex 6. 4. Assessment of Risk to Development Outcome 4.1 The risks to sustainability of project outcomes are considered Substantial. The Project has created income flows that accrue directly to beneficiaries, based on locally appropriate systems that the communities have identified themselves. The EFA indicates that over 70% of CIGs are sustaining benefit flows from their sub-projects. Agriculture activities in particular, (e.g. tubewells) are providing good returns, and there are signs that beneficiaries are likely to use these benefits for further income streams, e.g. diversifying cropping patterns. For agriculture and livestock CIGs (about half of all CIGs), formal links are being developed with line agencies. 4.2 On the other hand some non-farm based CIGs, and some agricultural CIGs (e.g. dairy in some districts) need stronger technical and marketing skills, a fact noted by beneficiaries. Some CIGs engaging in livelihoods that are new to members are at risk of being run by individuals as opposed to groups, with some members dropping out. Some emerging external risks need a system of response that was not in place. For example, there is increasing competition emerging from large Punjabi combine harvesters that threatens CIG sub-projects based on local small threshers. 4.3 CIG contributions have been placed in fixed deposits in village funds (Apna Kosh) in the name of the Panchayat. These funds were originally conceived for O&M purposes, but thinking evolved over the life of the Project and the intent now is to use them as revolving funds to support income generation by village members. At the time of the ICRR, a few villages were piloting this approach. The decision to change the purpose of the funds was based on community demands, as well as the fact that most infrastructure sub-projects were small and were undertaken by the GPs 24 rather than CIGs, and therefore O&M will mainly be a government responsibility (although some communities indicated to the ICRR mission that they will also be involved in minor upkeep). At the same time, the ICRR mission noted that O&M funds are not always assured, and some structures (e.g, community buildings) were not being fully utilized and there was a risk that maintenance may be neglected. 4.4 Some of these sustainability issues arise from the fact that the majority of sub-projects were undertaken very late in the project period and therefore will need to be addressed in the post-project phase. CIG Federations that could provide back-up to improve market access and help link CIGs to other agencies are still in their early stages of formation, will need technical and organizational handholding and do not cover the majority of CIGs yet. GoCG’s request for a one- year extension of the Closing Date was mainly to allow more time to work on consolidation and sustainability of CIG investments, but this was only partially achieved during the extension year because of procurement delays in hiring Support Organizations to assist the CIGs. 4.5 The proposed NRLM has great potential to provide some of the support needed to strengthen sustainability, and a multi-state Bank project to support NRLM is scheduled for Board presentation in June 2011. Pre-Launch Workshops for the NRLM have already been organized. GoCG authorities and technical staff attended initial workshop and discussed how to roll out NRLM in Chhattisgarh, building on and providing further technical assistance and capacity building for the CIG structures created under the Project. 5. Assessment of Bank and Borrower Performance 5.1 Bank (a) Ensuring Quality at Entry Rating: Moderately Satisfactory 5.1.1 Project design was generally sound, addressed important strategic priorities of GoI, GoCG and IDA, and built on good CDD practices emerging elsewhere in India and globally. The design team correctly identified the institutional and technical capacity constraints of the new State, but underestimated their severity and the likely impact of ensuing political developments. CIG sub-project costs were overestimated in comparison with similar projects elsewhere. Having decided to include local government strengthening in the Project, the team should have planned a stronger program of support than simply training. The M&E plan was comprehensive, although more attention should have been paid to the feasibility of tracking some indicators, such as migration. Fiduciary and safeguards arrangements were well-prepared and implementation arrangements were piloted in Bilaspur District. Most key risks were identified although, again, the severity of some was underestimated. (b) Quality of Supervision Rating: Moderately Satisfactory 5.1.2 IDA provided adequate supervision support throughout most of the Project period. Formal supervision missions took place and ISRs were prepared twice yearly, with more frequent short supervision inputs by both Headquarters and field-based staff, especially during periods when ISR ratings slipped into the unsatisfactory range. More intensive mission support helped to 25 get the Project up and running after initial launch delays, and were also quite important in the final two years when IDA redoubled efforts to ensure proper closure of sub-projects and introduce several initiatives to strengthen prospects for CIG sub-project sustainability. The skill mix of supervision teams was broadly appropriate and cooperation between Washington and field-based staff was good, but there was excessive turnover of Task Team Leaders (4), with some changes occurring at critical times for Project decision-making. This lack of continuity in team leadership was mirrored by inopportune and excessive changes of key staff on the Borrowers’ side. As noted above, the Project did achieve core PDO objectives and many key targets, but greater TTL/PD continuity by IDA and the Borrower could have strengthened these results (e.g. more attention to refining project processes, indicators and targets; possible dropping or reformulation of some activities, in particular trying to strengthen GPs, which might have been better addressed as a separate initiative in the Chhattisgarh context; greater focus on targeting issue; and planning for post-project sustainability.) (c) Justification of Rating for Overall Bank Performance Rating: Moderately Satisfactory 5.1.3 Bank performance was generally good, although with moderate shortcomings. Project design was sound and reflected lessons of Bank experience with CDD projects in India and elsewhere, but decisions on scale were overly optimistic and the GP activities should either have been dropped or treated in a more comprehensive manner. With those exceptions, technical, economic, M&E, fiduciary and safeguards preparation were satisfactory; the time and budget for preparation were efficient; and the skill mix which the Bank brought to bear through the task team and peer advisors were appropriate, although continuity in task leaders could have been better. Design of the Project was generally sound, though there could have been greater emphasis in ensuring the institutional readiness and build-up of capacity of implementing agencies. 5.2 Borrower (a) Government Performance Rating: Moderately Satisfactory 5.1.4 The Borrower (GoCG) provided important support contributing to the Project’s achievements. Despite heavy pressures on a very thin senior management team, owing to the newness of the State, GoCG enthusiastically provided considerable attention to Project design and financed the Bilaspur pilot with its own resources. During implementation, the State met counterpart commitments and financial oversight functioned well on the whole. Thanks to champions within GoCG, project performance improved over time, to the point that many core project targets were met. However there were also some significant shortcomings. There were serious delays in recruitment of some key staff positions, in putting in place HR policies to hire and retaining high quality staff, and there was excessive turnover of Project Directors – six in all – partly explained by the newness of the State, where a few key highly qualified staff are needed to hold several positions and address a range of institutional and other issues as they arose in several places. (b) Implementing Agency Performance Rating: Moderately Satisfactory 26 5.1.5 The implementing agencies, the State and District project management units, managed a challenging project for such a new State, achieving key targets in terms of outreach, livelihoods improvement and basic inclusion of the poor and marginalized groups. Basic systems were put in place to achieve this, albeit with delays or of limited structure and quality. The development of several key strategies was delayed (as for M&E and training, use of the Apna Khosh funds), and at times inconsistencies and/or lack of detail and clarity in guidance complicated implementation (on topics such as beneficiary targeting, implementing the innovations fund which was not followed through, and in the development of linkages with line agencies). Financial management improved over the Project period, while procurement performance, which had been strong from the start, deteriorated during the final two years due mainly to staffing constraints. A strong sense of purpose was driven by some very committed senior managers and field staff during some periods, but continuity of leadership and retaining competent staff remained chronic issues throughout. (c) Justification of Rating for Overall Borrower Performance Rating: Moderately Satisfactory 5.1.6 There were some very positive aspects of Borrower and implementing agency performance, which helped the project complete its core targets. However, on balance there was insufficient attention by the Borrower to important staffing and implementation issues which, had they been better managed, would have led to stronger results and better prospects for sustainability of Project investments. 6. Lessons Learned  A focused and participatory livelihoods approach through organized beneficiary groups can provide quick and tangible income results and improve well-being of poor rural households, and as a result contribute to their improved status in the community. There needs to be a strong technical support, especially in the case of marginalized groups who are exploring new livelihoods strategies.  A focus at the village level is also advisable for a CIG-based model. A village-level organization would allow for more effective social accountability and monitoring mechanisms, for instance by transferring the responsibility for certification of use of funds from the project team to the village organizations – as is now common practice among other South Asian projects – and the development of Village Development Plans that are truly participatory and inclusive. Institutional linkages between village organizations with Gram Panchayats would ensure appropriate coordination and be mutually beneficial.  The length of a project needs to be long enough to allow for proper sequencing of activities – first participatory identification and mobilization of the poor and formation of the institutions of the poor (which other projects have shown to require at least six months, and at least 9 months among tribals), then asset creation and aggregation through federations - and then for innovations and building of local capacities.  Direct control of funds, and a demand-driven process for investment decisions by beneficiaries, is critical for the empowerment of the poor, and to make sure resulting livelihoods sub-projects are suitable to the capacities, resources and social relations of poor community members. However for sustaining and developing such livelihoods it is important that funding support is matched by very good technical advice and linkage development. 27  Targeting needs a structured, locally appropriate and participatory approach to wealth ranking. This needs to be well prepared, communicated, transparently and fairly implemented, and consistently applied statewide. Social accountability mechanisms need to be introduced at the community level, to ensure compliance through self-monitoring and peer pressure.  A revolving fund mechanism is essential to maintain community based financing support, to build credit discipline and make groups creditworthy to access Bank loans. Revolving funds also spread the benefits of the project. Revolving fund mechanisms have also shown to facilitate social mobilization, especially among women and vulnerable groups. As wealthier villagers are normally not interested in accessing revolving funds, these also contribute to proper targeting. In the case of Chhattisgarh, a careful approach on how to manage the financial relationships between village level revolving fund based institutions/federations (composed various kinds of CIGs and SHGs), and the financial needs of activity-focused federations was being developed at the end of the Project. Delay in the development of such mechanisms causes considerable uncertainty for communities.  Aggregation through federation building is critical for maintaining support to CIGs in terms of marketing, technical advice and linkage to financial services. It takes considerable time and dedicated support to ensure that this benefits a large number of community organizations, with very particular individual sectoral and geographical needs.  Community contributions are important for promoting ownership, but there needs to be flexibility in implementation, especially in the early stages. Clear rules must be developed, consistently followed, and regularly reviewed to ensure they are contributing to project effectiveness and objectives.  Local monitoring at the village level, involving community members selected by the beneficiaries themselves, is critical to ensuring progress and transparency of activities. This needs to be formalized in the project processes, and supplemented by a range of feedback mechanisms including workshops, etc. which are followed through then at higher levels.  ML&E systems need to be more internalized and practically oriented. A good learning system is necessary to be able to adjust program processes to changing needs and circumstances. A core area of management needs to be a thorough mechanism for regular review and stakeholder agreements on project indicators and targets, and have MIS systems adapted to this management need. This should not be seen as a separate minor staff function.  Capacity building and institutional development needs special attention in the case of new governments. This should be based on a comprehensive initial needs analysis of all levels of actors, with due attention to the competing demands which will be placed on scarce technical and managerial staff when a new government is simultaneously trying to develop core institutional structures in many areas. 7. Comments on Issues Raised by Borrower, Implementing Agencies and Partners (a) Borrower/Implementing Agencies 7.1 The borrower has communicated via email on April 21, 2011 that they do not agree with the rating of overall project performance in terms of Development Objective (DO) and outcomes as Moderately Unsatisfactory but agree to the rating about GoCG and World Bank performance as Moderately Satisfactory. The borrower has not provided any elaborate comment on the draft 28 ICRR, but specifically on the issue of PDO indicator about migration the borrower had commented that “the state doesn't agree to it, as migration has reduced due to convergence of many development schemes such as NREGA, BRGF, SGSY & others”. (b) Co-financiers N/A (c) Other Partners and Stakeholders N/A (c) Other Partners and Stakeholders N/A 29 Annex 1. Project Costs and Financing (a) Project Cost by Component (in USD million equivalent) Components Appraisal Actual/Latest Estimate Estimate I. Institutional and Human Capacity Building 23.57 8.31 Human resource development 3.89 0.90 Strengthening of Panchayats 5.96 1.20 Formation and strengthening of organizations 5.80 2.72 Communication and information 1.35 0.56 Monitoring, learning and special studies 1.07 0.51 Project administration 5.50 2.42 II. Community Investments 105.78 58.94 Community investments 64.79 41.38 Panchayat plans 33.51 13.57 Village funds (Apna Kosh) 6.48 3.98 Innovation fund 1.00 0.01 Total Baseline Cost 129.35 67.25 (b) Financing Appraisal Actual/Latest Type of Percentage of Source of Funds Estimate Estimate Financing Appraisal (USD million) (USD million) Government 5.40 3.53 65.37 IDA Credit 112.56 57.59 51.12 Local Communities 11.39 6.04 53.03 Annex 2. Outputs by Component 1. The Project delivered most of the key outputs, despite considerable delays in the initial phase of implementation and the fact that it disbursed only about 52% of the planned budget at end of project. There were shortfalls in some major outputs, particularly in capacity building. This Annex provides a summary of outputs, further more detailed presentation on Outputs is provided in Annex 7 Borrower’s report. Component (1) Institutional and Human Capacity Building: (a) Project Facilitation Teams (PFTs): The project formed PFTs in each of the 40 project blocks, including 8 contracted NGO PFTs, to mobilize and facilitate communities’ engagement in the project. Up to MTR, there were considerable vacancies in the PFT positions. Nevertheless, as a final result, 20,689 Common Interest Groups (CIGs) of around 5 or more community members were formed to access project resources. (b) Communications and Information Dissemination: The Project provided information and messages to communities through a range of communication materials, radio and popular dissemination practices. (c) Capacity Building for Project Stakeholders and Implementers: Human Resource Development (HRD) for project stakeholders was provided through 5-day orientation programs to key stakeholders. A well-established NGO, CARD from Madhya Pradesh, provided initial training, and the State Institute of Rural Development (SIRD) provided further training to CEOs. Exposure visits for project implementers were made to neighboring States as well as overseas. A total of 19,583 community members were provided mainly classroom based technical training related to their CIG activity. (d) Capacity Building for Gram Panchayats: Basic orientation was provided to 20,000 Panchayat functionaries by SIRD. Some 193 Gram Panchayats also received support in the form of building of GP offices (close to the 200 target in the PAD). There seems to have been little activity in terms of demand-based support for strengthening participation processes or budgeting. (e) Monitoring, Learning and Evaluation (MLE): For MLE, the project to a considerable degree followed regular Block and District lines of administrative reporting, that was revised and consolidated after MTR. A Baseline study was conducted in 2005-6, a web-based MIS established after the MTR, and 5 thematic studies prepared around the time of the MTR. (f) Project Administration: The project was to be administered by an autonomous State Society established in 2003, which would allow the Project Director to make some important financial decisions. However in practice the Executive Committee of the Society took most key decisions. Implementation was managed by the State Project Management Unit (SPMU), harnessing the District and Block CEOs. The SPU and District project units operated with government staff on deputation from other departments. Some key State and District project units’ officers’ positions, especially relating to technical and monitoring, were frequently vacant, due to high turnover, as well as difficulty in filling positions. Component (2) Community Investments: (a) Income-generating Activities: A total of INR 1769.5 crore (about USD39.32million) in matching grants for collective income-generating activities were made to 20,446 completed CIG sub-projects identified by community members and approved by the respective Gram Sabhas. Communities contributed about US$ 2.06 million for these activities making a total of about 31 US$41.38 million of total investments in CGI sub-projects. These sub-projects were mainly for agriculture, livestock and traditional local activities. (b) Village Infrastructure: A total of INR 61.08 crore (about USD13.57 million) was made as Matching Grants to finance 3,314 completed investments in village infrastructure. 60% of these investments were for paved cement roads. (c) Village Revolving Funds: Village Funds (Apna Khosh) were created by CIG members contributing 10% of their CIG sub-project cost to a fixed term deposit account for each project village. A total of INR 17.92 crore (about USD3.98 million) has been deposited. Although originally intended mainly for O&M, the draft guidelines for the use of the Apna Khosh have been prepared, for their use as revolving funds, rather than the original design as funds mainly for O&M of community infrastructure. These revised guidelines were introduced in response to community demands for access to credit and the perceived need to build capacity at the local level, and strengthen village level institutions. (d) Innovations Fund: An Innovations Fund to provide small grants for demonstration sub- projects was initiated in 2009. Guidelines were uploaded in the project website and calls for bids were published online. 16 agencies submitted their proposals and a few of them were invited to Raipur to further refine their proposals with some fieldwork. As the project was closing, this activity was not pursued further. 32 Annex 3. Economic and Financial Analysis 1. The project’s PDO is to improve opportunities for the poor and vulnerable to meet their own social and economic development objectives. To achieve this, the project would create infrastructure and income opportunities for the rural poor; empower active groups of disadvantaged people; and support village governments in becoming more responsive and effective in assisting them. The project has two components namely: institutional and human capacity building, and community investments. Under community investments, the project has supported (i) formation of 20,689 CIGs in 2023 GPs, through which 111728 members are provided access to project resources and (ii) 3355 infrastructure sub-projects to support the income generating livelihood investments in the project villages11. A. PAD Estimated Project Benefits and Rate of Return 2. No specific economic and financial analysis was done to the proposed project investments on the ground that community investments are demand driven in nature. PAD quoted the experiences from similar projects in India and other regions wherein, demand driven projects with community contributions are cost-efficient besides being financially viable. Specifically, MP-DPIP experience was used to emphasize the high rate of return from such community managed livelihood sub-project investments. Financial net present values varied from USD 603 to USD 6,208 and Economic net present values ranged from USD 2,059 to USD 4,450 for goatery, buffalo dairy, land leveling, tailoring, and brass band based livelihood sub-project investments in MP-DPIP at 15% real interest rate12. B. ICRR Estimation methodology 3. The project has benefited 111728 members through 20689 CIGs. Women and tribals, together constitute half of the project T-1 Beneficiary CIGs beneficiaries. 20446 sub-projects are completed up to the end of the project with a total disbursement Sub-project Category CIGs % of INR 176.95 crore as village funds under Agriculture based 4993 24% community investment. Animal husbandry sub- Animal husbandry 5750 28% projects dominated the livelihood choices of the Traditional activities 5190 25% Trading activities 4171 20% CIGs closely followed by traditional, agriculture Service sector 585 3% and trading activities (T-1). The project has Total 20689 100% supported 3,355 community infrastructure sub- project investments covering cement concrete roads, culverts, drinking water supply, and other common facilities in the project villages. 4. For cost benefit analysis of livelihood sub-projects implemented through CIGs, data and documentation from multiple sources are compiled and used. They include; the rapid impact assessment study13 conducted during March/April 2010, WB mission reports14 and ICRR mission field visits and workshops, covering 151 CIGs proportionately representing five major livelihood sub-project categories and distributed across three socio economically diverse northern, central 11 "Nawa Anjor", Chhattisgarh Rural Poverty Reduction Project, Project Completion Report, Government of Chhattisgarh, Panchayat and Rural Development Department, June 2010. 12 Project Appraisal Document of Chhattisgarh District Rural Poverty Project, The World Bank, March, 2003. 13 Chhattisgarh District Rural Development Project (CGDRDP), Assessing the Financial and Economic Impact of the Project Interventions by Juan Morelli and Anand Shukla, WB/FAO Consultancy Report. 14 Non-Farm rural enterprise notes by Arup Barua, WB Mission Notes, March 2010. and southern regions. WB mission study on community infrastructure investments covered 12 sub-projects covering cement concrete roads, culverts, hand pumps and GP buildings. These databases are supplemented by the secondary data sources, Bank’s supervision reports and interactions with the project implementing agencies for quantifying the physical benefits and assessing the project costs. 5. Total project cost is INR 275.36 crore in current prices (T-2). Community investment component accounted for 86% of the project costs, which is distributed between village funds for CIGs (64%) and community infrastructure (22%). All project costs are included in the analysis at constant 2010 prices. Total project cost at constant 2010 prices is INR 304.0 crore. T-2 Project costs by Components and Sub-components (Rs Crore) Total Total (Current (2010 Real Components/Sub components Prices) Prices) Institutional & Human Capacity Building 37.26 41.00 Formation and Strengthening of Organizations 12.24 13.29 Communication and Information 2.54 2.70 Human Resource Development 3.86 4.07 Strengthening of Panchayats 5.42 6.49 M&E and Special Studies 2.31 2.59 Project Administration 10.89 11.86 Community Investments 238.09 262.98 Matching grants for Community Infrastructure 61.08 70.06 Village Funds 176.95 192.86 Innovation Funds 0.06 0.06 Total 275.36 303.98 C. Project Benefits T-3 Agriculture CIGs 1. Agriculture CIG Sub-Projects: The project CIG Activities CIGs % has supported 4,993 CIGs for agriculture-based Agriculture Based 4993 100% activities, accounting for 24% of project CIGs. Tube Wells 2517 50% Irrigation investments for tube wells, open wells, lift Open Wells 839 17% and sprinkler irrigation accounted for 75% of the Vegetable/Floriculture 698 14% agriculture CIGs (T-3). The analysis is based on 41 Lift Irrigation 356 7% sample CIGs distributed across diverse regions, and Agriculture Equipments 43 1% Sprinkler 29 1% representing different types of irrigation investments Medicinal Plant 11 0.2% and high value crop cultivation for quantifying the Others 500 10% incremental benefits. T-4 CIG level impacts due to irrigation investment 2. Irrigation CIGs: Average cultivated area per CIG is 1.2 ha. Without project, only Average Impact per rainfed paddy was cultivated during the CIG Unit WOP WP monsoon season, yielding 2 MT/ha, with a Net cropped area Ha 1.2 1.2 probability of losing the crop once in five years Gross cropped area Ha 1.2 2.0 due to drought. Irrigation infrastructure, Cropping intensity % 100 167 provided by the project, ensured reliable water Gross irrigated area Ha 0 2.0 supply for the crops, increased cropping Area under paddy Ha 1.2 1.6 intensity by 66%, diversified cropping pattern Area under non-paddy Ha 0 0.4 Paddy production MT 2.4 6.4 Annual gross margin INR 9600 38756 34 to non-paddy crops in 20% of the gross cropped area and enhanced average paddy yield to 4 MT/ha (T-4). 3. Average incremental financial benefits due to irrigated area intensification and diversification per CIG are estimated at INR 29,156 per annum. Rapid impact assessment survey estimated the incremental financial benefits per CIG, benefited by irrigation investment, ranging from INR 19,028 to 105,917 per annum across different regions15. Average per capita income in the CIG with irrigation infrastructure has increased by INR 5,830 per annum, which is adequate to lift at least one member of the project beneficiary household above poverty line16. Based on mission’s field visits and discussions with the CIGs, it was considered that full incremental benefits would be realized, in phases over three seasons, by the beneficiaries following the installation of irrigation infrastructure. 4. Convergence Impacts: Incremental benefits from CIG investments for irrigation infrastructure get maximized when there is convergence in the investments complementing each other as observed in one project village (Fig.1 and A-1 below). With 80000 irrigation investment only, CIG’s incremental benefits increased by 60000 INR 29,156 per annum. With 40000 complementary investments from community infrastructure and 20000 other public funds like NREGA 0 converging to have positive Without Project Irrigation Convergence impact on irrigated agriculture investment impacts through improved ground water recharge and reliable power Fig 1 Maximum annual benefits from convergence, INR/CIG supply, annual incremental benefits per CIG went up to INR 73,070, an increase of 89% over irrigation investment alone. Annual incremental benefit per capita (INR 12,690) more than doubled as compared to irrigation investment impact only. This will ensure minimum poverty line income for at least two members for each beneficiary CIG in the project area. 23 TW groups, consisting of 102 members are benefited by this convergence impacts. 5. Projected benefits from agriculture CIGs: Out of 4,993 agriculture CIGs, 70% are expected to sustain their improved farm operations during the project period. Rapid impact assessment survey and mission’s field visits recorded about 30% of the CIGs failed to function effectively and are considered unsustainable. Based on this, the financial incremental benefits from agriculture CIGs are estimated at INR 10.2 crore per annum, which will increase and stabilize at INR 11.2 crore at full development. For want of adequate data at project level, convergence impacts are not included and to that extent, estimated benefits will be conservative. 15 Rapid impact assessment study considered only the successful CIGs for this analysis and hence the reported incremental benefits are upward biased. 16 For rural Chhattisgarh, per capita poverty line income level, in 2010 real prices, is Rs. 4,875/annum. 35 6. Animal Husbandry CIG Sub-Projects: The T-5 Animal Husbandry CIGs   project has supported 5,750 CIGs for animal CIG Activities CIGs % husbandry based activities, accounting for 28% of Animal husbandry Based 5750 100% the project CIGs (T-5). Dairy, fisheries, piggery Dairy 3415 59% and goatery accounted for 91% of the animal Fisheries 759 13% husbandry CIGs. Poultry, sheep and bee keeping Piggery 706 12% Goatery 430 7% accounted for 6% of the animal husbandry CIGs. Poultry 291 5% 33 sample CIGs distributed across diverse Sheep 17 0.3% regions, and representing different types of animal Bee Keeping 11 0.2% husbandry based investments and cattle trading activities are analyzed for quantifying the incremental benefits. 7. Dairy CIGs: The project has supported 3415 dairy CIGs. A sample of 22 dairy CIGs, distributed across north, central and south regions, is analyzed. Average dairy herd size is five per CIG, consisting of two dry and three milch animals (T-6). CIG’s plan is to maintain at least three milch animals continuously in lactation every year during the first phase, while T-6 CIG level incremental impacts due to Dairy investment planning to expand the unit size. Average Average Impact per milk yield is 6 l/day/animal, expected to CIG Unit With Project reach 9 l/day/animal in the third lactation Herd size number 5 cycle, with an average lactation period of Milch animals number 3 290 days. Annual financial gross margin per Dry animals number 2 CIG is INR 34,450. Rapid impact Milk production l/lactation 7830 assessment survey estimated the Milk yield l/d/animal 9 incremental financial benefits per CIG, Lactation period days 290 benefited by dairy investment, ranging from Gross margin INR/annum 34450 INR 22,900 to 52,625 per annum across different regions17. Average per capita income in the CIG with dairy investment has increased by INR 6,890 per annum, which is adequate to lift at least one member of the project beneficiary household above poverty line18. Based on mission’s field visits and discussions with the CIGs, it was considered that full incremental benefits would be realized, in phases over three lactation cycles, by the beneficiaries following the investments on dairy activity. 8. Goat CIGs: The project has supported 430 goat CIGs. A sample of six goat CIGs, distributed across north, central and south regions, is analyzed. Average membership is five per CIG. Average herd size is 16 goats per CIG, consisting of one buck and 15 goats. CIG’s plan is to maintain at least 15 does continuously in the rearing cycle of the first phase and expand the unit size gradually further subject to the space and resource availability. At full development in first phase, realized in three years, annual rearing cycle yields 30 goats for sale, following three kiddings in two years with half of the time yielding twins, and 15% mortality rate. Annual gross margin from sale of goats per CIG is INR 29,203, yielding annual per capita financial income of INR 5,841 per CIG member. Rapid impact assessment survey estimated the incremental financial benefits per CIG, benefited by goat investment, at INR 53,600 per annum in central region19. Average incremental per capita income in the CIG with goat investment is adequate to lift at least one member of the project beneficiary household above poverty line20. 17 Rapid impact assessment study considered only the successful CIGs for this analysis and hence the reported incremental benefits are upward biased. 18 For rural Chhattisgarh, per capita poverty line income level, in 2010 real prices, is Rs. 4,875/annum. 19 Rapid impact assessment study considered only the successful CIGs for this analysis and hence the reported incremental benefits are upward biased. 20 For rural Chhattisgarh, per capita poverty line income level, in 2010 real prices, is Rs. 4,875/annum. 36 9. Piggery CIGs: The project has supported 706 piggery CIGs. A sample of eight piggery CIGs, distributed across regions, is analyzed. Average membership is five per CIG. Average herd size is seven pigs per CIG, consisting of six female pigs and one male pig. At full development in phase- 1, realized over four years, annual rearing cycle yields 25 pigs and 10 piglets for sale. With a mortality rate of 25%, annual gross margin from sale of pigs per CIG is INR 21,950, yielding annual per capita financial income of INR 4,390 per T-7 Traditional Activity CIGs CIG member. Annual gross margin is on the lower side since, mortality rate is high in the sample CIGs visited CIG Activities CIGs % and with adequate veterinary health care mortality rate Traditional Activities 5190 100 can be brought down to less than 10%. Rapid impact Masonry 1450 28 assessment survey estimated the incremental financial Tailoring 815 16 benefits per CIG, benefited by piggery investment, at Carpentry 741 14 21 INR 28,700 per annum in south region . Average Bamboo products 286 6 Leaf Plates 215 4 incremental per capita income in the CIG with piggery Spices 187 4 investment is marginally short of per capita poverty cut- Fabrication 179 3 off income level, but by reducing the high mortality Papad/Badi/Pickle 167 3 rate, scope for increased income for the CIGs exists22. Lac/Silk/Mushroom 145 3 The piggery CIGs are also planning to triple the unit Pottery 140 3 size gradually over time. 10. Projected benefits from animal husbandry CIGs: Out of 5750 animal husbandry activity based CIGs, 70% are expected to sustain their animal rearing activities during the project period. Rapid impact assessment survey and mission’s field visits recorded about 30% of the CIGs failed to function effectively and are considered unsustainable. Based on this, the financial incremental benefits from animal husbandry CIGs are estimated at INR 12.9 crore per annum, which will increase and stabilize at INR 28.3 crore at full development. The mission observed expansion of unit size in case of dairy, goatery and piggery CIGs and it was projected to double for dairy and triple for goatery and piggery in phases. Available space and resources will limit further expansion, as expressed by the CIG beneficiaries. 11. Traditional Activity based CIG Sub-Projects: The project has supported 5,190 CIGs for taking up traditional activities to improve their livelihood income. Masonry, tailoring, carpentry, bamboo products and leaf plates accounted for about 2/3rd of the traditional activities (T-7). 32 sample CIGs distributed across diverse regions, and representing different types of major traditional activity based CIG investments are analyzed for quantifying the incremental benefits. 12. Average membership is five to six per traditional activity CIG. Across major traditional activities, annual incremental benefit is estimated at INR 24,812 per CIG. The mission also observed that the turnover of the CIGs, started earlier, has expanded by about 25% in about three year time frame and the same ratio was adopted for all the traditional activity CIGs. Rapid impact assessment report revealed that about 70% of the traditional activity CIGs is projected to sustain their activities. Based on this, the financial incremental benefits from animal husbandry CIGs are estimated at INR 9.0 crore per annum, which will increase and stabilize at INR 11.1 crore at full development. 21 Rapid impact assessment study considered only the successful CIGs for this analysis and hence the reported incremental benefits are upward biased. 22 For rural Chhattisgarh, per capita poverty line income level, in 2010 real prices, is Rs. 4,875/annum.     37 T-8 Trading/Services CIGs 13. Trading and Services based CIG Sub-Projects: The CIG Activities CIGs % project has supported 4,171 CIGs for taking up trading Trading Activities 4171 100% and service related activities to enhance their livelihood Tent 986 24% income. Tent hiring, NTFP, Cement works, Light/sound NTFP 795 19% and Band accounted for about 2/3rd of the trading and Cattle trading 349 8% service activities (T-8). 42 sample CIGs distributed across Light/Sound 305 7% diverse regions, and representing different types of major Band 287 7% Holler/Flour Mill 246 6% trading and service activity based CIG investments are Animal Feed 227 5% analyzed for quantifying the incremental benefits. Cement works 131 2% Service Sector 585 14. Animal Trading CIGs: The project has supported 349 animal trading CIGs. A sample of six goat trading CIGs, distributed across regions, is analyzed. Average membership is six per CIG. Average herd size traded during one cycle ranges from 15 to 20 and annually, five trading cycles are taken up by these CIGs. Goats are procured from faraway places for which the CIG members are split into 2 to 3 groups. Goats are normally kept for few days, which is the time taken for finding the prospective buyers. Average sale price is 22% more than the cost price. After accounting for all other costs associated with the procurement, maintenance and disposal, annual gross margin for the CIG involved in animal trading is estimated at INR 24,000 per CIG and INR 4,000 per member. 15. Average membership is five to six per trade related activity CIG. Across all major trading and service activities, average annual incremental benefit is estimated at INR 29,286 per CIG. The mission also observed that the turnover of the CIGs, started earlier, has expanded by about 25% in about three to four year time frame and the same ratio was adopted for all the trading activity CIGs. Based on the rapid impact assessment report, about 70% of the traditional activity CIGs is expected to sustain their activities. Accordingly, financial incremental benefits from trading and service CIGs are estimated at INR 9.8 crore per annum, which will increase and stabilize at INR 12.2 crore at full development. D. Rate of Return Analysis 16. There was no estimation of IRR in the PAD. ICRR estimated rate of return (T-9) to the project investments varied from 15.2% (FRR) to 15.9% (ERR). All project costs are included in the analysis at constant 2010 prices. Across diverse sub-project categories, the ICRR estimated rates of return varied from 21.5% for T-9 Project analysis summary dairy CIGs to 33.4% for piggery CIGs. The variation between FRR ICRR Estimation PAD and ERR is due to the difference in Project Analysis PVC PVB NPV IRR IRR the economic and financial prices used for the agriculture based CIG Financial analysis 2,116 2,418 302 15.2% Not Estimated investment analyses. Economic analysis 1,905 2,236 332 15.9% 17. The project has initiated 3,355 infrastructure sub-projects, of which 3,314 are completed. 2/3rd of the infrastructure sub-projects relate to the CC roads and culverts. There was no data available to assess the cost efficiency of infrastructure sub-project investments and end of the project impact assessment also missed the evaluation of benefits generated by diverse infrastructure investments. Hence, estimated rate of return, which considered only the quantifiable benefits from the CIG investments alone, will underestimate the returns. 38 A-1 Impact of Investment Convergence in Agriculture CIG, Lapkasa Village NREGA (Rs 0.37 M) Tank Excavated (Rs 0.30 M) (Further deepening) (Nava Anjor) 10 TW CIGs Transformer (Nava Anjor) (Nava Anjor) Transformer 13 New Private TWs (Nava Anjor) (in Tank command area) Under Nawa Anjor, the community in Lapkasa village excavated one tank for harvesting and storing the runoff rain water. During off-season, brick making is taken up inside the tank bed. Using the NREGA funds, this tank was further deepened and storage capacity increased, which demonstrates the efficient use of project and public funds. Next, the community came together to install two transformers for minimizing the voltage fluctuations to ensure reliable power supply while promoting 10 CIGs with tube well investments for irrigated agriculture in the command area of the tank, which demonstrates the convergence in infrastructure and livelihood investments. With improved ground water recharge and improved reliability in power supply, 13 more tube wells have been installed through private investments. The farmers from all CIGs reported increase in the irrigated area commanded by the tube well by 33%, doubling of cropping intensity, reduced repair costs by two-third due to reliable power supply, reduced irrigation time and savings in power consumed for irrigation by 1/4th, increased crop diversification by 37% and enhanced paddy yield to 5.1 MT/ha. Power savings alone work out to 300 Kwh per acre of paddy. The cumulative impact cost savings, additional area coverage, and yield is annual incremental financial benefits going up to Rs 63,470 per   CIG, benefiting 23 TW CIGs/groups.               39 Annex 4. Bank Lending and Implementation Support/Supervision Processes (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Ali Awais Counsel D.J. Baxi Procurement Geeta Sethi Economics Grace Domingo Program Assistant Hans Binswanger Economics (peer reviewer) Community Development (peer Luis Coirolo reviewer) Luis Constantino Task Team Leader Madhavi Pillai Co-Task Team Leader Manvinder Mamak Financial Management Community Development (peer Parmesh Shah reviewer) Paul Martin Environmental Economics Ranjit Nayak Tribal Issues Sara Gonzalez-Flavell Senior Counsel Sarita Rana Program Assistant Syed I. Ahmed Senior Counsel Warren Waters Social Development Yoshiko Masuyama Program Assistant Supervision/ICRR Ai Chin Wee Senior Operations Officer SASDA Arun Kumar Kolsur Procurement Specialist SARPS Arun Manuja Sr Financial Management Specialist SARFM Assaye Legesse Senior Agriculture Economist AFTAR Senior Rural Development Spec., Barbara Verardo SASDA Former Task Team Leader Benjamin Powis E T Consultant EAPVP Dhimant Jayendraray Baxi Sr Procurement Spec. SARPS Kalyani Kandula Consultant SASDI Madhavi M. Pillai E T Consultant ARD Mam Chand Consultant SARPS Manvinder Mamak Sr Financial Management Spec. SARFM Martien Van Nieuwkoop Program Coordinator AFTAR Norman Piccioni Former Task Team Leader SASDA Parmesh Shah Lead Rural Development Spec. SASDA Priti Kumar Sr Environmental Spec. SASDI Ramachandran R. Mohan Senior Social Development Spec SASDS Reena Gupta Natural Resources Mgmt. Spec. SASDA Salimah Haiderali Samji Consultant SASDI Senior Rural Development Spec., Samik Sundar Das SASDA Task Team Leader Sanjay Gupta Consultant ETWSA Sitaramachandra Machiraju Consultant EASER Supriya P. Kumar Junior Professional Associate SASDA Talib B.K.Esmail Former Task Team Leader LCSDE Vibhuti Narang Khanna Program Assistant SASDO Vinayak Narayan Ghatate Consultant SASDU Warren Waters Consultant EAPCO (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle USD Thousands (including No. of staff weeks travel and consultant costs) Lending FY02 12 66.64 FY03 43 157.72 FY04 0.00 FY05 0.00 FY06 0.00 FY07 0.00 FY08 0.00 Total: 55 224.36 Supervision/ICRR FY02 0.00 FY03 0.00 FY04 16 55.37 FY05 38 108.22 FY06 27 107.02 FY07 27 121.22 FY08 37 268.80 FY09 24 146.51 FY10 17 109.91 FY11 03 6.62 Total: 167 923.67 41 Annex 5. Impact Evaluation Results 1. An end of the Project Impact Evaluation (IE) was conducted by an independent agency, the Centre for Bhartiya Marketing development (CBMD) during May 2010-February 2011. The IE measured the three key indicators of the Project Development Objectives (PDOs) as well as five out of nine intermediate outcome indicators as set out in the PAD log frame, as follows23: PDO Indicators (i) Increase in beneficiary incomes. (ii) Reduction in migration among beneficiaries. (iii) Participation of the most disadvantaged people, particularly women and tribals, in sub-project activities and decision-making. Intermediate Outcome Indicators For Empowerment of Disadvantaged People (i) Percentage of CIG members belonging to poor and vulnerable groups. (ii) Number of beneficiaries with access to the banking system. For Supporting Village Governments Become More Effective and Responsive (i) Improved people’s perceptions of Gram Sabhas (GSs) and Gram Panchayats (GPs). (ii) GSs and GPs practicing participatory budgeting. (iii) Increase in access to clean drinking water by the poorest habitations. 2. Methodology. A baseline survey was conducted in 2005. It included 5,141 households in 257 villages across 16 districts in Chhattisgarh. The initial IE methodology aimed at using only baseline households for its end line samples. However, the IE team faced severe difficulties tracing down the baseline households. As an alternative, traceable households for both treatment and control group were selected from the baseline survey, along with additional randomly selected households. In the end, a total of 1,278 households were surveyed, among which 667 households were CIG members that directly benefitted from the Project; 302 households were non-CIG members but indirectly benefitted from the Project through village infrastructure; and 309 were control households that did not benefit at all from the Project. 3. The IE used both quantitative and qualitative research methods comprising: (i) household survey; (ii) 21 focus group discussions (FGDs) with different categorical groups such as women, SCs/STs, CIG members, and non-CIG members; (iii) semi-structured interviews with community leaders; and (iv) in-depth case studies documenting both successes and failures. Quality control measures were introduced and regularly monitored by the project team, the Bank and the ICRR team. The IE used the Difference in Difference method, which measures difference in pre-project and post-project outcome indicators for both project and control, and measures the difference between the two to arrive at project impact. Key findings summarized below were derived from both quantitative and qualitative analyses. 23 The IE does not cover the following four intermediate outcome indicators as they have been regularly measured by day-to-day project monitoring data: (i) number of sub-projects submitted, funded and implemented; (ii) number of families benefitted; (iii) average cost of sub-projects compared to government standards; and (iv) community contribution in cash. A. Key Findings PDO Indicators 4. Increase in beneficiary incomes. According to the IE, the annual household income for treatment group (CIG households) increased by 20% from Rs. 18,809.80 at baseline to INR. 22,635 at end line. This compares favorably against 4.3 % income increase of the control group. This finding coincides with the increase in overall annual household expenditure. The expenditure for treatment group (CIG households) increased by 80% from Rs. 22,810.40 to Rs. 41,201.78, as compared to a 58% increase for the control group. All groups who participated in the FGDs attested that the Project generated livelihood opportunities for them and that the biggest contribution of the Project has been in promoting self-employment as an alternative to wage labor or simply waiting for government jobs. Beneficiaries explained that the Project was particularly successful in scaling-up and adding quality to already existing economic activities (wood carpentry, metal work; etc); while activities that started afresh have encountered more challenges due to lack of strong forward or backward linkages. However, FGDs also revealed that about 300 livestock CIGs (1460 households) are forming a marketing federation. 5. Reduction in migration among beneficiaries. The IE found that overall migration rate dropped by 0.93% for the treatment group compared to the 2.3% increase for control group. Most notable reduction was the migration for 6-8 months where migration rate dropped by 1.14%. Contrastingly, migration rate for the control group increased overall, the largest increase in the 3- 5 months bracket by 0.61%. 19 out of 20 FGDs stated that migration for livelihood in the Project area dropped, and that the only migration is to urban cities for more lucrative opportunities rather than to meet their basic needs. This finding is supported by the fact that the treatment group managed to diversify their livelihoods after the Project. The percentage of both men and women engaged in three or more livelihood activities increased by 8.9% for treatment group, while for control group, it increased by only 6.2%. Moreover, the IE found that more treatment households were engaged in non-farm labor compared to the control households. 6. Participation of the most disadvantaged people, particularly women and tribals, in sub- project activities and decision-making. IE findings show that disadvantaged groups are participating in sub-project activities. Using households that have adopted saving habits as a proxy for improved livelihoods, IE data indicate that disadvantaged groups in the project areas are benefitting from the project more than the non-poor (OBCs) and more than their counterparts in control areas. There are 35% more saving households among project STs than among control ST households; similarly, there are 63% more saving households among project SCs than among control SC households; while there are fewer (-39%) saving OBCs among project households than among control households. This finding seems to be supported by the fact that household dependency on forest for livelihood has decreased from 12.6% to 5% in project areas. Dependency has also decreased in control areas but not to such an extent. As forest-dependant households are normally amongst the Poorest of the Poor, and collection of forest products is mainly a female activity, this is an indication of increased access to alternative livelihood opportunities for poor women. During FGDs, anecdotal instances of mis-targeting were reported, both in selection of beneficiaries and in the selection of project villages. In terms of decision- making, FGDs found that men came to accept the importance of women participation in decision- making over domestic and social issues through the Project’s participatory process. This finding is partially supported by the fact that 48.20% of women (members of CIGs) felt they have control over most of all decisions at end-line compared to 33.6% at baseline. These scores slightly higher than the control group where 42.4% at end-line felt they had more control compared to 35.8% at baseline. Similarly, the number of women perceiving they have no control over decisions has 43 decreased from 33.04% at baseline to 13.3% at end line (female members of CIGs) as compared to a smaller decrease from 25.73% to 21.1% in control areas. Women’s participation at GS meetings also increased by 20.6% for treatment group, though it fared poorly than control group’s 28.1% increase. From the FGDs, it also emerged that CIG female leaders have contested and won Panchayat elections, an indication of women empowerment through the Project. Intermediate Outcome Indicators for Empowerment of Disadvantaged People 7. Percentage of CIG members belonging to poor and vulnerable groups. According to the Project monitoring data, STs and SCs accounted for 37.0% and 13.0% respectively of the total Project beneficiaries. Women constituted 25.0% of the beneficiaries, and 20% of CIGs were women only CIGs. CIG households have a higher percentage of disabled members than no-CIG households in the end-line survey. However, FDGs revealed a perception that wealth ranking was not used for CIG formation or sub-project selection24, 8. Number of beneficiaries with access to the banking system. At baseline, the biggest source of loans was from friends/relatives, followed by private money lenders. For example for the treatment group, 54.2% requested loans from their friends/relatives; 25.0% from private money lenders; and merely 4.2% from banks for loans. At end line, 54.8% requested from banks; 19% from private money lenders; and 7.1% from friends/relatives. These data clearly show that beneficiaries have became increasingly more confident in accessing banks and that their dependency on money-lender has decreased. Control areas show a similar confidence in accessing banks; however reliance to money-lenders has increased from 6.25% to 27.9%. In terms of actual success rate in taking out loans from banks, 48.2% of the treatment groups were more successful than control group at 37.0% at end line. Treatment group’s primary use of loans was for agriculture (45.9%) compared to the control group’s health care (26.0%). This implies that treatment group came to borrow more for productive purposes at the end of the Project. Intermediate Outcome Indicators for Supporting Village Governments Become More Effective and Responsive 9. Improved people’s perceptions of Gram Sabhas (GSs) and Gram Panchayats (GPs). Participation of women, who were traditionally not actively, involved in GS meetings, increased from 19.6% at baseline to 40.2% at end-line for target group. More women also became aware of the activities undertaken by GSs, where 75.9% responded that they were well aware of the activities as compared to 72.1% at baseline. All respondents at FDGs attested that they were more aware of the GSs and GPs responsibilities and started holding them more accountable after the Project. However, respondents also revealed that the participation rate dropped once the Project was over, as people had little incentive to participate unless they were going to get some tangible benefits. Therefore, while the Project was successful in encouraging people’s participation, it could not have a sustained impact on people’s behavior. In terms of GPs’ responsiveness to community needs, people’s perceptions deteriorated for both treatment and control groups. The rate of approval of the treatment group dropped from 51.5% at baseline to 30.4% at end-line, while disapproval rate increased from 8.1% to 13.4%. Control group showed similar ratings. This signifies that GPs have changed little through the Project, and possibly became less responsive to their people as they have become more overloaded with various programs. 24 Wealth ranking is not used for sub-project selection but only to identify villagers that are eligible to become CIG members. 44 10. GSs and GPs practicing participatory budgeting. The Project provided some training to the GPs on the importance of participatory planning process and implementation. However, since there was no specific training on participatory budgeting, no evidence could be found that GSs and GPs were practicing it. 11. Increase in access to clean drinking water by the poorest habitations. The IE found little difference in changes in the distance to drinking water facilities between treatment and control groups, although both groups reduced the percentage of households that were 1km or further away from such source. This is probably due to various development programs that provided drinking water facilities across the State. While the Project did provide some sub-projects that improved access to water, the IE could not find any attribution effect of the Project. B. Key Recommendations 12. Quality of groups. To ensure the sustainability of the economic activities and a transparent, efficient and equitable functioning of the CIGs, it is essential that the groups acquire group management and enterprise management skills and that their level of maturity is regularly monitored. Groups should also conduct thrift activities to increase their access to finance but also to strengthen their group management skills and their institutional base. 13. Village-level infrastructure. FGDs emphasized the need to invest on productive infrastructure, namely irrigation and watershed, rather than community infrastructure, like roads and bridges. Irrigation and watershed management have a direct positive impact on people’s livelihoods and reduce their vulnerability to draughts and other climatic shocks. On the other hand, there is an abundance of funds already available at the Panchayat level for community infrastructure. 14. Introduce better targeting mechanism coupled with strong monitoring and redress mechanism. To ensure a more transparent and equitable distribution of benefits, villages and beneficiaries must be selected through a participatory selection mechanism based on vulnerability and wealth ranking. In order to realize this, there needs to be a strong third party monitoring mechanism where project facilitators as well as other stakeholders such as independent media or NGOs are present at each and every step of the project implementation. The Borrower and the Bank must also conduct surprise spot checks. A grievance mechanism needs to be established as well to allow beneficiaries to anonymously submit their complaints to the project staff. 15. Enforce stronger monitoring at every level. All 21 FDG groups confirmed that Project coordinators could not pay regular visits as they were assigned a large area and there were no village facilitators to follow up the activities on a day-to-day basis. Lack of manpower to ensure sufficient monitoring was also confirmed at the block level, where the Chief Executing Officer had the Project as one of his many responsibilities and therefore could not devote ample time for the Project. For a large-scale decentralized project to succeed, it is imperative that both male and female community representatives are recruited locally and trained as village facilitators to monitor the day-to-day activities, and people fully dedicated to the project are placed at higher levels to conduct spot checks. Project staff should be held accountable to conduct regular field visits and reporting. Monitoring information system needs to be strengthened so that it captures all the necessary quantitative and qualitative information throughout the course of the project. 16. Devote more time and resources to capacity building and institutional development. Beneficiaries pointed out that the Project’s focus was more on quantity rather than quality. CIGs were formed and given financial resources prematurely, which led to lack of people’s ownership 45 and unsustainable income generation activities in some cases. More than 70% of the FDG respondents pointed out that capacity building and skills training were not sufficient to make CIGs capable to run businesses. To avoid this, more time and resources should be devoted to capacity building and institutional development for CIGs to nurture ownership and enable them to undertake complex income generation activities. Due attention also needs to be paid to the government agencies at all levels, as it takes time to change one’s mindset. 17. Ensure sustainability of income generation activities with backward/forward linkages. Selection of appropriate income generation activities should be based on thorough analysis of the available resources, skills, markets, technical services, and access to finance. This should be backed by necessary training on technical as well as business management skills. Project facilitators need to be complemented by technical staff in helping CIGs develop their micro plans that are viable and sustainable. 18. Strengthen communications on Project information. To allow all beneficiaries adequate knowledge about the project activities and process, a dedicated communications team needs to be placed within the project team. The communications team will not only be responsible for disseminating project information, but preempt any unnecessary rumors or misunderstandings that could undermine the beneficiaries’ perception of the process. Communications team should be in charge of the beneficiaries’ complaints redress as well to ensure accountability and transparency. 46 Annex 6. Summary of Feedback from Beneficiary Workshops 1. During the ICRR Mission, a beneficiary stakeholder workshop was held in each of the four regions of Chhattisgarh covered by the project, to learn the views of beneficiaries with regard to the achievements of the project and what had worked less well. The workshops were facilitated by a local consultant facilitator (member of the ICRR team). CIG and some Gram Panchayat representatives were invited from a range of districts in each region, who in working groups of 5-10 members first discussed their opinions amongst themselves and then consolidated them for presentation and discussion. Close to 150 beneficiaries were involved in total. 2. The tables below categorize and rank the responses of the individual groups in plenary. The figures in the totals column correspond to the number of CIGs or other sub-group that identified the benefit or issue as important. The table below also shows in more detail the number of CIG of men (M), women (W), Gram Panchayat (GP), or mixed (U=unspecified) who considered a particular benefit or issue to be important, and thus how the totals were arrived at. The table below highlights those categories of benefits or issues that were identified by four or more groups as important. Table 1. Achievements of project to beneficiaries Jashpur Rajnand- Mahasa- Kanker Total gaon mund Economic benefits: Increase in incomes / profit / economic conditions U W, M W, Ux2 W, U, GP 9 Moving from un/employment to self-employment Ux2 GP W, Ux2 W 6 Increased employment U M U U 4 Better marketing (without middlemen) U 1 New enterprise activity (poultry) W 1 Community interactions: Improved awareness (about services etc), information U W, U Ux2, W, 7 GP Improved interaction with GS, GPs, Block and district U W Ux2 W+1, GP 7 councils Improved interactions with line agencies and businesses W, M 2 Improved interaction with banks U U 2 Improved household interactions: Increased respect for women’s roles / cooperation in family W, M Wx2, GP 6 Women more mobile in community W W 2 Community spirit: Improved community spirit / unity U M, GP GP 4 More dignity, respected, self confidence M U, W, GP 4 Working collectively U W, Ux2 4 Providing mutual support W 1 Household benefits: Children enrolled in (private) schools U Ux3 4 Reduced indebtedness U W, U 3 Started up HH savings W GP 2 Improved living standards Ux2 2 Increased ownership of assets M 1 Improved capacities and skills U 1 Improved nutrition U 1 More products for consumption (NTFP) U 1 Improved youth independence W 1 Increased access to health and education W 1 Communal benefits: Convergence to get other government services Ux3 W, U 5 Improved water supply for irrigation and drinking U U 2 Improved transportation access U 1 More available at village level M 1 Support to GP through Apna Kosh GP 1 Decreased nepotism GP 1 Better distribution of benefits U 1 Targeting free from BPL list GP GP 2 Interest free amount /grant GP 1 Table 2. Aspects which worked less well / issues Jashpur Rajnand- Mahasa- Kanker Total gaon mund Financial allocation: Financial support was less than expected (less than 150,000 U W, M Ux2 GP 6 entitlement) Allocation needed according to proposal U 1 Money not sufficient W U 2 Additional support to well performing groups could have M 1 been done Further funding would have been needed (e.g. on tubewells) GP 1 Contributions and disbursements: Should be lower contribution requirements (5%)/ 10% too W GP W Ux2 5 high Slow disbursement of installments (and between them), or not W Wx2 GP 4 on time Delayed approval of proposal U 1 Waiting for remaining amount /installment M 1 Sub-project implementation: Insurance information and claims weak W M W, U 4 Marketing of products still an issue /needs further support W U W, U 4 Further support to allied activities needed M U 2 Income based on seasonal activity so limited M 1 Financial problems M 1 Sub-projects not completed U 1 Training requirements: Need training before disbursement? W, U 2 Better quality timely trainings needed U GP 2 Difficulties in filing and maintaining books, inconsistencies M U 2 48 Project support and interactions: Poor relationships to project staff U 1 More govt supervision of NGOs needed / direct contact M 1 Review of CIG at GP level GP 1 Service standards too low (?) U 1 Supply driven choice of activities (e.g. dairy) GP 1 More mobilization and support to collective efforts U 1 Need more feedback from communities (like workshop) U 1 Better internal monitoring needed U 1 Beneficiary overage in village: Greater coverage needed GP 1 Sticking to BPL restrictive W 1 More participation from other people needed U 1 Despite wealth ranking some wealthy villagers benefited GP 1 Looking into the future, lessons learned: Waiting for 10% FDR / Apna Kosh U M U 3 The project should be renewed/ continued GP U 2 Need to support individuals W, M 2 Should not have been grant W 1 49 Annex 7. Summary of the Borrower’s Completion Report and Comments on the Draft ICRR Nawa Anjor: Chhattisgarh Rural Poverty Reduction Project Government of Chhattisgarh Panchayat and Rural Development Department Project Completion Report Chhattisgarh, a new state formed on 1 November, 2000 demonstrated its commitment to development of weaker sections and weaker areas, to poverty reduction, and women’s empowerment. Chhattisgarh was in reasonably good fiscal health compared to most other states and was focusing first on strengthening its basic institutional framework. The state had also committed itself to maintaining a lean administration and relied on the local governments for delivery of public services. As per the 1991 census data total population of Chhattisgarh was 1,76,14,928. 43% of the population lived below the poverty line (BPL). The tribal and Scheduled Caste population together comprised 57% of the BPL population. Largely backward and under developed the state of Chhattisgarh was rich in forest and mineral wealth. Rural areas of Chhattisgarh were characterized by poor connectivity poor access to electricity, safe drinking water and sanitation, high levels of migration, which worsened in periods of drought. While the Gender Ratio is the second highest in India after Kerala, female literacy was low (21% in rural areas) and infant mortality and maternal motility rate is high. Literacy was very low amongst tribal people (25%) and tribal women (11%). Tribal families live on the margin of the agrarian economy serving it with their cheap labor and skills. In order to have a balanced growth and development, Government of Chhattisgarh aimed at developing partnerships with funding organizations and Non-Governmental Organizations to reach grassroots level beneficiaries more effectively; promoting accountable Panchayats Raj Institutions that deliver pro-poor planning and effective services; greater empowerment of the poor, especially women and Scheduled Castes and Scheduled Tribes, through the formation of Self-Help Groups; a more effective investment in infrastructure to provide sustainable employment to the poor, and to reduce migration; and above all, improving delivery of education, health, drinking water, sanitation and roads. With a view to make a dent on rural poverty, Chhattisgarh Rural Poverty Reduction Project aided by World Bank was initiated in 2003. An apex society for the implementation of the Project was registered on 24 March, 2003. The implementation of the project began in the year 2004-05. Original project cost was INR 617.25 crore (SDR 94.461 Million) of which there was a loan part of INR 537.13 crore (USD 112.56 Million or SDR 82.200 Million). After 2004 Tsunami an amount of 13.264 Million SDR (approximately USD 20.7 Million) was transferred as Tsunami Aid, hence SDR 68.936 Million was left as available loan part from IDA. The project was implemented in 2,023 village Panchayats of 40 blocks, of all the 16 districts of Chhattisgarh. To improve the income opportunities for the target population investment was made in the sub-projects of livelihood activities. 20,446 sub-projects have been implemented, in which more than half are in the area of agriculture and livestock. Besides, 3,314 small infrastructural works have been implemented as part of Panchayat Plans and 192 Village Panchayat Buildings have been constructed in the project area. To improve the prospect of sustainability of sub-projects investments linkage with other government departments and banks have been initiated. Government departments have during the project period, imparted training to various groups involved in implementation of sub-projects. On the whole, the physical targets set at the beginning of the project in terms of coverage of families and, participation of the disadvantage Schedule Castes and Schedule Tribes communities have been met. The average cost of the sub-projects implemented is lower than the average cost of projects in SGSY, which is the main self-employment program. The average investment per family also is lower than the average investment in the SGSY. The accurate assessment regarding increase in income of the target beneficiaries and other social indicators in the project area can be made from end of project impact evaluation. Project Development Objectives The Project's Development Objective was to improve opportunities for the poor and vulnerable, especially women and tribals, to meet their own social and economic development objectives. To achieve this objective the project was to: 1. Create infrastructure and income opportunities for the rural poor Common Interest Group was the key organization of the poor in the project villages. A target of forming 20,000 CIGs in the coverage area of 2,000 villages was set. The target population was about 150,000 families among the most socially and economically disadvantaged of coverage area. Project was implemented in 16 districts and 40 blocks. A total of 2,023 Village Panchayats were covered against the target of 2,000 villages. 20,689 CIGs were formed as against targeted 20,000 CIGs. The 5 years of project period & one year of extension period, a total of 20,446 CIGs sub- projects (99%) were implemented. The incomplete sub-projects include such CIGs where due to groups falling apart and improper use of the fund provided the sub-projects could not be successfully completed. In 65 sub-projects, 10% community contribution could not be made as the concerned beneficiaries were unable to contribute the amount due to their very poor economic conditions. A wide range of activities was chosen by the beneficiaries themselves, which included agriculture, livestock, traditional skill & craft, trading activities etc. 1,11,728 families have been the beneficiaries of the CIG sub-projects. In a Centrally Sponsored Program SGSY, which is a self-employment, program targeting the rural poor and which has both subsidy & loan component, the average investment per family in Chhattisgarh is INR 34,000. A detailed analysis of various activities/ sub-projects undertaken in this project, shows that the average investment per family is approximately INR 18,000, which is lower than in the SGSY program. This is mainly due to the reason that CIG sub-projects investments supported by this project are of demand driven nature and largely decided by the community and approved by the Gram Sabha. Apart from this total 2,284 rural BPL unemployed youth were trained in different skills as computer operators, drivers, masons, electricians and it is being assured that these youth will get placements. End impact evaluation of the project is being carried out at present. Therefore an accurate evaluation regarding the increase of income of the CIG s can be made only after the completion of this study. As per the assessment reports of the District Project Supporting Units, the agricultural activities, fisheries, dairy in a few districts, trading activities, and the activities based 51 on local skill/craft and stitching/tailoring have an appreciable impact on the economic life of the families. In some locations the CIGs are moving to the next level & are in the initial stage of federation formation or large producers’ group. The economic activities are fisheries, dairy, NTFP, vegetable production & tailoring. 2. Empower active groups of disadvantaged people The project aimed at creating community infrastructure by giving grants to Panchayats. The initial amount of which was INR 3 lakh per Panchayat (@USD 5 per capita). The Panchayat plans in the form of village infrastructure & other public goods were decided by the Gram Sabha in a participatory manner. Main preference of the community was for cement concrete internal roads (60%) & culverts (8%), hand pumps (7%), extension of rural water supply system, bathing ghat for women (4%), electrification of tube wells for irrigation purposes, small common facility centre were the other preferences. In the newly created state of Chhattisgarh massive investment has also been made for creating all kinds of rural infrastructures. Therefore, the main assets created under the sub-component of Panchayat Plans in this project were related to the demands of those community members who were also benefited by the CIG sub-projects. Participation of Scheduled Caste and Scheduled Tribe population: the demographic profile of the 40 project blocks shows that 57% population comprises of SC and ST population. The beneficiaries of this category in the covered villages comprise 51% of the total beneficiaries. A total of 26,152 beneficiaries are women. Access to Banking of Vulnerable Section: in 26 project blocks all families have been provided access to banking system. Community Contribution: in the project the CIGs had to contribute at least 15% of the cost of the sub-project in cash (5% to the sub-project and 10% to a Village Fund) and GPs had to contribute at least 5% of the cost of the sub-project. The original target of expenditure for CIG investments was USD 64.79 Million (INR 309 crore). The expenditure on the implementation of all CIG sub-projects is INR 176.95 crore (USD 39.32 Million). The 5% group contribution (total INR 9.26 crore – or about US$2.06) at the commencement of the sub-project was contributed by the groups. Before the release of second installment 10% Village Fund (Apna Kosh) was deposited as Term Deposits. The total amount in the Village Fund is INR 17.92 crore (approximately USD 3.98 Million). 3. Support village governments in becoming more responsive and effective in assisting the most disadvantaged people Project has adopted participatory approach as a strategy to associate beneficiaries in the project right from very beginning and empower them. It was mandatory that the CIG sub-projects and the village Panchayat infrastructure are presented before the Gram Sabha and after being duly approved by the Gram Sabha it was presented before the project sub-committee at the district level. The district level project sub-committee is chaired by the president of district Panchayat and representatives of the concerned blocks. Also some members of Common Interest Groups participated in the district level sub-committee meetings. 52 The second installment for the sub-project was released only after the approval of expenditures of first installment by the Gram Sabha. Gram Sabha had right to monitor the activities of sub- projects. Implementation of the interventions planned by them with sense of ownership and accountability towards the performance of the program. Hence the involvement of Gram Sabha and District Panchayat in the finalization of sub-projects and village plans. Though the participation level in the Gram Sabha meetings do not show any remarkable increase (approximately 10%), in the various sub-projects’ planning and approval the Gram Sabha & District Panchayat played an important role. Project Components 1. Human Resource Development: A non-governmental organization CARD was given the responsibility for conducting capacity building programs of project staff. All the staff recruited were provided orientation programs. Five days orientation program was conducted for a group of 25 trainees each. Workshops were organized at State Institute of Rural Development (SIRD) for the entire project Chief Executive Officers Training. Refresher and vision building training courses in 8 Groups were organized for members of District Project Supporting Unit and Project Facilitation Team. District project supporting Unit and Project Facilitation Team members were trained on group management and vision building. Exposure visits were organized to foreign country of Indonesia and Sri Lanka and to Madhya Pradesh and Andhra Pradesh, Himachal Pradesh, Rajasthan. Training for district Panchayat presidents and Block Panchayat Presidents was organized at SIRD. Over 20,000 Panchayat functionaries were trained at SIRD. Training to CIG members was imparted on project management & accounting practices, technical training, capacity building of PRI members, training on sub-project planning, formulation and management was organized for the CIG leaders. Total 19,583 members got training in different economic activities. 2. Strengthening of Panchayats: In the newly created state many of the village Panchayats did not have their own buildings to conduct the business of Gram Panchayat. The project provided one building each in 192 Gram Panchayats. 3. Communications and information: To create awareness amongst target population, Kala Jattha, Chaupal, wall writing, posters, pamphlet, electronic & print media were extensively used. Literature on the project, its implementation mechanism, role of Gram Sabha, choice of beneficiary, collective planning etc., was distributed in the project villages for dissemination of information regarding project. Kala jattha is a popular mode of information dissemination. 27 kala jatthas were formed for this purpose & they proved to be very effective tool for communication & information. A popular program called “Chaupal” was broadcasted through all centres of All India Radio on every Wednesday. Good practices were documented in the form of small films in all the districts. For the dissemination of specialized techniques in the area of dairy, agriculture, fisheries, NTFP, handICRRaft & piggery, video films have been prepared and used. 4. Formation and Strengthening of Organizations: During the project period 1,11,728 members of families were mobilized to form 20,689 CIGs. 20% of CIGs comprised of “only women” members, towards the end of the project approximately 904 groups started mobilizing to reach the next level, federation. In some locations the mobilization also took place with groups and associations formed under different schemes but involved in the same activity. 5. Community Investments of CIG’s: During the entire project period 20,689 CIGs did collective planning. The project facilitation Teams played an important role in fostering the process. CIGs 53 selected their own activities and after formulating their sub-projects submitted them for the approval of Gram Sabha. The total expenditure on the implementation of all CIG sub-projects is INR176.95 crore (USD 39.32 Million). The 5% group contribution (total INR 9.26 crore) was contributed by the groups with first instalments. Before the release of second instalment, 10% of project cost was deposited as Village Fund (Apna Kosh) for future maintenance of village sub- projects. The total amount in the Village Fund is INR 17.92 crore (approximately USD 3.98 Million). 6. Panchayat Plans: The project aimed at creating community infrastructure by giving grants to Panchayats. The initial amount of which was INR 3 lakh25 per Panchayat (@USD 5 per capita). The Panchayat Plans in the form of village infrastructure & other public goods were decided by the Gram Sabha in a participatory manner. Main preference of the community was for cement concrete internal roads (60%) & culverts (8%), hand pumps (7%), extension of rural water supply system, bathing ghat for women (4%), electrification of tube wells for irrigation purposes, small common facility centres were the other preferences. Safe drinking water: under Panchayat plans 217 works for providing safe drinking water to rural habitations have been executed. These cover habitations where the average distance villagers had to walk to obtain drinking water was 1,500 meters which has been greatly reduced. It is remarkable to note that in Chhattisgarh one hand pump caters to approximately population of 85 whereas the national norm is coverage of a population of 250. Fund Flow: The project was budgeted in the Panchayat and Rural Development Department budget as an identifiable line item and this was continued for the subsequent years. Govt. of Chhattisgarh passed on the funds to the State Project Unit in the Panchayat and Rural Development Department through a treasury cheque. GoCG ensured adequate advance fund reserves with the State Project Unit during the entire project period. Total fund made available by was INR 291.52 crore. The State Project Unit retained a part of the funds to finance its own activities (costs associated with M&E, Special Studies etc.) and passed on the balance to the 16 Zilla Panchayats for project implementation. The Zilla Panchayats used the funds for activities at the district level, including sub-projects being implemented by Common Interest Groups and GPs. Both at the State Project Unit and at the Zilla Panchayat levels, the project funds were maintained in separate bank accounts earmarked for the project. In accordance with their proposals fund was directly transferred into Common Interest Group separate accounts. The account is opened in some nationalized/ scheduled bank in the name of Common Interest Group and is operated jointly by two members of the group. All the financial transactions were recorded in simple cashbook. 8. Village Fund (Apna Kosh): The fund was the fund created through Common Interest Group's contributions to cover operation and maintenance costs and further village development beyond the lifetime of the project. They would thus help ensure the sustainability of sub-projects. These Common Interest Group contributions would be deposited in the Village Fund bank account and would not be used to finance the specific sub-project investment they are associated with. Towards the end of the project it was decided to revise the guideline of village fund and use it as working capital/revolving fund for the federations of the CIGs. Before the release of second installment 10% Village Fund (Apna Kosh) was deposited as Term Deposits. The total amount in the Village Fund is INR 17.92 crore (approximately USD 3.98 Million). 25 1 Crore (1,00,00,000) is 100 Lakhs 54 9. Monitoring, Learning and Special studies: The in-built monitoring mechanism within the project was the monitoring of Common Interest Groups by village Panchayats and Project Facilitation Team and monitoring of Project Facilitation Team and village Panchayats by District Project Supporting Units and Zilla Panchayat. The sub-project plans were presented before sub- committee of District Panchayat, hence the overall monitoring by the District Panchayat. Such monitoring was done through the monthly reporting system and the regular review by the Chief Executive Officer, Zilla Panchayat. After the Mid Term Review of the project, in accordance with the Key Agreed Actions, a computerized management information system (MIS) has been developed and is being used so that as and when needed, the project progress is available at different levels and in desired format instantly. A comprehensive database management of MIS for the project was maintained in which the web-enabled technologies were introduced. Complete MIS was provided on the web site for monitoring various activities of the project up to village level. A website was developed with the domain name www.cgrurallivelihood.org where MIS data were updated regularly.  Baseline study was carried out in the year 2005-06 to measure current poverty situation in the project area through collection and assimilation of qualitative and quantitative information and then use the model to capture & compare changes in poverty situation after end of the project. A total of 3,742 project household & 1,399 control household total 5,141 household were surveyed.  Socio- Economic and Legal study of tribal land study was done in 2006 to generate a typology of various processes of tribal land alienation in the state and to explore their impact on people's livelihood.  A study on Performance Rating of Gram Panchayats in Chhattisgarh was done to develop a performance monitoring and rating system for decentralized local government institutions using Community Score Card Methodology.  A study on the Traditional Structures and Panchayat Local Governance was done to identify various informal & formal governance structure and system which influence conduct of village affairs and development, role in village development.  Cost Benefit Analysis was done to find out the activities which are economically viable and suitable to the poorest of the poor. Fifteen activities were chosen for the study; e.g. tube well, dairy, piggery, goatery, fishery, carpentry, tailoring, tent making, NTFP, weaving, bamboo & masonry. It was found that though the activities like dairy, fishery, Sambalpuri Sari and agarbatti making are the viable activities but not suitable for the poor, while weaving, NTFP, piggery, goatery, carpentry, bamboo & pottery are viable and suitable for the poor.  A study regarding strategy for working with poor and potential of cluster approach was done for exploring the potential of the cluster development approach to increase this impact. It is found that piggery, goatery, NTFP, bamboo, weaving, carpentry, masonry, pottery, tailoring, tent and tailoring are the most suitable activities for the poor. The activities with high potential for clusters are piggery, goatery, NTFP, bamboo, donapattal, bell metal & weaving with area specific approach. Environmental Audit Based on the environment strategy, project has developed an environment management framework (EMF) identifying the linkages between the project activities and the surrounding environment and laying down mitigation measures. The primary objective of the environment audit is to assess the extent to which the EMF has led to the implementation of appropriate environment mitigation measures in sub-projects. It is found that out of 75 sub-project activities 55 visited 38 sub-projects (51%) are following the mitigation measures prescribed. Environment friendly interventions are practiced in dairy, carpentry, paddy cultivation (SRI), vermin-culture, fisheries, NTFP, waste recycling, nursery, sprinkler activities whereas bamboo waste, fish pond for bathing purpose, un hygienic piggery sheds, use of chemical fertilizers in vegetable cultivation needs interventions for improvement. It is concluded that the sub-project activities are too small, dispersed and localized in nature that there is no opportunity for cumulative environment impacts. Project Administration The apex society for the implementation of project “Rural Poverty Reduction Society” was registered on 24 March, 2003 and the Poverty Reduction Project aided by the World Bank was started in June 2004. An Executive Committee was in place to carry on smooth and timely decision at state level under the Chairmanship of Principal Secretary, Panchayat & Rural Development Department and representative from Finance Department was also a member in the Committee. State Project Unit was responsible for the project implementation with Project Director at the head along with multi-disciplinary team taken on deputation. At district level, Chief Executive Officers were the Coordinator and a team of five specialists were hired on contract basis. At block level, along with Project Coordinator there were four other specialists. As per the Project Appraisal Document 20% of the project block was to be implemented by Non Governmental Organization. For this purpose 16 NGOs worked as Project Facilitation Teams. Financial arrangements 1. Funds Flow: The project was budgeted in the Panchayat and Rural Development Department's budget as an identifiable line item and this was continued for the subsequent years. GoCG passed on the funds to the State Project Unit in the Panchayat and Rural Development Department through a treasury cheque. GoCG ensured adequate advance fund reserves with the State Project Unit during the entire project period. Total fund made available by GoCG during the project period was INR 291.52 crore. Budget was released equivalent to the fund requirements on submission of the annual action plans by the State Project Unit after the approval of Executive Committee of the project. The State Project Unit retained a part of the funds to finance its own activities (costs associated with M&E, special studies etc.) and passed on the balance to the 16 Zilla Panchayats for project implementation. The Zila Panchayats used the funds for activities at the district level, including sub-projects being implemented by Common Interest Groups and GPs. Both at the State Project Unit and at the Zilla Panchayat levels, the project funds were maintained in separate bank accounts earmarked for the project. The Financial Management Manual and Operational Manual along with the sub-project agreements entered into between Common Interest Group /GP and Zilla Panchayat on approval of the proposal included information on the agreed financial management arrangements. Common Interest Group /GPs opened a separate sub-project bank account. The funds in this bank account were used for project related payments. Project through Zilla Panchayats deposited the 95% of the sub-project cost in two installments. The balance 5% was contributed by the beneficiaries. An amount equivalent to 45% of the cost of the sub-project was immediately made accessible to the Common Interest Group as first installment. The balance 50% of the amount was released as second installment on the confirmation of the Project Facilitation Teams, that the Common Interest Group has met the conditions as laid out in the MoU and is eligible to draw the 56 second installment from the bank. Of the total cost of the sub-project, beneficiaries contributed an extra 10% contribution towards Apna Kosh before the release of the second installment. 2. Accounting system: State Project Unit and District Project Supporting Units were centres of accounting. The data captured in accounting were in accordance with the cost tables prepared for the project so as to facilitate project monitoring and comparison of budgets with actual achievements. The chart of accounts categorized into two heads account head and activity code. These heads were defined at the commencement of project in accordance with disbursement categories. The accounting system was developed in such way that presentation of financial information is in clear, simple, user-friendly and consistent format, ensuring that all project- related activities were reflected in the project accounts and financial statements. There was need for timely preparation of preparation of project FMRs so that claim can be sent in time. Hence a computerization of accounts was done in Tally Software using double entry system. Accounting system also enabled an improvement in internal controls on advances. In the above context it can be mentioned that financial accounting system fulfilled all the requisite services as and when required. 3. Yearly & cumulative disbursement: Project displayed almost a linear trend of expenditure. Total expenditure during the project was INR 275.36 crore comprised INR 245.35 crore is loan part, INR 12.2 crore local community contributions to Village Fund and about INR 8.8 crore to CIG sub-projects, while the state contributed about INR 17.69 crore. 4. Category wise expenditure: Category wise expenditure is indicated in the table below: Expenditure SN Category (In INR Crore) 1 CIG 176.95 2 Panchayat plan 61.08 3 Innovation 0.06 4 HRD 3.86 5 Strengthening of Panchayat 5.42 6 Strengthening of PFT 12.24 7 Communication 2.54 8 Monitoring 2.31 9 Project Administration 10.89 Total26 275.36 5. Reimbursement Status: The bank would appreciate that project authorities were quite punctual in submitting the reimbursement claims. Till date claims of third quarter of financial year 2009-10, has been sent and reimbursement has been received. The FMR of last quarter is being prepared and will be sent to World Bank by 15 may 2010 for the approval. An amount of INR 83.68 lakh was deducted from the previous claim as this amount was mis-utilized by the groups. 6. Physical Verification of Assets: It was agreed in February 2009 mission that a physical verification of assets of project would be done by an independent agency. Accordingly physical 26 CIG (line 1) and Total figures do not include INR 8.8 community contributions to CIG sub-projects. 57 verification was carried out by treasury officers of the respective districts and the reports were shared with the bank in their follow up visits. 7. Sub-project Audit: As per requirement of the project agreement, the project auditors (a panel of chartered accountants) were appointed to audit the accounts of Common Interest Groups. However due to slow progress of audit work it was agreed in the February 2009 Mission that only 20% those Common Interest Groups will be audited which get their cc in the current year. Project made quite a good progress in audit of sub-projects. Out of total 24,236 sub-projects (20,689 Common Interest Group + 3,355 Infrastructures + 192 Panchayat buildings), 13,020 sub-projects (53.72%) were audited. Factors affecting implementation Chhattisgarh a newly created state (in 2000) was in its infancy when the project was started and was still to overcome many of the teething problems. It comprised of that part of erstwhile Madhya Pradesh, which was largely backward & underdeveloped. Though rich in mineral wealth and forest, the livelihood potential of state was not tapped. Chhattisgarh faced dearth of skilled human resource and huge vacancies in all the sectors of governance. Hence setting up of new structure at the state, district and block level and recruiting able staff was a big challenge. The qualified technical staff available at the local level was not sufficient enough to fulfill the expectations and goals of any major developmental initiative. The resource centre for rural development, State Institute for Rural Development though started in 2001, could not begin functioning well until 2005 and recruitment of faculty and staff could only take place towards end of 2007. Hence there was very limited faculty to provide training and orientation. On creation of new state there were huge vacancies in all departments of government. All the Project Facilitation Team at block level required support and synergy with the line departments. Vacant posts at many levels created difficulties in providing such support to Project Facilitation Team and Common Interest Groups. Since there were no linkages with the technical departments in the initial stage, the activities selected by the beneficiaries could not get adequate guidance. Since human resource strategy could not be formulated there were frequent turnover of staff, which led to lack of inconsistency in approach and facilitation at the field level. In state like Chhattisgarh which was new and which was largely a backward and under developed area of erstwhile Madhya Pradesh any project or program faced is an initial hurdle due to social and cultural factors those restrict women to come forward and take part in formal deliberations in any forum. Linkages with the line departments and convergence with other schemes and programs was initiated only in the last year (extension period) of the project. The project implementation was also affected during the conduct of the elections, general as well as local. Borrower's Learnings from the Project The experience of implementation of a poverty reduction initiative for six-years provides a number of lessons for development projects in general and poverty alleviation initiatives in particular. Some of the key learning emerging from the project is briefly summarized in the following: 58 1. Wealth ranking to identify poorest of the poor and the vulnerable is a good tool for any such initiative. However the lesson drawn from this project shows that utmost care should be taken to conduct this exercise. The execution of the entire project is based on the correct identification of the target population. Hence more time, and preparation is needed for this kind of ranking, as a good preparation leads to better implementation of such initiatives. 2. Importance of early social mobilization and capacity building: the project experience shows that social mobilization and capacity building is process oriented, and therefore, takes time to start-up but produce good results if given adequate times. It implies that project planning should emphasize social mobilization as early in the project cycle as possible. 3. The project has adopted a demand driven approach: the activities were selected by the group members. Project Facilitation Team did not decide about the activities or about the composition of the groups, but in turn guided the groups. This approach provides more space to beneficiaries and thus leads to enhancement in their capacities. However in an area were plenty of livelihood opportunities are available, strong and robust technical support in the implementation of the selected sub-projects is also a requirement therefore it is essential that a good synergy should exist not only between the beneficiaries and the project facilitator teams but also between them and all the concerned government departments. 4. Design of the project: the sub-project submitted by the Common Interest Groups are funded in a very transparent and time bound manner. A provision is made in the project that 95% amount of the total cost of the sub-project will be deposited in the account of the Common Interest Group in single or two installment as per the requirement and deposition of Apna Kosh money. Though the design of the project was simple and well understood by all the stakeholders, it is felt that for providing fund for the implementation of sub-projects, a judicious mix of both subsidy and loan component should be there. This will teach the beneficiary to develop repayment ethics as well as to own the sub-projects better. 5. In a few cases the cash contribution as a condition for release of second installment became difficult because of the very poor economic condition of the concerned beneficiaries. Therefore during the implementation of the project it is felt that at some stages during the periodic reviews this could have been reexamined. 6. An efficient Management Information System needs to be in place at the state, district and block level, from the very beginning of the project. Rather it can be considered a pre-requisite before launching any such developmental project. Also the members of Project Facilitation Team and members of District Project Support Unit should have been given adequate training to be able to handle the MIS. This would have ensured smooth and regular availability of detailed data at every level and thus could have ensured a quick and meaningful monitoring. 7. A detailed review of the implementation progress and the assessment of the constraints as well as the strength of it, at the block levels and district levels, periodically by the State Project Management Unit together with the World Bank team or otherwise, could have yielded much better results. 8. Different approaches and different strategies to respond to the progress in implementation, in different blocks/districts could have been a better alternative. Though the design implemented was the same throughout the covered areas, the progress the strengths and weaknesses visible in different areas and communities show that once the project was started and the first dose of support was given, an analysis of the initial results sector wise and area wise should have been 59 done and accordingly different strategies could have been decided to provide support in the remaining period of the project. This would have ensured consolidation of the initial gains during the implementation of the project itself; also such grading can help in deciding the future doses of support to the community through any other scheme, project or program. 9. There was often close monitoring done by the state government and the World Bank. Nevertheless it would have been appropriate to set quarterly and yearly milestones for proper implementation. Since the design of the project was new and different from the poverty reduction initiatives implemented by the government under different programs, different problems, constraints were faced by the different stakeholders. It is more appropriate for any such project to revisit it at certain intervals in order to make it more workable and practical. Other Recommendations: The Supervision and Implementation Support Mission of the World Bank especially from the second year of the implementation could have reviewed the structure and the design of the project and made necessary changes to restructure it to suit the local requirements and demand. Particularly the Mid Term Review by the World Bank Team was an important opportunity to modify the Project Development Objectives and Monitorable targets of the project as two years of implementation of the project and the results thereof could provide some indication towards the constraints faced by the stakeholders, the capacity of the available human resource and the progress in terms of making the Village Panchayats more effective and responsive. A detailed review of the implementation progress and the assessment of the constraints as well as the strength of it, at the block levels and district levels, periodically by the state project management Unit together with the World Bank team or otherwise, could have yielded much better results. Different approaches and different strategies to respond to the progress in implementation, in different blocks/districts could have been a better alternative. Though the design implemented was the same throughout the covered areas, the progress the strengths and weaknesses visible in different areas and communities show that once the project was started and the first dose of support was given, an analysis of the initial results sector wise and area wise should have been done and accordingly different strategies could have been decided to provide support in the remaining period of the project. This would have ensured consolidation of the initial gains during the implementation of the project itself; also such grading can help in deciding the future doses of support to the community through any other scheme, project or program. Workshops and brainstorming sessions at the focal point of implementation i.e. District level could have been a good idea to assess the gaps in terms of achieving the goal of economic and social empowerment of the covered beneficiaries. It could also have provided an opportunity to assess the availability and capacity of the human resource involved in the support at district level and the project facilitator's team at the block level. The process of procurement within the project required no objection from the work bank Task Team Leader (TTL) at various stages. During the hiring of the agencies obtaining no objection from the World Bank TTL took very long, which reduced the total time available for completing other formalities and sometimes left no time for the project to actually undertake that particular activity. Hence, it is suggested that a member of the World Bank team may become a member of the procurement team at the state level, which will simplify and expedite the entire process. A 60 resident consultant of World Bank was placed in Raipur to provide support to the state team and was to act as a liaison between the state team and the World Bank. However, did not effectively contribute towards achieving this. Good practices, Sustainability and the Way Forward Though there was no arrangement or planning in the project implementation plan regarding convergence of technical knowledge, finance or provision of credit with other government schemes and programs, yet in the later stages of the implementation some convergence initiatives were undertaken for sustainability of the Common Interest Groups and consolidation of the gains. 411 Common Interest Groups have been linked to the banks and credit facility under different government programs like SGSY, Khadi & Village Industries Board, Fish Farmers Development authority. 2,800 cases have been submitted to the banks to access credit in order to fund the working capital requirements and upgrade their investments. More than 50% of the implemented Common Interest Group sub-projects are in the areas of land based and agricultural and livestock activities. With the support of the government department of agriculture and veterinary it has been initiated to provide more than 10,000 Common Interest Groups with further training, skill upgrading, inputs and implements. With the support of this sector prospects for sustainability of half of the sub-projects investments under this project will be enhanced. Yet another form of convergence has been initiated in the form of providing small physical infrastructures, which are helpful in facilitating and supporting the income generating activities of the groups. In very few locations demands are being created from the Government departments implementing public welfare schemes. One such example is the involvement of women Common Interest Groups involved in tailoring and stitching are able to take work orders for stitching of school uniforms for government schools. Since the first year of the implementation of the project certain well performing Common Interest Groups who have been able to increase their income and upgrade their activity, have been participating in the regional Saras Melas and other exhibitions. Such exhibitions and Melas have been able to provide forward linkages to these groups. This has also helped the Groups to increase the volume of production of their goods Gradual mobilization of similar economic activity based Groups is taking place in some project areas. The project supported Common Interest Groups are mobilizing with Groups formed under other departments and schemes and societies farmed for achievement of specific economic goal e.g. fisheries, dairy, NTFP collection and processing. The Common Interest Groups are in the initial stage of federating as production Units in 6 districts namely Mahasamund for dairy with 553 Common Interest Groups, in Dhamtari for fisheries with 124 Common Interest Groups, in Kanker for NTFP with 100 Common Interest Groups, in Rajnandgaon for tailoring with 55 Common Interest Groups, in Janjgeer Champa for vegetable production with 22 Common Interest Groups and in Bastar for tailoring and stitching with 5 Common Interest Groups. 61 Annex 8. Comments of Cofinanciers and Other Partners/Stakeholders NA Annex 9. List of Supporting Documents 1. Project Appraisal Document on a Proposed Credit to India for Chhattisgarh District Rural Poverty Project, March 24, 2003 2. World Bank Supervision Mission Aide Memoires 3. Mid Term Review and Implementation Support Mission (March-June, 2007 Missions) 4. Project Implementation Plan Department of Panchayat and Rural Development Government of Chhattisgarh 5. Baseline Study for “NAWA ANJOR” Chhattisgarh District Rural Poverty Project, June 2006 6. Borrowers’ Completion Report, May 2010 7. Final Report of Impact Evaluation Survey, February 2011 8. Implementation Completion and Results Report Madhya Pradesh District Poverty Initiatives Project, December 30, 2008 (ICR00001029) 9. Economic and Financial Analysis, April 2010 10. Gram Panchayats and Infrastructure sub-projects ICR Thematic Study April 2010 11. System of Rice Intensification (SRI) ICR Thematic Study April 2010 12. CIG Federations ICR Thematic Study April 2010 13. Livelihoods ICR Thematic Study, FAO, April 2010 63 IBRD 32254 60° 62° 64° INDIA Son STATE OF CHHATTISGARH DISTRICT RURAL Ka POVERTY PROJECT 24° nha 24° r HIGHWAYS Ramanuj Ganj JHARKHAND MAIN ROADS RAILROADS SURGUJA SELECTED CITIES KORIYA DISTRICT CAPITALS Nagar Son STATE CAPITAL Jainagar Baikunthypur - Ambikapur DISTRICT BOUNDARIES STATE BOUNDARIES Ramgarh Kunkuri JASHPUR ELEVATIONS IN METERS: Jashpur 1000 500 KORBA Katghora 200 Dharmjaygarh BILASPUR Korba Kota RAIGARH Pandara Gharghoda - Bilaspur Janjgir 22° 22° - Kawardha JANJGIR Raigarh MADHYA PRADESH KAWARDHA Manha di na nath Seo RAIPUR DURG UND Durg RAIPUR MAHASUM - Mahasamund - - Raj-Nandgaon RAJ- NANDGAON Chhura Kusumkasa Dhamtari DHAMTARI MAHARASHTRA - Bhanupratappur - Kanker ORISSA KANKER Bardula 20° 20° - - Narainpur BASTAR I Jagdalpur Kutru Dhantewada - Bhopalpatam G da ari DHANTEWADA o v ANDHRA PRADESH Sukma - Chintalnar ri Saba 18° 18° This map was produced by the Map Design Unit of The World Bank. The boundaries, colors, denominations and any other information Konta 0 25 50 75 100 Kilometers Bay of Bengal shown on this map do not imply, on the part of The World Bank Group, any judgment on the legal status of any territory, or any endorsement or acceptance of such boundaries. 0 25 50 75 Miles 60° 62° 64° MARCH 2003