Speakers:
John Farrington - ODI
Craig Johnson - ODI
Rob Jenkins
Birbeck College

Chair:
John Harriss, LSE

1. The study selects a sample from four types of poverty reduction schemes: from those involving direct income transfer (The National Old Age Pension Scheme and the Employment Assurance Scheme); from those concerned with promoting self-employment opportunities (the Integrated Rural Development programme and the National Credit Fund for Women); from those concentrating on employment creation (the Complete Village Employment Scheme and the Employment Assurance Scheme); and from those aiming to improve the quality of infrastructure (the Accelerated Rural Water Supply Programme and the Drought Prone Areas Programme). These schemes were then used to demonstrate how differences in design and implementation methods influence performance.
   
2. Farringtons presentation compared two of these programmes: the National Housing Scheme and the National Old Age Pension Scheme. The National Housing Scheme involves substantial lump sum payments and is subject to corruption. There is evidence that benefits have been allocated to party supporters. The National Old Age Pension Scheme, by contrast, is a much better performer. This scheme involves repeated delivery of the very modest sum of about 75 rupees per month. Payments are made via the post office and the process is straightforward.
3. The policy conundrum, Farrington notes, is that there are no policy champions for these modest well-performing schemes; therefore, many are not expanded. Champions prefer to focus on the schemes that allow them greater visibility and votes; these tend to be the ones that are most subject to corruption.

1. The state has substantial powers over land, tenancy and other agrarian institutions; over agricultural policy, especially related to inputs such as pesticides; and over the organisation of public administration. The most recent reforms (decentralisation) were formalised in 1993, when the Federal government amended the constitution and granted constitutional status to the district and village administrations. At the village level, the panchayat is elected to 5-year terms; the gram panchayat and gram sabha are granted powers defined by the state; positions are also reserved for scheduled castes, tribes and women.
   
2. MP considers itself as a state that has effectively devolved power to the people under its "Gram Swaraj" of 1993 and 2001. AP, by contrast, has a more top-down approach. The SMART government system bypasses elected panchayats and uses parallel groups. Both administrative systems are the result of party politics and political accommodation. However, neither is particularly pro-poor, despite the rhetoric.
   
3. There is less caste mobility in MP than in AP and incomplete land reform in both, though possibly more so in AP. There is enormous dependence on local elites and on the administrative system for livelihoods. This, combined with electoral politics, creates strong incentives to control the development process.
   
4. In both AP and MP, (i.) there is little substantive decentralisation; (ii) the poorer groups are particularly dependent on government schemes; (iii) the reservations and public resources are often captured by local elites; (iv) there is great confusion arising from the reforms; (v) red tape is providing huge opportunities to extract bribes for all matters of documentation.
   
5. Preliminary findings indicate that (i) scheduled castes are challenging local elites "outside" the system, though at a price; (ii) previously marginalized groups are infiltrating PRIs (Panchayat Raj institutions), either through reservations or as a result of economic mobility; (iii) there are few external allies; and (iv) democratic reforms are having little impact on existing power relations.

Rob Jenkins focused on two questions: (a) How politics affects the design/reform of political institutions and (b) How the institutional differences between states in India affect whether or not anti-poverty programmes turn out to be pro-poor.

1. MPs direct democracy form of government does not necessarily produce better outcomes. This is not only because of the existence of informal forces of domination, such as caste or community. Preliminary findings indicate that this apparent decentralisation at the local level actually masks a process of recentralisation. The district governments are more under the control of state government than the formal scheme would leave us to believe. All key decisions on implementation management are taken at the district level before being passed on to the villages. The district collector, who is appointed by the Chief Minister, has a central role. In this way, the Chief Minister retains control over what the districts do. The district officials role in determining the position of the watershed schemes creates a stake for project implementers. The implementation agency must pay if it wishes to run the scheme and will, therefore, try to recoup its costs. There is a lack of accountability since the people who illegally distribute these benefits also sit on the monitoring boards.
   
2. In AP, where the Chief Minister implements schemes by by-passing the local institutions, the process has, inadvertently, created competition and marginalized groups have found leverage. This is especially the case in certain villages where there is open intra-elite conflict between the panchayat and self-help groups.