The design of recent and current economic policies in developing countries increasingly reflects the recognition that local communities are capable of addressing their specific economic needs in highly decentralized ways. Thus governments are simultaneously implementing a mixture of community based economic strategies and more formal local government structures. For example, the 2003 poverty reduction strategies suggested by the South Asian Association for Regional Cooperation (SAARC) include an emphasis on expanding local government structures, ensuring free and fair local elections, and ensuring political representation of socially marginalized groups in these local institutions. Additionally there has also been a shift in focus from lowering barriers to credit access for individuals, to encouraging group-based microcredit schemes and group-based economic cooperatives (Sobhan 2009).
A good representative of the types of poverty alleviation policies currently being formulated in India is the National Urban Livelihoods Mission (NULM), an urban poverty specific policy drafted by the Government of India just this year . It states one of its goals as providing support “from the national level to the community level” in order to “induce social mobilisation, institution building and livelihoods promotion”. Lest this sound too top-down an approach, the draft states in the next sentence, “strong institutional platforms of the poor support them to build up their own human, social, financial, and other resources” and finally, that the entire poverty alleviation strategy can succeed “only if it is driven by the poor and their institutions”. In order to help the poor with their ‘institution building’, the policy proposes a number of strategies, including the mobilisation of “neighbourhood womens groups” (where neighbourhood is defined as a spatially contiguous area targeted by the policy, such as an urban slum), the creation of “self help groups” which are registered group entrepreneurships that are eligible for specific types of bank loans, and the building of “community institutions” which are partnerships between these communal groups, local government bodies, and private financial institutions.
While the vague goals and the excessive emphasis on “institutions of the poor” can lead to endless criticism of a policy such as the NULM, there are at least two points of critique and caution specifically related to inequality. First, the policy is extremely unclear about who exactly is supposed to undertake the responsibility of initiating these different group enterprises. It removes the onus of responsibility from the government since this is supposed to be a “grassroots mobilisation” approach but does not lay the responsibility anywhere else. While the ability of communities to self-organise for different purposes is well established, it is also very well known that the costs of such collective action are enormous and that communities do not always succeed. A completely hands-off approach such as this may result in either aggregate poverty reduction goals not being reached, or some communities lagging very far behind. The latter case has serious inequality consequences in an already unequal country and should be carefully considered by policy makers.
Second, the policy is also extremely unclear about the membership criteria of these urban poor “communities”. It lays out no guidelines for how these groups are to be created, about who decides membership rules, and whether there are certain social, cultural or economic attributes that would be considered unacceptable, discriminatory criteria upon which to decide membership. The problem here is that creation of such groups is associated with relatively large economic incentives, there are no disincentives associated with already rampant social discrimination, and the result may be the formation of economically valued, socially exclusive groups that are efficient but inequitable. Even in the case of “neighbourhood groups” where spatial contiguity is provided as a criterion for group formation, social exclusion is still likely to play a prominent role. For example, urban slum neighbourhoods are usually sociocultually very diverse.
Personal experience with slum-based collectives has shown that households often segregate into groups on the basis of caste or other forms of social hierarchy. While these groups may act cohesively for purposes of political or economic mobilisation, the problem occurs when the success of a group is linked to attributes such as caste but the membership into this group is also based on attributes such as caste. This means that certain groups might be unable to exploit the opportunities created by a wide scope, economic decentralization scheme such as the NULM, but also that individuals from these groups would find it prohibitively expensive or socially taboo to join the other, more successful groups. All of these are plausible mechanisms by which the existing inequalities in a socioeconomically heterogeneous country such as India might be exacerbated.
 Sobhan, Rehman. 2009. Agents into principals : Democratizing development in South Asia. In: Arguments for a Better World : Essays in Honor of Amartya Sen / ed. by Kaushik Basu, Ravi Kanbur (2009), Vol. 2
 Draft policy notes can be found online at http://pib.nic.in/newsite/erelease.aspx?relid=79845