Ebook Institute for Financial Management and Research Centre for Micro Finance
Nearly all poverty alleviation programs target a particular sub-population. This feature is most readily apparent in programs designed to aid those who have suffered a particular tragedy, such as grants to widows of debt-ridden Maharashtra farmers, but is also generally true of large, broad based development interventions. At first blush, this may seem unremarkable and not to warrant particular consideration. But effective identification of the target population is crucial to the success of aid programs. If, for instance, households which are adequately nourished are identified as eligible for subsidized food, the program is unlikely to significantly reduce malnutrition.
When the targeted population is not distinguished by a well-defined, observable trait, however, identification of the intended population may be complicated. Evidence suggests that the targeting efficiency of aid programs is less than perfect. A report by the Indian National Sample Survey Organization found that 18% of the wealthiest 20% of the rural population (ranked by monthly per capita expenditure) held Below Poverty Line (BPL) rationing cards. That targeting inefficiency has real consequences is apparent from a 2006 story in The Hindu which reported on street protests carried out by families who had been denied their ration cards.
In this study, we evaluate the targeting efficiency of various assistance programs operated by the government of India. We find that the methods used to identify eligible house-holds do not particularly target the poorest of the poor. In our sample, those who receive government assistance do not appear worse off, according to our measures of poverty, than households which do not.
We also evaluate the targeting efficiency of a process used by Bandhan, a Kolkata based micro finance institution, to identify households eligible to participate in one of their programs designed to assist the poorest of the poor. This program offers eligible households grants consisting of income generating assets (livestock, inventory, etc.) as well as training and assistance operating a small-scale enterprise. The goal is to assist destitute households to establish a regular income source.
Our results indicate that Bandhan’s process successfully targeted a group which appears poorer in various respects, particularly land ownership, assets and credit access. Along other dimensions of poverty, such as expenditure, the results are less crisp; it does not appear that per capita consumption among the identified group is less than among those not identified as eligible. This may be in part due to the fact that these families are smaller, so that their total expenditure (not per capita) is made smaller.
While we focus on this particular intervention, our study has broader relevance since the identification process employed in this setting included a Participatory Rural Appraisal (PRA). PRA’s are widely practiced by NGOs, both within India and internationally, when conducting development interventions. Increasingly, PRA methodologies are used to identify beneficiaries for assistance programs. Consequently, it is important that the information collected from a PRA accurately reflects the conditions within the village where it was conducted.
There is some evidence suggesting that certain types of information can accurately be obtained using PRAs (see Chambers (1994) for an overview of various results). Specifically in West Bengal, Chattopadhyay and Duflo (2001) collected information on the infrastructure (water systems, etc.) of various villages using PRAs. The infrastructure was subsequently inspected to verify the information from the PRA, revealing that this information was highly accurate. In this study we assess the reliability with which PRAs can accurately rank village residents according to economic status.
Specifically, we evaluate how well our measures of poverty accord with the evaluation of poverty established by the PRA. This analysis reveals that those ranked as most poor in the PRA are in fact poorer than others in very observable dimensions such as land and asset ownership. They also have less access to credit. This suggests that the PRA can generate a reasonably good indicator of economic well-being and can serve as the basis for targeting.
Contents
1. Introduction
2. Data and data collection
3. Targeting efficiency of government aid programs
4. Analysis of Bandhan’s identification process
- 4.1. Overview of Bandhan’s “targeting the ultra poor”
4.2. Details of the identification process
4.3. Analysis of the PRA process
4.4. Analysis of Bandhan’s verification process
5. Conclusions
References
Annexure 1: Figures
Annexure 2: Tables
Download
PDF Ebook Institute for Financial Management and Research Centre for Micro Finance
Posted in :
