House of Commons - International Development

Select Committee on International Development Third Report

5 DFID's programme in India: the nature of engagement

89. It is right that DFID works through a variety of aid instruments at a range of scales; selecting their instrument on the basis of their particular objectives in each different context. A key issue for our inquiry was whether DFID was achieving the right balance between different aid instruments within their India programme, and whether these instruments were being deployed on the basis of appropriate evidence.

Poverty Reduction Budget Support and aid conditionality

90. In recent years, DFID has moved in the direction of providing more of its aid in the form of Poverty Reduction Budget Support (PRBS). [154] PRBS refers to funds supplied directly to a Government's Budget, to support its efforts to reduce poverty. DFID told us that whereas in Africa, budgetary support is used to address long-term financing gaps in the public sector, in India it has been mainly used at state-level "to assist with the costs of reform and to provide an incentive for state governments to take difficult decisions and tackle harder reforms."[155] PRBS has also been used to allow state governments to increase spending in key development areas and to substitute for higher cost financing. DFID told us that PRBS provided "an opportunity for donors to engage in a broader range of policy dialogue than is normally afforded by projects," although as mentioned in paragraph 27, the influence which DFID is able to have over policy making at any level in India is always limited.

91. DFID has never attempted to use PRBS as an aid instrument at a central level in India, although it provides programme-specific budgetary assistance at central level and, until recently provided PRBS to two of its focus states (Orissa and AP). In 2004 however, the Ministry of Finance told DFID that it should no longer provide bilateral grant assistance as PRBS to individual states. DFID told us that this pronouncement was being questioned by state governments who would prefer to continue receiving PRBS, but that it seems "very likely" the GoI will stick by its decision. Assuming this is the case, DFID plans, wherever appropriate, to redistribute funds provisionally allocated to PRBS to state-level sector budget support in four sectors: health, education, rural and urban poverty reduction.[156] Providing support to sectoral budgets may enable a state to reach the minimum contribution threshold which will allow them to draw down the maximum allocation of funds from the GoI for a particular CSS. Without donor support, the imperative for states of reaching these thresholds, in order to maximise their total available budget, makes it difficult for them to plan their future investments strategically, and can encourage fiscal irresponsibility. We think it is sensible for DFID to reallocate funds previously directed though PRBS to the state-level budgets needed in order to secure the release of centrally sponsored scheme funds.

Programmatic support; Centrally Sponsored Schemes

92. Responding to critiques of narrow, project-based development interventions, international donors have, over the last ten years, begun to shift their support towards broader sector wide approaches (SWAps). In many countries, donor understandings of what constitutes a sector are wider than government understandings. A health SWAp, for example, might include work on sanitation and water purity, whereas a government's health ministry might purely address medical service provision. In India, the size and strength of the GoI is such that, in practice, donors' sectoral approaches tend to conform with government-defined sectors, as defined by the GoI's Centrally Sponsored Schemes (CSSs). Employed since India's first five year plan (1951-1956), these are national programmes which are incorporated into state-level policies.[157] The GoI currently allocates RS 250bn. (£3.5bn.) annually to its CSSs. Some CSSs are wholly funded by the national government, but most involve state counter-funding.

93. DFID has sought to maximise its influence on the delivery and monitoring mechanisms of CSSs by strategically focusing on schemes within the sectors of health and education. DFID told us that GoI officials value the Department's involvement in CSSs because of the monitoring conditions which DFID imposes. Without DFID establishing these conditions as requirements of their assistance it would be politically unacceptable for them to be introduced. Once they have been included, they establish a valuable precedent for the design of future schemes. DFID's involvement in CSSs enables them to have an impact across all states, rather than just in their focus states, multiplying the effects of their involvement, as Dr Seymour-Smith told us "a small policy change leverages quite a lot of increased effectiveness."[158]

94. We heard criticism however, that because CSSs are 'top-down' programmes, conceived and designed by central government, they often fail to take account of the specific circumstances of individual states. In relation to DFID, we heard that the Department's growing focus on CSSs has led them to engage more and more with central government and to take a 'hands off' approach to working in non-focus states.[159] Although we saw little evidence of the impacts of this shift, we note that DFID's growing support for CSSs would be problematic if it led the Department to neglect more 'hands on' work with non-focus states. The Secretary of State told us however, that CSSs are only one of the four ways in which DFID is considering increasing its engagement in UP and Bihar:

160

95. Perhaps the clearest example of the positive impact which international donors have been able to have on India's CSSs has been in relation to HIV/AIDS. International donors and organisations played a key role in motivating the GoI to initiate its first responses to HIV/AIDS during the early 1990s. The influence of UNAIDS, the UN body responsible for coordinating the responses of all UN agencies to the HIV/AIDS pandemic, was essential in prompting the GoI to acknowledge that HIV/AIDS was an issue affecting the country. In 1992, the World Bank offered India a soft loan of $87m., until that point its largest HIV-related loan to any single country, to establish a National AIDS Control Organisation (NACO) and to begin the first stage of a CSS addressing HIV/AIDS; the National AIDS Control Programme (NACP). In 1997 the World Bank was able to use the leverage of the renewal of this loan to convince India to withdraw a draconian AIDS Bill which had been introduced into the Lok Sabha.

96. Since 1999, DFID has been involved in the GoI's NACP, working at national level, through NACO, as well as at state level, through State AIDS Control Societies (SACS) in six states; Kerala, AP, WB, Gujarat, Orissa and MP. DFID's financial support to the NACP, which by 2007 will total £123m., has probably added weight to the Department's efforts to shape India's agenda on HIV/AIDS. Dr. Joanna Reid, a Senior Health Adviser with DFID in India, told us that DFID is also attempting to 'mainstream' the issue of HIV/AIDS throughout all its programmes in India. Historically, the GoI's reluctance to acknowledge the significance of HIV/AIDS in India made it important for DFID to become involved in the NACP. Early indications are that the new UPA government is taking the challenge seriously, and a recent change of leadership at NACO lends further grounds for optimism. Now that the GoI has shown the political will to engage with the issue of HIV/AIDS, we think that it is vital that DFID continues to support the programme. We were therefore delighted to hear that DFID has recently accepted an invitation from the DEA to extend financial support to the SACS in Bihar and UP.[161] With India likely to be acknowledged as the country with the largest number of HIV positive citizens in the world during 2005, we are pleased that DFID recognises the country's window of opportunity to act decisively to tackle the epidemic. We strongly encourage DFID to do all it can to support the GoI in taking such decisive action.

97. Although we approve of DFID's signalled intention to stop attaching policy conditions to its development assistance, we are convinced of the need for the Department to continue to attach process conditions, which aim to improve the quality and effectiveness of aid. In the course of our inquiry we were concerned to establish to what extent DFID had been able to negotiate process conditions in India, given the limitations of DFID's influence on GoI policy making. DFID told us that the UK's financial assistance to India had enabled DFID to engage the GoI in a "genuine and fruitful" dialogue.[162] We wished to establish what evidence there was of the impact of this dialogue.

98. One example which DFID gave us of their impact on a CSS related to the GoI's universal elementary education scheme, the Sarva Shiksha Abhiyan (SSA), to which DFID will contribute a total of £210m. over the next four years.[163] Education is acknowledged, by the GoI as well as other actors, to be the sector in which donors have had the most positive influence on GoI policy.[164] Roger Cunningham, DFID's Senior Education Adviser in India, told us that although the SSA was essentially 'home-grown' — the GoI was responsible for its policy design — the Indian government had invited DFID, together with the World Bank and the European Union, to become involved in appraising and monitoring the scheme. DFID was invited to undertake an appraisal of SSA, addressing issues of process, including financial management, procurement, fiduciary risk, targeting, monitoring and evaluation. Now that the appraisal has been completed DFID will be involved in twice yearly evaluations of the scheme's implementation. Although we were encouraged to hear that DFID has been involved in the appraisal of the SSA, we remain concerned that we have not seen any evidence of the impacts which DFID was able to have as a result of this involvement.

99. The problem of measuring the value which DFID adds to CSSs relates to the more general issue of how effectively CSSs are monitored. A problem inherent in the design of CSSs is that the GoI cannot control the implementation of the schemes at state or district level. In addition to the problems of leakage and the corrupt diversion of funds (discussed in Chapter 3), the funds allocated to CSSs are also highly fungible. This is a significant issue for donors who want to ensure that their money makes an additional contribution to the sectors which they have chosen to target. Nayak et al. have found that:

165

In light of the fungibility of assistance allocated to CSSs, we were particularly interested to hear from DFID staff about the use of incentive-based financing models to encourage central and state governments to contribute their share of funds for CSSs. Under such models, donor funds are released in response to government contributions reaching certain thresholds, established on an annually rising baseline. Incentive-based financing is a sensible strategy, which discourages the movement of government funds away from those sectors where DFID has become engaged, as well as enabling DFID to maximise their leverage over the financial management and auditing of CSSs. We encourage DFID to continue to develop and deploy such mechanisms. In addition to the general issue of monitoring, there is the more specific question of how adequately the pro-poor impacts of CSSs are being assessed. We have not seen evidence which convinces us that the GoI or DFID have developed mechanisms which can adequately assess the extent to which CSSs are impacting on India's poorest people.

100. In 2003/4, DFID spent £79m. through its National Programme in India.[166] DFID told us that its current spending forecast for the National Programme in 2004/5 is £147m.[167] Given that around three-quarters of National Programme spending is through CSSs,[168] this shift represents a significant increase in investment in CSSs. As we have already commented (paragraph 70), we do not think DFID has adequately justified this substantial budgetary shift in the public domain. We are concerned that DFID has decided to substantially increase its investment in CSSs while convincing evidence of the value which DFID has been able to add CSSs to date remains scant. Furthermore, DFID has not made clear how their growing focus on CSSs at central level fits into their overall strategy for working with Government in India, nor how the Department plans to create synergies between its central and state-level work.

Sector support and project support

101. Outside of its four focus states, DFID engages in project work. This is funded through its National Programme, and is usually undertaken by partner organisations who have established some sort of comparative advantage by working in particular states. During our visit, a number of witnesses told us that DFID's project work in India was particularly valued for its innovation and the demonstration effect it could achieve.[169] We were very impressed by the individual examples of DFID's project work which we saw during our visit (some of which are discussed below).

EDUCATION

102. As well as engaging with education-focused CSSs, such as the SSA (discussed in paragraph 98), DFID has become involved in state-level projects in the education sector. During our visit we saw two examples of such projects: in Orissa a post-cyclone school reconstruction programme; and, in AP, an ILO programme addressing the issue of child labour. We also heard about Kerala's experiences in the education sector.

103. Orissa is a natural disaster-prone state, affected by cyclones, flooding, drought and earthquake over the past decade. In October 1999, the eastern, coastal region of Orissa, was hit by two severe cyclones within a span of just 11 days. The second, a rare 'super cyclone' with wind speeds of 300km per hour, killed more than 10,000 people and 406,000 livestock. DFID provided emergency assistance in the immediate aftermath of the cyclones and the flooding which followed. At the time there were complaints in the press and from NGOs of the slowness with which donors reacted to the disasters.[170] Subsequently there was criticism of donors for failing to fulfil their pledges of assistance for rehabilitation.[171] We acknowledge that financial and bureaucratic factors may restrict the speed with which DFID can react to natural disasters in India, but think that improved coordination between DFID, other donors and UN agencies in the interim periods between disasters, would help to facilitate a more rapid and coordinated response in the event of future disasters. Fortunately Orissa was not badly affected by the recent tsunami earthquake.

104. Working in the longer term following the super-cyclone, DFID has contributed a grant of £33.2m. and technical expertise to the Post-Cyclone Reconstruction of Primary Schools project (PCRPS) in Orissa. The PCRPS is a project led by the State government's Primary Education Programme Authority, to reconstruct primary schools which were damaged by the 1999 super-cyclone. DFID's funding is to support the reconstruction of 3500 primary schools in 14 districts. Intended to double as community shelters in the event of future disasters, the schools reconstructed by the PCRPS project are designed and built to a high specification in order to withstand extreme natural events. Often the only concrete building within a village, these schools also provide an important community resource during interim periods between disasters. We heard that in villages where the school reconstruction scheme had been implemented, local people were beginning to attach a higher priority to educating their children. Raised daily attendance at school, particularly by young children, has been aided by the implementation of the GoI's midday meals programme in Orissa. DFID told us however, that by October 2004, only 670 of the 1,334 schools scheduled to be constructed in 2003/4 had been completed, and that although the remainder may be finished by the end of 2004, it will be a challenge to complete the additional 1,800 schools scheduled for construction by June 2005. It is unfortunate that five years after the super-cyclone, only 670 of the 7000 primary schools damaged by the disaster have been reconstructed.[172] We believe DFID should encourage the GoO to accelerate the PCRPS project, as well as extending it more widely across the State, perhaps seeking to involve the voluntary sector and private firms in the financing of new school buildings.

105. An innovative aspect of the PCRPS project has been the involvement of village education committees (VECs), composed of local people, parents and teachers, in the planning and management of the construction of the school buildings. The strategy of involving the VECs in the reconstruction process was intended to increase communities' feelings of ownership and involvement. We think that the involvement of grassroots institutions in schemes such as the PCRPS project is important, especially if the VECs can continue to be involved in the day-to-day running of the schools once they re-open. In particular, VECs could play an important role in combating the pervasive problem of teacher absenteeism in remote rural areas; if teacher salaries were administered through VECs, these could be withheld if a teacher failed to turn up for lessons. It is important however, to recognise that although the involvement of grassroots institutions, such as VECs, is designed to amplify the 'voice' of the poor in the formation and implementation of development policy, it may in reality serve mainly to augment the influence of relatively powerful individuals within a community. There is also a question about the sustainability of societal institutions established in the context of development interventions, and whether working through existing institutions would be a more sustainable approach.

106. In AP, DFID contributes to the International Labour Organisation's International Programme for the Elimination of Child labour (ILO IPEC). This programme mobilises a wide range of stakeholders, including parents, employers, trade unions, civil society and government officials, to develop a 'state-based approach' to eliminating child labour. Although the ILO IPEC is primarily educational, it also involves a significant degree of cross-departmental working, with the innovative aim of integrating prevention and education work. We saw one key feature of the programme, a 'bridge school' which aims to enable former child labourers to make the transition to mainstream schooling. The school was an example of cooperative working between various partners: the building was provided by employers' organisations; the staffing was provided by NGOs; and, the funds came from GoAP. The ILO IPEC is also looking at community monitoring of the incidence of child labour and at ways of providing mothers with additional income in order to allow their children to stop working. As with the other programmes we saw in AP, we were told that DFID's financial contribution had been vital in enabling the programme to produce a demonstration model which would encourage the government to develop their work in a new area.

107. The bridge school approach taken by the ILO IPEC was pioneered by Indian NGOs including the MV Foundation in Hyderabad and CINI-Asha in Calcutta. It is now being taken up by the GoAP, and more widely across India, to provide temporary education centres for children who enter school late. The aim is to bring these children, often former child labourers, up to the right grade for their age as quickly as possible. The approach is also used to target potential child labourers, particularly those around the vulnerable age of ten, as a way of ensuring that they become involved in education rather than work. We heard however, of concern at the high drop-out rate when children who had attended a bridge school re-entered mainstream schooling. This problem could be tackled by the development of more vocationally-oriented post-basic education. This might encourage parents to believe that it was worth their children's while to continue their education after finishing at a bridge school.

108. DFID's money has helped leverage GoAP funding for phase two of the child labour project, which Hilary Benn launched during his visit to India in November 2004. This second phase will involve tracking the children to ensure they continue to attend education as well as addressing the complex problems of urban child labour. The other major area on which progress still had to be made was the economic rehabilitation of parents whose children work; and the role of employers, particularly private and informal sector operators, in helping provide social security for parents.

109. Since Kerala has met its non-poverty MDGs, the State's concern has shifted from issues of quantity to those of quality. During our visit to the State, the main criticism we heard was of the GoK's inability to build on its impressive achievements in primary level education at other levels. Tertiary education was underprovided and the GoK had failed to compensate for this by creating a climate for private investment in higher education; Bihar, it seems, has more elite schools per capita than Kerala. Another problem is that educational achievement and universal literacy have raised people's employment expectations but the GoK has failed to generate the jobs to meet these expectations. Some of those we met attributed Kerala's low overall productivity to a culture of activism (encouraged by dynamic trade unions) and a lack of professionalism in the Keralan approach to work. This explanation is contradicted by the high productivity of Keralites working abroad, however.

HEALTH

110. In UP we saw an example of a health-related project which is being partially funded by DFID and implemented by UNICEF. Through its Polio Eradication Initiative (PEI), UNICEF aims to eliminate polio from UP, one of its last footholds in the world. Mr Ray Torres, UNICEF's State Representative in UP, and his team, told us how the PEI has sought to achieve sustained behaviour change on immunisation and to improve vaccine acceptance. It has pursued these objectives through: an Information, Education and Communication (IEC) campaign; the development of partnerships with over 230 Muslim institutions at block, district and state-level; and, capacity building with partner organisations. The project works particularly with Muslims, a particularly marginalised community, which was previously underserved by immunisation programmes. One reason for the lower rates of immunisation among Muslims was the prevalence of myths claiming that polio immunisation was in fact a covert means of sterilising the community. The PEI has worked to dispel such myths. We were impressed by what we saw of the initiative and particularly by the committed and enthusiastic representatives of its cadre of community motivators, who we met in Sitapur District. The PEI has the advantage of working through existing state and local institutions, reinforcing the government's existing immunisation programmes, which has the potential to render its impacts more sustainable.

111. DFID's involvement in the UNICEF PEI in UP is a good example of the way in which the Department can work in states where it has no focus state programme. By working through UNICEF, DFID is able to benefit from the agency's comparative advantages in UP: an established relationship of trust with the GoUP; a large pool of committed sector specialists; and, an effective field presence in the State. Although working through partner organisations introduces new elements of risk, which DFID must manage, the UNICEF PEI demonstrates the potential of partner organisations to deliver systematic, well managed and effectively monitored programmes on DFID's behalf.

LIVELIHOODS

112. DFID has established a good track record and reputation for its livelihoods and social forestry projects, which have been in place in many parts of India. During our visit we had the opportunity to see several of the livelihoods projects with which DFID is involved. In Orissa we visited a watershed project in the district of Bolangir. In AP we visited a rural livelihoods project and a programme to provide services for the urban poor, both of which are described in some detail below.

113. Research on rural poverty and livelihood diversification in AP shows that although poor people are already seeking to move out of farming, they frequently move into other poorly paid activities and are consequently unable to exit from poverty.[173] Assisting farmers in dryland areas is a significant priority for the GoAP. Rain-fed agriculture makes up the largest sector of AP's economy, but is threatened by indiscriminate exploitation of ground water, droughts and a strong preference among farmers for growing rice, which have together contributed to a rapid drop in the water table in parts of the State. In one village that we visited we saw an almost total absence of vegetative cover and were told that this had reduced significantly over the past few years. We talked to villagers who were desperate to see their lands brought back to life, their priority being the restoration of defunct water tanks, which raise the water table and increase the availability of drinking water. The availability of more water would allow farmers to grow a greater diversity of crops and vegetables, which could be sold in towns.

114. The AP Rural Livelihoods programme (APRLP), to which DFID has committed £46m over 7 years, operates in 5 of AP's poorest districts which depend mainly on rainfed agriculture. The aim of the APRLP is to strengthen the GoAP's watershed development programme by adopting a participatory approach to the development of rural livelihoods, in order to improve agriculture and create other income earning opportunities. In key respects the APRLP adopts the approach to livelihoods and participatory development which was successfully pioneered by DFID in the Western and Eastern India Rainfed Farming Projects. The APRLP's approach is to develop locally appropriate and sustainable agricultural practices to tackle poverty and food security by building on local knowledge and traditional crops, as well as by bringing in experience from elsewhere. The project expects to impact directly on 300,000 poor and vulnerable people and indirectly on another 1.2m.

115. Through DFID's involvement, a high level of participation and a focus on the poor have been introduced into the government's original, supply-driven watershed programme. The landless poor tend to be the very poorest of the poor, and previously the livelihoods programme had paid little attention to bringing them into the programme. The civil servants administering the APRLP are now required to consider whether their policies involve expanding the livelihoods of the rural landless poor, particularly women. The APRLP aims to address the needs of the poorest people, without alienating the community as a whole. Its efforts are focused on strengthening existing Self Help Groups (SHGs)[174] and networking them within villages to form Village Organisations (VOs). Each VO manages a Livelihood Fund (comprised of the group's pooled resources with a matching grant from the APRLP). The APRLP only releases its money if, by the end of the first year, 20 of the poorest households in the village have been included in the SHGs, and if, by the end of the second year, all of the poorest households have been included. Additionally, at least 30% of the fund must be spent on addressing the livelihood needs of the poorest. The remaining 70% is for the wider community, which includes the poorest.

116. The APRLP draws attention to the importance of processes in the delivery of development interventions, and demonstrates the importance of these being internalised and sustained by the community. The type of participation encouraged under projects such as the APRLP, reflect a common dilemma in the design of development interventions; whether to set up project-specific institutions (like SHGs) or to rely on existing organisations (like local government bodies known as panchayats). We were told that SHGs are less prone to being captured by elites than panchayats. While project-specific institutions may be more pro-poor than the panchayats, they may be less sustainable and inadequately integrated with existing institutional structures.

117. Gender divisions increasingly correlate with livelihood divisions among poor landless households; poorly paid agricultural labour is becoming feminised. At the same time male members of poor households are increasingly taking up migratory work, creating new gender-specific livelihood insecurities in agricultural slack seasons. The APRLP needs to demonstrate that it can deal with gender specific aspects of poverty. Despite this concern, we believe that that the programme has developed a sustainable and integrated approach to livelihoods, as well as incorporating previously excluded factors, such as migration. The District Collector in Mahabubnagar told us that DFID's technical assistance and new ideas were as important to the project as the money they provided, although DFID's investment had also been crucial to allowing them to prove that their new ideas worked. We were told that the GoAP was now rolling out similar initiatives across AP at its own cost.

118. The Andhra Pradesh Urban Services for the Poor programme (APUSP) is a partnership between the GoAP and DFID aimed at achieving a sustained reduction in the vulnerability and poverty of the urban poor. The APUSP programme aims to improve public services in slum areas of 32 towns and is expected to benefit 2.2m. slum dwellers by 2006. The programme supports government reforms in the municipalities, funds improvements in infrastructure in slum areas (including water supply, sanitation and solid waste management) and focuses heavily on engaging poor communities in decision-making. In each town, a participatory poverty assessment is undertaken in consultation with local politicians, officials in the municipal government, civil society and the public, in order to formulate a Comprehensive Municipal Action Plan for Poverty Reduction (CMAPP). The CMAPP identifies the needs of poor communities, and prioritises and targets infrastructural improvements in the town. We were told that ward councillors had initially pressed for programme funds to be shared equally between districts, but that working with a poverty focus, DFID had ensured that a matrix was developed to ensure that the poorest areas were identified and targeted. From this point onwards the project had been demand driven, with infrastructure improvement priorities identified by the poor through neighbourhood street groups. These groups worked with contractors to design and implement sanitation improvements with technical assistance from the APUSP programme.

119. An innovative feature of the APUSP programme is that municipal infrastructure is provided on the basis of performance improvement. There is no automatic entitlement to improvements unless municipalities can demonstrate a commitment to change, dedication to addressing the needs of the poor, and the ability to maintain the infrastructure provided. We were persuaded of the robustness of the project when the Secretary of the GoAP complained to us about the slowness with which DFID's funds were being disbursed. The release of funds had been conditional upon the achievement of certain performance targets, which had been met more slowly than expected. The Secretary therefore wanted the targets to be relaxed. DFID accepted that funds had been flowing more slowly than anticipated, but explained that this was because the first two years of the programme had been spent on capacity building. The process of empowering communities was part of the project and this meant delay as money was not simply handed over to the contractors.

120. As well as improving the appearance of the slums, the APUSP programme has had health and educational benefits. Local people told us that child health had improved, enabling increased school attendance and savings on medical bills. By allowing people to see that there was funding available to back their ideas, the programme had promoted genuine community participation in decision-making. Revenue from property tax had increased by 108%, largely because of an improved process for identifying properties liable for local tax. This process had developed as the result of certain 'conditionalities' imposed by DFID. We were told that the GoAP and local municipalities have found it politically expedient to be able to 'blame' DFID for the introduction of such bureaucratic reforms. The authorities have been able to ensure the implementation of otherwise unpalatable reforming policies on the basis that they are donor requirements. We were also told that the APUSP had been an important demonstration project. The GoAP would not have been convinced by the project idea alone, but once the effects of the pilot scheme had been demonstrated, the CMAPP had been taken up as GoI best practice and also adopted by the World Bank.

121. The livelihoods projects we saw seemed very successful. While visiting DFID's individual state programmes however, we heard of concern from among DFID staff that the rigidities of projects undertaken within certain sectors were limiting the potential benefits of a more holistic approach to development work. DFID staff in Orissa told us that DFID's legitimacy in the State was derived from the Department's breadth of understanding of Orissa's development situation. They feared this understanding could be lost if DFID focused too narrowly on certain sectors. The recent changes in the organisation of DFID's advisory groups may have begun to overcome this problem, but DFID's continuing focus on achieving the MDGs seems likely to reinforce the narrow foci of sectoral approaches. We hope that DFID will maintain its awareness of this problem, and attempt to resolve it.

122. The development of effective replication mechanisms is a cost effective way of multiplying the impacts of relatively cost-intensive projects. A key issue for DFID is how to disseminate the lessons which it learns from its project work. It is essential for DFID to ensure that the high quality, individual projects which it develops are replicated in other places, and do not remain isolated as 'islands of excellence'. DFID staff in India acknowledged this: in relation to a livelihoods project in Orissa, for example, we were told that it was crucial "that the lessons learned from [the project] transcend localised impacts and input effectively into the state-level policy-making process."[175] DFID's strategy of undertaking projects through existing government programmes seems likely to promote the chances of replication and lesson learning taking place. Hilary Benn believed that DFID had had some success in the replication of projects within states but admitted that lesson learning between states was harder to achieve.[176] Charlotte Seymour Smith told us that DFID was trying to build its relationship with the GoI's Planning Commission,[177] which is currently undertaking work on lesson learning between states. We see the effective replication of DFID's project work as a key factor determining the Department's impact in India and encourage DFID to place a substantial focus on the issue.

Working through civil society

123. DFID works through civil society in India through two formal mechanisms: the Poorest Areas Civil Society programme (PACS) and the Orissa Civil Society Poverty programme (OCSP).[178] The £27 m. PACS programme, which began in 2001, funds NGOs working in states other than DFID's focus states, to help poor people to obtain access to government services. The £1.5 m. OCSP was initiated in March 2004, to experiment with the possibility of a state-specific funding stream, to increase the capacity of civil society organisations in Orissa. In addition to these formal mechanisms for delivering funds to civil society organisations, DFID staff maintain informal links with individuals and organisations in Indian civil society, who provide an important source of information and feedback on DFID's programmes and policies.

124. Historically the GoI has been uncomfortable about foreign agencies funding and engaging with individual civil society organisations. Its concerns have been linked to concerns about Pakistani/foreign influences within India. We heard from DFID staff that the Department for Economic Affairs' process for approving DFID funding of civil society organisations is slow and bureaucratic. The DEA has approved only five civil society organisations' proposals for funding to date.[179] The head of DFID's state programme in Orissa told us that the GoI had been more nervous about the OCSP mechanism than DFID had anticipated; the central government had demanded detailed information on all the organisations which DFID proposed to fund, before granting approval. This has significantly raised DFID's transactions costs. A group of NGO representatives told us that the process of approval is also extremely time and resource intensive for civil society organisations; numerous police checks and visits by officials are required before the government can be satisfied that DFID can fund the organisation. This was a significant disincentive to their seeking funds from DFID programmes. We were pleased to hear however, that the new UPA government is currently reviewing the procedures for donor funding of civil society organisations in India, with a view to simplifying these.[180] Despite the difficulties which have been encountered, DFID should continue to persevere with funding civil society organisations through the PACS and OCSP programmes. Indeed, we encourage DFID to explore the possibility of replicating its OCSP programme in other states. As well as funding individual organisations, DFID should develop mechanisms to encourage civil society networking within India, in order to promote lesson learning and the transferral of best practice between organisations.

Technical assistance

125. DFID delivers technical assistance (TA) to governmental and non-governmental partner organisations in India as part of its National Programme. Most projects undertaken under the National Programme entail an element of TA, while some are composed entirely of TA. Some examples include: the TA which DFID has provided directly to SACS in six Indian states as part of the Department's engagement with India's NACP; the £20 m. Technical Assistance Cooperation Fund for India, established with the ADB in June 2001; the Sexual Health Resource Centre established in New Delhi in 2003; and, the £5.9 m. programme to establish a Centre for Good Governance in AP from September 2001.

126. In December 2004, the NGO 'War on Want' criticised DFID for contracting Western consultancy firms to provide TA in India, thereby creating "aid-funded business".[181] We would be concerned if the provision of these contracts affected India's ability to gather its own evidence and to make its own analysis of the likely costs and benefits of policies such as privatisation. We were reassured to some extent by Hilary Benn's recognition that "Technical assistance only really works if the partner wants to receive it, if together an issue has been identified where it is felt that technical assistance would help, and that technical assistance will enable the Government or the state in the end to be able to do it for itself."[182] Wherever possible, DFID should make use of the considerable expertise of Indian consultants. DFID needs to ensure that all its technical assistance is provided in such a way as to enable recipients to come to their own conclusions about the value of the policies advocated.

127. During the course of our inquiry, the quality of DFID's TA was widely praised. We were repeatedly told that the value of DFID's contribution to India lay in its innovative ideas and practices rather than its money, a view which the Secretary of State acknowledged in his evidence to us.[183] DFID should build on the existing strengths of its India programme by devoting a greater proportion of its resources towards technical assistance, research and the development and dissemination of good practice.

128. If TA is the most important aspect of DFID's programme in India, then perhaps the Department should begin to consider how it might continue providing this to India, if (or more likely when) its aid programme to India is scaled back. One option might be to move to a consultancy-based framework, although we were told that Indian national pride would be likely to preclude the purchase of development advice from a foreign agency. Another option might be to develop the links between UK institutions (such as the ODI and IDS), and Indian development institutions. DFID has already been active in this area; during the 1990s for example, it funded the IDS to train the main Indian civil service training institute (the Lal Shastri National Academy of Administration, in Mussoorie) on Gender Planning. We encourage DFID to continue developing links with Indian development institutions, through international secondments, collaborative programmes and joint research initiatives. In many areas India is now at the cutting edge of international development policy and practice,[184] and so the promotion of such links would be mutually beneficial for development practitioners in the UK and India. The encouragement of such global networking could be seen as DFID's legacy in the sub-continent.

154 IDC, Eighth Report of Session 2003-04, DFID: Departmental Report 2004, HC 749, paragraph 53. Back

155 Ev 64 [DFID supplementary memo]. Back

156 Email communication from DFID India to the Clerk of the Committee, not printed, placed in the Library. Back

157 CSSs constitute one of the three main ways in which funds from central government are transferred to the states, the others being transfers under the provision of the Finance Commission and direct support to States' five year Plans. Back

158 Q 159 [Dr Charlotte Seymour-Smith, DFID India] Back

159 Ev 94 [ActionAid India memo]. Back

160 Q 156 [Rt Hon. Hilary Benn MP, Secretary of State for International Development] Back

161 S. Jha "DFID to help Bihar, UP combat AIDS", Times News Network (9 January 2004). Back

162 Ev 65 [DFID supplementary memo]. Back

163 See http://www.dfid.gov.uk/news/files/pressreleases/pr-indiaeducationfunding12oct04.asp. Back

164 This may possibly be because of the synergies between the advocacy of civil society organisations and donors working in the sector. Back

165 Radhika Nayak, N.C. Saxena and John Farrington, Reaching the Poor: The Influence of Policy and Administrative Processes on the Implementation of Government Poverty Schemes in India, ODI Working Paper, No. 175 (2002) p.12. Available online at http://www.odi.org.uk/publications/working_papers/wp175.pdf. Back