Andhra Pradesh Tribal Development Project (2001)


Completion Evaluation


This document being submitted for the consideration of the Evaluation Committee contains two distinct sections related to the completion evaluation of the Andhra Pradesh Tribal Development Project, India. The first section contains the evaluation's Executive Summary, and the second section includes the Agreement at Completion Point that is a mandatory feature of IFAD evaluations. This Agreement at Completion Point was reached among various evaluation partners on 30 April 2001 in Hyderabad. It is based on an in-depth analysis and reflections on the main evaluation findings. The Agreement at Completion Point is the product of a learning exercise and an intense dialogue among the members of the Core Learning Partnership.1 It consists of five key evaluation insights.

Abbreviations and acronyms

APTDP Andhra Pradesh Tribal Development Project
GCC Girijan Cooperative Corporation
ITDA Integrated Tribal Development Agency
NGO Non-Governmental Organization
NTFP Non-Timber Forest Product
PRA Participatory Rural Appraisal
PWG People's War Group
SHG Self-Help Group
VTDA Village Tribal Development Association

Section one: executive summary


Project background. The Andhra Pradesh Tribal Development Project (APTDP) was appraised by IFAD in December 1990. The loan became effective on 27 August 1991 and closed on 31 March 1999. The total project cost was USD 46.5 million, funded by the Governments of India/Andhra Pradesh (USD 19.5 million), IFAD (USD 20.0 million) and The Netherlands (USD 7.0 million). The cooperating institution was the United Nations Office for Project Services. The Tribal Welfare Department of the Government of Andhra Pradesh was the executing agency, with overall responsibility for project implementation.

Project area. The project area is located in the north-eastern regions of the state of Andhra Pradesh. The project was implemented in four contiguous Integrated Tribal Development Agencies (ITDAs), with a high concentration of families engaged in podu (shifting/slash-and-burn) cultivation: Seethampeta (Srikakulam district), Parvathipuram (Vizianagaram district), Paderu (Visakhapatnam district) and Rampachodavaram (East Godavari district). The project area comprises wide plains, hills and narrow valleys. The altitude ranges from 200 m to as high as 1 600 m in some areas of Paderu, and the area contains a wide range of microclimates. The areas where podu cultivation is practised often have steep slopes in excess of 28 degrees. Rainfall averages about 1 100 mm per year, with around 80% of precipitation occurring during the south-west monsoon (June-September). The project area is endowed with a generally good climate, productive soils, rich biodiversity and is subject to less demographic pressure than exists on the plains.

Target group. At the time of project appraisal, the entire project was inhabited by scheduled tribes. About 70% were engaged in podu cultivation, and the majority of the remainder relied on rainfed crop production. The degree of poverty was greatest among the 13% of households with only podu land, and average incomes per year, including off-farm income, were estimated at 2 660 Indian rupees (INR) for these households. The comparable figure for families with both podu and permanently cultivated land was INR 3 710. Both figures were well below the annual poverty level of INR 4 800 per household used by the Integrated Rural Development Programme. Especially vulnerable groups include significant numbers of landless households, the poorest women and under-employed youth. About 60% of tribal families were reported to be in debt to non-tribal traders/moneylenders, with an average debt of INR 1 390. The target group comprised all 63 370 families living in the 2077 villages of 16 selected watersheds.

Objectives and strategy. The main objective of the project was to foster self-reliant household food security by increasing food production and raising the income of tribal families, with specific focus on households practising podu agriculture. The project strategy included several elements: (a) planting the hill slopes with tree crops to provide food and cash income, as well as utilizing soil conservation measures to restore the ecological equilibrium; (b) improving the productivity of food crops through expanded irrigation and improved farming technology; (c) formulating a community development strategy to ensure the sustainability of economic development; and (d) identifying measures to address the chronic indebtedness among tribals by strengthening credit and marketing systems and, in particular, by building up the assets and capacities of the Girijan Cooperative Corporation (GCC).

Project components and implementation. The APTDP had the following components: (a) community and women's development (6% of project costs), (b) health and education (9%), (c) natural resources development (70%), (d) credit and marketing support (9%), and (e) project management support, including monitoring and evaluation (6%). The Tribal Welfare Department had the overall responsibility for project implementation, with the commissioner for tribal welfare as the project coordinator. Project activities were implemented through the regular ITDA channels. The respective ITDA project officer was designated as project director and was responsible for day-to-day management of the project.

The political factor: Naxalism. The main risk involved in the implementation of the APTDP was the history of socio-political unrest in the region. During the 1970s and 1980s, the project area witnessed a period of turbulence as the epicentre of the Naxalite revolt, a radical insurgency movement. The initiation of the project represented a bold move on the part of both the Government of India and IFAD and provided the opportunity to assess the relationship between a grass-roots-led protest movement and the operation of a participatory tribal development project.

Completion evaluation process. The main objective of the completion evaluation was to assess the performance and impact of the APTDP. An additional objective was to document the experiences of the project and develop a series of lessons learned that could assist in the design and implementation of similar ongoing and future projects in the country and elsewhere. The evaluation was jointly conducted with The Netherlands, the project's cofinanciers, and was planned and implemented to promote maximum local participation and ownership.

A participatory rural appraisal (PRA) was undertaken from 15 September to 15 October 2000 by Outreach (a non-governmental organization (NGO) based in Bangalore, India). The objective of this exercise was to collect primary data from project stakeholders in order to assess the performance of the project from the beneficiaries' perspective, including its targeting, impact and sustainability. The results and analysis of the PRA were discussed in a stakeholders' workshop in Hyderabad in October 2000 and were made available to the evaluation team.

The completion evaluation team was then fielded in November 2000. It held discussions with representatives of the Indian Government and the Embassy of The Netherlands in New Delhi before proceeding to the project area, where discussions were held with officials of the Government of Andhra Pradesh, NGOs and other project participants. Two workshops representing major stakeholders were held at the outset and at the end of the mission to create a forum for consultation, dialogue and knowledge-sharing. Mission members spent about two weeks in the four project areas, during which they were able to interact directly with beneficiaries in 35 villages and gain valuable insights from their experience. They also examined some of the physical achievements of the project and participated in in-depth discussions with key project staff. A video conference (Delhi/Rome) was organized during the evaluation mission's wrap-up meeting in New Delhi. This provided concerned IFAD staff from the Office of Evaluation and Studies and the Asia and the Pacific Division with the opportunity to participate in discussions. Finally, a stakeholder workshop was held on 30 April 2001 in Hyderabad to discuss the draft evaluation report and to formulate the evaluation's Agreement at Completion Point.

Implementation performance

Natural resources development. The physical results under the natural resources component have generally matched or exceeded targets. Total food production in the project area has increased by an estimated 500% during project life. However, these figures mask important problems, such as the need for better maintenance of irrigation systems, a lack of sufficient know-how regarding horticultural techniques and, perhaps most significantly, an increased vulnerability to drought because of the dependence on irrigated agriculture at the expense of traditional techniques that included built-in measures to counteract periodic drought conditions. When questioned, the farmers admitted that they were continuing podu cultivation as a fall-back mechanism, or that they would revive it if necessary.

More specifically, about 20 000 ha of rainfed lands were provided with irrigation, representing an increase in the irrigated area of six and a half times as compared to the pre-project period. More than 54 000 ha of land belonging to some 27 000 households were subjected to soil and water conservation measures. More than 55 000 families now directly benefit from highly productive horticultural plantations and orchards, where the emphasis is on mixed plantings of cash and food crops. A most significant achievement has been the establishment of nearly 40 000 ha of orchards, mostly cashew, but also mango, guava, citrus and silver oak for coffee. Under the APTDP, in addition to greater irrigation and soil conservation methods, determining factors in enhancing food production were the development of improved seed varieties, the wider application of fertilizers and the adoption of improved agricultural practices.

Agricultural extension is provided by village extension officers. The outreach capacity of the village extension officers is constrained by the lack of training necessary for the dissemination of innovations and technologies, and of transportation facilities. To overcome these constraints, some support has been provided by village-level workers and the posting of agricultural graduates as agricultural development consultants in villages that are not easily accessible. The APTDP achieved its target of establishing 6 500 demonstration plots and 2 400 seed-production sites. About 900 farmers were given training to enable them to establish satellite nurseries. Adaptive research was initially made available through Andhra Pradesh Agricultural University, and since 1995 the responsibility has been entrusted to the National Institute of Agricultural Extension Management. Research focused on cropping systems, inter cropping experimentation, natural resources management and so on.

Community participation and village institutions. Participatory approaches designed to motivate and empower men and women have been a key process in project activities. The APTDP established a variety of local-level institutions, including self-help groups (SHGs), cluster-level associations of SHGs, user groups/village development committees (for example, for education, health, irrigation, grain banks) and a nodal institution in the form of village tribal development associations (VTDAs). The latter were conceived on the one hand as a forum for the expression of community priorities and concerns and on the other as a means of delivering project and programmes to the community. A total of 1 029 VTDAs were formed in the project area, fives times the original target. Similarly, 1 231 SHGs were formed. However, more than half of the SHGs are inactive, most likely due to the fact that they can be set up easily, but are difficult to sustain. The cohesiveness and sustainability of groups have been weak, and too much emphasis has been given to the mere existence of SHGs. A total of 467 grain banks were established. They have been the most significant of village-level institutions and have enabled members to do without the services of moneylenders for the greater part of the year. Community coordination teams - groups of young, dedicated professionals who live in villages for up to three years - have had notable success working with villagers by encouraging genuine participation in helping tribal communities identify village priorities and implement and monitor development activities in the sphere of health and education.

Education. The principal activities in education focused on infrastructure and access improvement, training, awareness-building and performance monitoring. The support for community initiatives included the provision of matching funds for the construction of community (maabadi) schools and for the payment of their teachers, as well for basic equipment such as blackboards, stationery and slates. Support was given to 1 323 community schools, covering nearly 20 000 schoolchildren. A total of 81 educational resource centres were constructed. These were used for ongoing teacher training programmes, access to educational materials and teacher conferences. Training sessions included teacher training, programmes for community school volunteers and guidance for village education workers. Awareness campaigns were mounted to emphasize the importance of education and increase enrolment in primary schools and to reduce drop-out rates. Through this component, the project also ensured the provision of midday meals and the undertaking of a comprehensive survey in terms of access, capacity and enrolment projections, as well as the rationalization of educational institutions and school complexes through a scientific school-mapping exercise.

Considerable emphasis was given by the educational resource centres to improving teaching methods and monitoring educational standards. The basic model of these centres included a training hall equipped with audio-visual facilities, a library, a laboratory, a kitchen and some accommodation for visiting teachers. The centres normally served 30-40 schools of all kinds under the aegis of the head teacher of the school that hosted the complex.

Health. Considerable emphasis was placed on the health sub-component at the design stage, but, as with education, little was achieved in the first years of the project. The programme focused on the promotion of community-based preventive health care. The principal objectives of the health sub-component were to promote accessibility to primary health care and monitor mother and child health/epidemics, as well as to raise the awareness of tribal people and capacity-building among medical and paramedical staff and tribes. Another key element in this initiative was the deployment in remote villages of well over 1 000 community health workers, each of whom was provided with a month's intensive training and a basic medical kit. The basic training focused on hygiene, malaria prognosis and first aid. Community health workers were selected from among married women in their twenties and thirties, preferably with a modicum of education. Other major achievements included the provision of drugs, equipment and vehicles for 32 primary health centres and of vehicles for mobile medical units; matching grants for the construction of 181 subcentres; the establishment of a referral fund for medical emergencies; and the provision of training kits. Thirty-seven jeep-cum-ambulances were supplied to primary health centres. Training sessions were provided for medical officers at the Indian Institute of Health Services in Hyderabad, and drugs and other equipment were supplied to primary health centres and subcentres.

Housing. Upon the recommendation of IFAD, the Tribal Housing and Habitat Improvement Facilities Programme was proposed as an additional activity in 1997-98, and the loan agreement was accordingly amended. As podu lands were converted to orchards and plantations and forest regulations were more strictly applied, the growing shortage of roofing and thatching material provided the rationale for inclusion of the housing programme. The basic objective was to provide low-cost pukka (bricks) housing to tribal villagers. A total of 14 292 houses were constructed, only eight short of the target. The housing programme was to be funded through a grant to VTDAs to be distributed among villagers as loans. In the majority of villages, however, payment for the houses has been regarded as a grant and not as credit, although beneficiaries have supplied the labour or labour payments and, in many cases, additional materials. The procedures for selecting beneficiaries under the housing scheme were not consistent, and the involvement of beneficiaries in design was limited.

Credit and marketing. The project provided resources to improve the operational capacity of GCC, in particular its ability to manage credit operations, improve marketing and support research and development activities that would benefit tribal people.

The project's performance in credit delivery was inadequate. Crop loans were provided both in cash and in kind, with approximately 59% being distributed in cash, and the rest in the form of fertilizers and other agricultural inputs. Consumption loans were also provided. SHGs were the principal vehicle for the channeling of credit. About 60% of the SHGs formed during the project period were inactive. In the six villages studied during the PRA, two SHGs were totally defunct, and the four remaining could be categorized as functioning moderately. It was noted that SHG activities were better in the initial years and gradually worsened. The initial enthusiasm is partly explained by the provision of matching grants to the SHGs by the project irrespective of their performance.2

In terms of credit delivery, the amount disbursed more than doubled from 1991-92 to 1992-93 (from INR 6.6 to 13.6 million) at the time that IFAD funding to GCC became available, but then gradually fell, reaching its lowest level, INR 2.2 million, in 1996-97. It picked up slightly in the next two years, but had still only reached INR 4.9 million by 1999-2000. The trend was similar in terms of loans in the form of agricultural inputs. From a 1991-92 level of INR 2.3 million, the total amount of such loans increased to INR 6.4 million the following year and thereafter began to fall sharply, until in 1998-99 no loans at all were disbursed, and only INR 2.1 million was disbursed in 1999-2000. One of the reasons given by GCC for the reduced levels of credit was the sharp decline in recovery rates. The GCC figures indicate about 60% recovery in 1999-2000, which is more than the rate in some earlier years. For example, in Paderu district the collection rate declined steadily from 88.7% in 1994-95 to 39.3% in 1999-2000.

In recent years, GCC has put more emphasis on marketing than on the provision of credit. With regard to marketing, GCC took a number of steps in organizing grass-roots channels, strengthening the supervision of procurement and sales and developing new products. In particular, the reach of the Girijan Primary Cooperative Marketing Societies was greatly extended. VTDAs organized periodic meetings of villagers to discuss who needed credit and how to improve the marketing of produce and obtain better prices. The project also provided training for functionaries of the Girijan Primary Cooperative Marketing Societies and considerably improved methods of processing, packaging, transportation and promotion.

Research and development activities constituted a major aspect of implementation, and GCC has tackled these activities systematically by commissioning market surveys and studies on processing techniques, financing opportunities and pricing, and undertaking the regular monitoring of the activities of tribal people. This has led to the identification of new non-timber forest products (NTFPs) and the creation of a database on forest products and their use. The emphasis on research and development has supported a process of synthesis between modern techniques and the tribal people's indigenous knowledge systems and has boosted efforts to market forest produce more actively. GCC initiatives have generated value-added items derived from many forest products such as gum karaya, gum olibanum, cleaning-nuts, marking-nuts, jatropha curcas, gymnema leaf and neem. The generation and processing of gum karaya have been particularly successful, providing much employment to tribal women. Gum karaya has been developed in four forms, powder, granule, cream and gel, with numerous usages promoted commercially. Cleaning-nuts are another item of technical processing, which has led to the production of Natfloc-1001 with a variety of functions in the clarification of water, including in large-scale industrial operations. These and many other familiar items, such as soap, honey and washing powder, have been developed by the research department of GCC.

The overall value of the forest and agricultural produce purchased by GCC rose sharply to a total value of INR 220 million in 1999-2000, compared to an average of about INR 100 million during 1992-96. The collection and sale of gum karaya are an increasingly important part of GCC operations, with top-grade gum being purchased from villagers at INR 125 per kg, a fivefold increase during the project period. The GCC organizational expansion facilitated by the APTDP led to the creation of the Commercialization, Research and Development Division, illustrating the emphasis GCC assigns to issues related to market linkages.

Project management and coordination. Project implementation was marked by frequent changes in management personnel in the Tribal Welfare Department. Nevertheless, the overall project coordination by the Tribal Welfare Department was effective. ITDA project officers were particularly enthusiastic in their efforts to institutionalize participatory approaches, and their efforts culminated in the Government of Andhra Pradesh issuing an order requiring that all works in tribal subplan areas should be executed by village SHGs. One major problem has been the fact that intermediaries between ITDA and the grass-roots level are limited, which has not facilitated project operations. Project supervision by the United Nations Office for Project Services was highly incisive and constructive, and the appointment of the same cooperating institution to supervise the follow-up project (the Andhra Pradesh Participatory Tribal Development Project) has provided good opportunities for the transfer of experiences between the two projects.

Analysis and impact

Food security and natural resources. The combined effect of better agricultural practices, input supply, horticultural initiatives and infrastructure interventions has led to marked increases both in productivity and in total production and thus a significant improvement in food security and income levels. The PRA reveals that the food security of selected households across the four ITDAs had improved by 20-30% by project year six. More specifically, at project inception only six households out of 24 enjoyed food security for nine months or more of the year, and this figure had increased to 18 by project completion. For the four poorest households, which were described as enjoying only three months' food security initially, the period of food security had doubled by the time of the PRA. In the village of Diggavu Solamalu, in Paderu, there were seven households in the pre-project period that had a full 12-month food security, and this figure had doubled to 14 by project completion. The number of families enjoying at least nine months' food security increased from 12 to 17. Generally, those households that were less food secure did not possess irrigated land.

The GCC initiative in gum karaya has been a major source of income for almost 12 000 tribal people and an important source of employment for tribal women. Coffee has proven to be popular with tribal communities and has generated greater economic returns. There has been a per-unit yield increase of 84% for Seethampeta, 94% for Paderu and Rampachodavaram and 115% for Parvathipuram. The average yield of paddy throughout the project area has increased nearly threefold (from 1 100 to 3 100 kg/ha), and, in cases of optimum fertilizer application, up to 4 500 kg/ha. Given an irrigated area of 20 000 ha by the end of the project, total food production from the irrigated area can be estimated at 62 000 t, without the application of fertilizer. If only 20% of farmers applied fertilizer, production would be in the region of 80 000 t, an increase of almost 500% over the pre-project era.

Generally, the perception of farmers is that irrigation and plantation have been the main contributions of the project. Extensive podu areas have been converted into orchards, with a positive impact on the environment. This transformation has also been sound from an economic standpoint, since the potential economic returns from orchards are several times greater than the returns from podu farming. However, price fluctuations and continuing exploitation by middlemen have meant that farmers have frequently been compelled to sell their products at uneconomic prices. In several villages, the progressive diminution of podu farming can make the tribals more vulnerable to severe drought conditions, and in some cases they are uneasy at not having their traditional mix of rainfed grain and other low-intensity crops to fall back on. This problem has been exacerbated by extended periods of drought in the past few years. These difficult conditions underline the need to promote agricultural strategies that rely on irrigation and rainfed systems. Further, to ensure the sustainability of plantations, there is a clear need for more training to educate farmers in agricultural know-how concerning new crops and techniques and the ongoing management of their plots.

The sustainability of small-scale irrigation is especially dependent on the effectiveness of community action, and, where the local communities have not been properly consulted in planning, the actual location of the check dam has sometimes been inappropriate. Similarly, the quality of construction has suffered when local people have not been involved in implementing the scheme and the necessary land development has not taken place, with the result that the benefits of irrigation have failed to reach the maximum number of households. Greater attention must still be focused on certain key technical aspects of irrigation development. These include the need to ensure that each scheme is selected on the basis of cost effectiveness, the permanency of water supplies and equitable land distribution and that it is designed and executed on the basis of detailed field investigations and PRA exercises. Greater financial and technical support to farmers is also required for downstream on-farm water projects to ensure the maximization of benefits and equitable distribution of the water through the formation of permanent water users associations. The expansion of irrigation has greatly contributed to reductions in the pressure on podu land and increased the area cropped under rice. However, there are striking disparities with regard to cropping intensity, improved crop husbandry and maintenance.

Environment. The evaluation generally noted an abundance of natural vegetation, and the project areas on the whole are well watered and fertile. Strict forest regulations are in place to prevent deforestation. Attempts to curtail the practice of podu cultivation are important for maintaining and improving environmental conditions, and terracing and bunding have in many cases led to immediate improvements in the orchard-level conservation of soil and water. However, it remains unclear whether thinly planted orchards will suffice to prevent localized erosion, especially where appropriate soil-conservation measures have not been put in place or where other ground cover is scarce. There is also concern about the development of low-lying land for paddy cultivation, with the application of non-organic inputs for high-yield varieties of rice such as fertilizers, pesticides and herbicides. Care must be taken to avoid a build-up of chemicals in these regions, and impact assessment and research into alternative agricultural practices may be necessary. The potential for organic methods of agriculture should be explored, particularly in view of the rapidly expanding market that exists for organic produce.

Beneficiary participation. Participation was promoted through SHGs, VTDAs and users groups. Such institutions have been replicated on a large scale in Andhra Pradesh and cover a wide range of activities. However, one problem has been that there are too few intermediaries between ITDA and the grass roots. Village-level workers and agricultural development consultants have had to cover up to 60 habitations, which often are widely scattered and remote, making it virtually impossible to meet villagers regularly. Thus, often only a small proportion of tribals have been able to grasp the benefits of ITDA inputs, which in some cases has led to strengthening those who are less poor. Despite some good results in the institutionalization of participatory approaches, the concept of participation has been conceived differently by people. Participation has also invariably been linked with an activity, so that social mobilization and community participation were primarily seen as a mechanism for preparing people for the delivery of services, rather than as a process for boosting empowerment. More could have been achieved, but perhaps the time and interaction required for such new thinking to take root in government-led programmes were underestimated. Furthermore, executing agencies require greater incentives to institutionalize structures in support of participatory development processes that can gradually lead to the transfer of authority 'downwards'. The concept of participation promoted through the APTDP has been extremely important, as it is has provided the stepping stone for future development programmes and activities. Participation has contributed positively to changes in social relationships not only within the state and grass-roots institutions, but also among tribal people themselves, as well as between tribal people and other actors in the informal economy, such as moneylenders, traders and other service providers. The APTDP also assisted in initiating a trend to change administrative and bureaucratic approaches, bringing them more in line with the fact that sustainable development is best achieved from the 'bottom'. The PRA signaled that there needs to be a much greater emphasis on training of project staff and enhanced support from ITDAs for village institutions, as well as capacity-building and the streamlining of institutions for the sustainability of the programme.

Credit and marketing. In the past decade, there has been a fluctuating trend in terms of credit delivery and a clear reduction in the amount of loans because of the sharp decline in recovery rates, implying that tribal capacity to repay loans improved only marginally. However, the presence of the GCC credit facility has given a certain confidence to tribal communities, and, through the creation of SHGs, the propensity to save has now become more well established among tribal people. The practice of borrowing in advance and pledging the produce of the following season has not been entirely displaced, despite GCC's expanded programmes. However, wherever there are active social workers or committed officials, the GCC schemes have had a noticeable effect. Greater efforts need to be made to sustain the flow of credit and improve recovery rates and to devise measures so that the landless are able to participate in such schemes.

The marketing of NTFP has been given a major boost. GCC undertook research on the processing and marketing of the forest produce collected by tribal peoples. Its initiative on gum karaya was a good example of the benefits of combining a concern for tribal people with the dissemination of scientific knowledge and professional marketing techniques. Gum karaya is the most important NTFP procured by GCC, accounting for about one-half of total procurement, and is a major source of income for almost 12 000 tribal people. The employment of a pharmaceutical specialist by GCC led to the development of scientific tapping and post-harvest practices, and storage and quality control were modernized. Nearly 80 consultants and 400-500 liaison workers were engaged to train and supervise the collection of the gum. Within two years, the price of Grade-I gum tripled, from INR 30 to 90 per kg, and tribal income rose proportionally. Tree life was also extended as a result of improved techniques of tapping. Similar successes were achieved in the development of bioflocculant from cleaning-nuts, in the identification of a large number of medicinal herbs available in the forest and in establishing markets for them. GCC attempted to institutionalize the involvement of tribal communities in these efforts and also to involve tribal youth through forest-produce gatherers associations. However, the trader-cum-moneylender continues to have a strong presence in the marketing of NTFP, and even though the law has given monopoly rights of procurement of NTFP to GCC, it has been able to procure only about 70% of these products according to its own assessment. The remaining portion, often including the best quality NTFP items, is left to the private trader. The linkage with the moneylender continues because of transactions based on the dynamics of tribal contingent needs. The trader's agent arrives at the villager's doorstep as soon as the produce is available and collects the item against advances provided earlier. Even though the interest on the loan is as high as 10 or 15% per month, the practice continues, and in some places VTDA or the liaison worker has intervened only to redirect this sale to GCC. Further improvements need to be made in ensuring greater communication at the village level as regards the availability of information on NTFP prices fixed by GCC. Because of illiteracy and lack of information, traders are in a position to manipulate prices for the purchase of both agricultural and forest products.

Social impact. The PRA indicates that the project successfully reached the great majority of households within five of the six settlements. In the six villages assessed, three villages identified the construction of the check dam as the most important achievement, and the other three cited the construction of pukka houses. A total of 14 292 houses were constructed, and tribal people felt pride of ownership because of their involvement. The overall standards of health and nutrition are surprisingly high, and the placement of more than 1 000 community health workers to make basic health facilities available in remote villages has been the most important innovation under the health component, especially in pre-natal care and hygiene.

The demand for education now has its own dynamic: a conclusion that may be the single most significant development in terms of long-term changes in perceptions in tribal areas. The introduction of community schools and direct involvement of parents in managing the schools have played an important role in raising enrolment and attendance rates. In all four districts, there have been annual increases in enrolments (27% for Seethampeta, 9% for Parvathipuram, 12% for Paderu and 19% for Rampachodavaram), of which approximately 40% are accounted for by girls. School drop-out rates fell from 71% in 1995 to 53% in 1998 and continue to fall. One crucial educational element that needs to be addressed immediately is the provision of vocational training in tribal areas.

According to the PRA, the status of women is generally higher in tribal areas than in non-tribal societies, and in some instances, the leader of VTDA was a woman. The PRA shows that the perception of gender equity improved noticeably in all six villages under review, mainly because of the existence of SHGs. However, the dynamic connection among women's groups, credit access and income generation has been absent in the APTDP area. Female literacy rates in the project area have risen substantially, although perhaps the main impact of the project activities on women has been the steadily increasing proportion of girls in school. One negative impact on gender equity has been the fact that men now concentrate more on work in the irrigated areas, whereas men and women were previously accustomed to a large degree to working together on podu land. Women play a key role in agriculture, village institutions, education and rural marketing, and it is essential that gender issues be woven into all aspects of IFAD strategy.

Peace-building. The project area suffered from an insurgency campaign mounted by the Naxalite People's War Group (PWG), which, in the early stages at least, appeared to have been motivated by a genuine desire to improve the welfare of tribal people. At that time, the PWG served to highlight the problems of tribal people and the urgent need of measures to protect them, in particular in terms of their alienation from tribal lands and exploitation by unscrupulous moneylenders and middlemen. The APTDP contributed to reducing insurgency activities in the project area, illustrating that the Fund has a facilitating role to play in peace-building, particularly in those areas in which protest movements are supported by people at the grass-roots level that have little choice but to join due to their social and economic vulnerability.

The project demonstrated that IFAD has a development philosophy characterized by participatory and bottom-up approaches, with the objective of transferring decision-making and ownership to tribal people themselves. IFAD also contributed to promoting changes in attitudes among government and related authorities on the one hand and people at the grass-roots level on the other - particularly participants in the insurgency movements. In its earlier days, the PWG gained momentum through the support of disenchanted tribal people, relegated to lives of abject poverty, who looked to the PWG as a possible escape from misery. However, the APTDP brought about changes and development in ITDAs that created an atmosphere of optimism and empowerment, leading tribal people to gradually distance themselves from the PWG, thereby weakening the Naxalite movement in the project area. Finally, the APTDP promoted better functioning of the informal sector in the tribal area, including among moneylenders, contractors and traders.

Insights and recommendations

Participation and empowerment. The APTDP has been a pioneer in stressing community participation and has brought about changes in tribal development during implementation by creating space for a multi-stakeholder approach with a specific focus on tribal people. One achievement has been that programme management has been initiated, executed and monitored by communities and tribal people have been empowered as partners in the improvement of their natural resource base and their means of livelihood. However, despite some good progress towards the institutionalization of participatory approaches, the concept of participation has been differently understood by different people at different times. Consequently, community participation strategies were applied differently in the four ITDAs, and there was inadequate continuity in the approaches and the emphasis on participation. Moreover, because participation was linked with an activity, social mobilization and community participation were primarily seen as a mechanism to prepare people for the delivery of services. It is important to build a culture where people are not accountable for the results achieved, but for their ability to 'manage for results'. Village institutions should be organically linked, nurtured and facilitated. Training and capability development for social mobilization should be an integral and regular aspect of project design and implementation, as should the involvement of competent NGOs. Careful attention should subsequently be given to the institutionalization of the expertise and skills generated so that the capabilities developed are successfully sustained.

Community institutions. The performance of various community institutions varied widely from village to village, depending largely on the level and regularity of project inputs, both economic and motivational. The design of future projects should be more realistic as to the formal extent of community involvement, and competent NGOs should be involved in both the creation and support of nascent institutions. With the knowledge that villages with strong traditional institutions find it easier to internalize new approaches and technologies, the intensity of project delivery needs to be based on the institutional capacity available at the village level. The fact that the project lacked a coherent strategy to ensure the convergence of community institutions with the traditional power structures on the one hand and government institutions on the other may have created difficulties in achieving better bargaining power for the community. In order to avoid the emergence of parallel and potentially divisive structures, the nature of traditional authorities must be investigated, and the necessary links between new and old institutions explicitly recognized. Although the project was successful in bringing about changes in attitudes among development workers at all levels, the building of strong, long-lasting community institutions that allow tribals to become self-reliant and to reduce their dependency on external institutions is a process that requires time, the continuity of efforts and sustainable support mechanisms.

Insurgency movements and development projects. The APTDP offers an unusual opportunity to examine the role IFAD can play in conflict resolution, in particular through assessment of the relationship between an insurgency movement (Naxalism) and the operation of a participatory development project in tribal areas. The mere involvement of IFAD in such a sensitive area was seen as a serious effort by the Government to respond to tribal disaffection and exploitation. IFAD provided a silent bridging leadership, playing the role of a facilitator that could be trusted and was committed to furthering the interests of tribal communities. The Fund's facilitation role significantly contributed to greater awareness within the Government and the civil-service cadres and to a process of change in bureaucratic attitudes that favoured more listening and greater partnership-orientation in the effort to boost the development of the girijans (tribal people). Emphasis was also placed on the need to address the movement through social and economic advancement schemes, as well as through a law-and-order approach. The APTDP also brought about the development of ITDAs, creating an atmosphere of optimism and empowerment, thereby weakening the Naxalite movement in the project area. In order to maintain the physical and social assets created under the APTDP, further investment in infrastructure, training programmes and adaptive research is essential.

Exit strategy and post-project monitoring. Although the development of a post-project completion plan is integral to project design, it is common for projects to be designed and implemented with limited attention to a strategic phase-out. It is necessary to devote serious attention during design to a completion plan that identifies institutional responsibilities and roles and, to the extent required and where possible, recognizes that IFAD's continued assistance in the post-project period is critical. This task would be facilitated by the creation of a post-completion monitoring system to highlight important issues requiring the Fund's follow-up and guidance. There is a need for the immediate assistance of IFAD and the state in consolidating the achievements and impact of the APTDP, particularly with respect to capacity-building in communities and community institutions. This would require the allocation of additional resources specifically allocated to social mobilization and training.

Food security, podu farming and the environment. Total food production in the project area has increased by an estimated 500% during project life. However, this figure masks important problems, including poor maintenance of irrigation systems, a lack of sufficient know-how regarding horticultural techniques and, most significantly, increased vulnerability to drought because of the dependence on irrigated agriculture. The degree of change involved in the replacement of a tried and tested agricultural pattern by an approach requiring different techniques, different seasonal patterns and a different attitude towards natural resources must be recognized. For example, emphasis needs to be placed on preliminary research concerning the choice of horticultural crops in particular areas to determine which crops would be successful in given soil, altitude and climate conditions. Wherever radical agricultural transformation is being attempted, there is a vital need for training, guidance, adaptive research and strengthened extension services in order to support and sustain the transformation. The training of trainers is the first requirement of training programmes, and this must take into account the fact that the line department officials themselves may have little knowledge of either local conditions or tribal societies.

Water users associations. A notable achievement of the project in terms of the increase in irrigated land has been the utilization of indigenous knowledge in the design and construction of cost-effective irrigation systems. However, in various villages the full impact of these improvements has yet to be realized, owing to the underperformance of water users associations. The responsibilities of these associations include the construction and maintenance of irrigation channels, but in many cases no associations existed, or, where they did exist, they were ineffective. In some cases, irrigation channels are already in need of maintenance. The main reason for this appears to be the lack of training provided to water users associations concerning their duties and responsibilities. Irrigation development includes the appropriate construction of field channels, land development, efficient on-farm water-management measures and efficient land-use and crop production, as well as post-harvest
management practices. In each of these areas the training of association members is the crucial element. Despite the repeated recommendations of supervision missions, the necessary training programmes were not put in place. In many schemes, there is now sufficient water for a second crop, but there has been no investigation of the possibility of advancing the second planting, as was done in earlier programmes in the plains of the Godavari districts during the Green Revolution of the 1960s.

Importance of vocational education. The demand for education in tribal areas has grown rapidly over the last decade, even in the absence of the necessary infrastructure and personnel. However, educational gains have not been matched by employment opportunities. The ashram schools, which are geared towards basic academic education, meet the overall mental- and social-development needs of the child but are not directly linked to post-school employment. Although these schools are admirable, tribal communities have an immediate need for craftsmen, irrigation engineers, nurses, teachers and tailors, and, above all, there will be an increasing need for farmers with knowledge of slope-agriculture land technology, pesticides, orchard maintenance and animal husbandry. Hence, it is important that the teaching of these skills be given the central importance it deserves, and this should be achieved within the existing ashram system. This shift in the purpose and philosophy of the ashram schools will require the full support of the Government of Andhra Pradesh, and it must be emphasized that the educational system cannot stand outside the development process as a whole. The Government of Andhra Pradesh should be encouraged to set up a working group within the Department of Education to explore the possibilities for vocational education within these schools. As part of the suggested post-completion process, IFAD should perhaps consider funding comparative studies on vocational education and its applicability to tribal areas.

Health care. Community health workers have been one of the undoubted successes of the project. The basic training package is sound and should be refreshed and extended where possible. In particular, basic midwifery training should be included in all programmes, particularly in view of the resistance to institutional care among tribal women in certain places. The Government prohibition on intramuscular injections by unofficial health staff might perhaps be reviewed. After two or three years of experience and a series of training programmes, such injections ought to be well within the competence of capable community health workers. The relationship between official health staff and the community health workers needs to be one of partnership. In some cases, there is some suspicion on the part of trained officials towards the 'barefoot doctors', and indeed this is to be expected to a certain extent. This could be minimized by ensuring that the health staff immediately superior to the community health workers be personally implicated in their training, as well as in supervision. In this way, they could themselves identify with the progress and achievements of the community health workers. Some workers were given very specific instructions to deny the validity of traditional medicinal practices. In view of the emphasis on the part of the project and ITDAs on respecting, utilizing and building on traditional tribal culture, more attention might be given to an assessment of traditional practices regarding herbal teas, roots and other medicinal plants. Wherever possible, beneficial therapies of this kind should be married with modern therapeutic practice. The dismissal out of hand of local knowledge runs counter both to the spirit of the project and to the increasing attention being accorded to traditional medicine in general.

Land and parity issues. The disparities that exist between remote and less remote villages must to some extent be regarded as inevitable given the difficulties of communication. Where reaching a village demands long journeys on foot and overnight stays, the evaluation noted that the village does not receive the same attention as more accessible habitations. In effect, more remote villages require a greater proportional investment in terms of time, effort and money, and this needs to be reflected in project design. Wherever possible, a housing scheme should include all householders of a hamlet or village, provided they meet the targeting criteria, in order to avoid the emergence of a two-tier social structure with hut-dwellers as a conspicuously less-favoured group. Certain long-standing settlements are located in areas designated as reserve forests. Communities have been allowed to remain where they are, but it is still legally impossible for secure land titles to be issued in such circumstances. This is a matter that requires consideration by the Forestry Department and ITDAs in order to ensure greater integration of all segments of tribal society. Finally, tackling the problem of landlessness has not been a major focus of the APTDP, and the majority of project activities have focused on households with a modicum of land. However, the seriousness of the problems of the poor and of marginal farmers in tribal areas can be gauged from the fact that approximately 20-25% of the households in tribal areas are estimated to be landless.3 Marginal farmers having up to two acres of land constitute 30-35% of tribal households. In other words, almost half of the tribal families possess little or no land and rely either on NTFP, or on work on other people's fields or on podu cultivation.

Maintenance of physical and social assets. Although the project achieved good results in terms of infrastructure-related improvements and attempts at social mobilization, an urgent concern is the extent to which these assets can be maintained in the post-project phase. Continuing efforts are necessary, particularly in terms of training programmes for ITDA staff and beneficiaries alike. This means that levels of investment must be maintained. During the year 2000 and beyond, state financial problems have meant that ITDAs have been starved of needed funds. The effect of the recent shortage of funds on the operation of ITDAs has clearly been profound, and in many cases it was difficult for the evaluation to determine how far this alone was responsible for operational deficiencies. The evaluation witnessed examples of damaged irrigation channels, vehicles out of commission, buildings requiring immediate repairs, cashew farmers struggling due to lack of training and guidance, primary school children without slates, and SHGs without motivation. In Paderu, the closure of the maabadi schools affected nearly 14 000 schoolchildren. All these problems can be addressed through relatively small interventions, but recently ITDAs have not had sufficient resources even for these.

1/ These included the commissioner of the Tribal Welfare Department of the Government of Andhra Pradesh, the development cooperation officer at the Embassy of The Netherlands in New Delhi, Outreach (a non-governmental organization) and IFAD (represented by the Asia and the Pacific Division and the Office of Evaluation and Studies).

2/ Cluster-level associations have been set up in many areas of the Andhra Pradesh Participatory Tribal Development Project with the principal aim of ensuring that communities can take full advantage of all ITDA programmes. The associations consist of ten or more SHGs formed according to the norms of the National Bank for Agriculture and Rural Development. Individual SHGs were considered as too small to be effective borrowing units, and their capital did not go far enough as investments in productive and profitable activities. The formation of cluster-level associations is a significant step in facilitating credit from GCC, ITDA and banks, and, in the Bhadrachalam ITDA, the institution has made remarkable progress. The emphasis was on the formation of women's cluster-level associations that combined several SHGs. The evaluation team observed two cluster-level associations in Bhadrachalam which had been able to purchase tractors and lease land for cultivation. Shops had been set up; the management of girls hostels had been taken over, and other profitable activities had been planned. However, the notion of the cluster-level association is still in its infancy and has not been fitted into the overall scheme of the Girijan Primary Cooperative Marketing Societies, nor coordinated with the activities of VTDAs. Although the Bhadrachalam ITDA adopted a clear policy of promoting women's groups, GCC as a whole does not have a specific gender dimension built into its programmes. In Bhadrachalam, the evaluation team noted a new wave of women's awareness and a determination to generate income independently and take organizational initiatives with the support of state agencies, NGOs and other public bodies. Women's groups campaigned against alcoholism and the trade in liquor and were addressing significant issues, including primary education and health.

3/ In some villages, the definition of landlessness is understood in terms of the lack of ownership of 'wet land', and podu land is not reckoned as cultivable land.