Often, the implementation of anti-poverty programmes has been evaluated on the basis of their outcomes rather than the process of implementation. Such methods of evaluation assume that good outcome indicators are the result of good implementation. 

This evaluation methodology assumes that the programme was 'prescribed' as a solution to a problem (the 'rational' approach), and better the outcome indicators, more is the problem reduction. However, with the emergence of public choice theories, 'policy process' has been the key point of inquiry. In developing economies, where the public authorities responsible for designing and implementing redistributive interventions are not sufficiently independent from the interests of the dominant sections of society, the approach of investigating policy process can yield to useful results.

Such an approach also can help us to understand the deficit of theory and practice of social protection programmes. This paper looks into process deficits of income maintenance programmes designed for agricultural labourers in the Indian state of Maharashtra, and its impact on people in persistent poverty. Process deficits have been studied not merely as 'lapses of government programmes'. 

They are critically examined in the contexts of the political conditions that perpetuate these deficits. Loyalty to local elites over the span of one's work life increases the possibility of income maintenance at times when work is not available. This enables local elites to avail cheap and instant labour for their private works. Protesting voices of the poorest people are well-managed by elites through the provision of such incentives. 

The conclusion of the paper is that anti-poverty programmes, at the implementation stage, strengthen local elites' capacity to wield power and support their own private interests. On the other hand, at the evaluation stage, outcome indicators are used to bolster the legitimacy of the state. Above all, the ability of local elites to manufacture outcome indicators point to a need for rethinking evaluation methods.