Network Past Issues

IDENTIFYING THE POOR: AN ASSESSMENT OF THE NEW BPL METHOD
H.S. Shylendra

Background
Poverty has been a major challenge both on the conceptual and policy front. A clear understanding of the nature and magnitude of poverty is considered essential for a meaningful action on poverty. The official poverty methods, in vogue for measuring poverty and identification of the poor, have come under criticisms for being highly top-down, and narrow in perception. About the methods adopted for actual identification of poor at the micro level for providing various developmental assistances, a major issue has been the error of high exclusion of the poor and inclusion of the non-poor.

A new methodology has been proposed for the identification of the poor in the recently brought out ‘Report of the Expert group to Advise the Ministry of Rural Development on the Methodology for Conducting the Below Poverty Line (BPL) Census for 11th Five Year Plan (August 2009).’ The Expert Group (EG) headed by N.C. Saxena was constituted by Ministry of Rural Development with the explicit purpose of identifying a more suitable methodology, to recommend institutional system for conducting survey and redressing grievances, and to look at the issue of the relationship between estimation and identification of poor and putting a cap on the total number of BPL households. This paper examines critically the proposed BPL methodology focusing on its relevance from the perspective of decentralised planning, and suggests broad directions for addressing the challenge of identification of the poor.

Poverty Ratio and the Cap
The EG starts with the right premise that there is a need for universal coverage for basic entitlements like food. However, the EG goes on to argue that there are several programmes which may need targeting and a cap on the number of BPL households. In order to arrive at a reasonable cap, the EG has gone into the issue of poverty estimation on its own. The EG based its analysis on available evidence, and reached the conclusion that there is a large scale underestimation of rural poverty by official agencies. A large section of the poor who are calorie deprived are not being captured by the official poverty line. The official poverty line initially arrived based on a calorie norm of 2400 for rural areas updated based on consumer price indices no longer corresponds to the level of expenditure required to meet the minimum calories. There is a major deviation in the calorie consumption at the poverty line level as compared to the prescribed norm. As per EG, the actual level of calorie intake is not only lower for the poverty classes (bottom 50 per cent) but has been showing a declining trend. The EG is of the view that this decline is attributable more to the declining purchasing power of the poor than to the perceived notion of declining calorie needs. The EG, hence, argues that the official poverty line and rural poverty ratio of 28.3 per cent are artificially low.

The EG argues that the poverty line has to be corrected to arrive at realistic numbers before the BPL census could be held. The EG has recommended that the level of rural poverty be fixed at 50 per cent as it compares well with other indicators like the actual nutritional status and poverty estimates made by World Bank for India. The EG advocates that this ratio be accepted and a pro-rata adjustment is made in rural poverty ratios of all the states. These state level poverty ratios are then translated into district, block, and village (panchayat) level ratios. The poverty ratios so arrived would serve as the cap or limit for conducting the new BPL survey. The districts with a poverty ratio of over 80 percent are to be excluded for the purpose of census.

BPL Census Methodology
According to EG, a BPL method has to be simple and less prone for errors. The EG has arrived at three broad methods for identification of the poor: (i) Automatic Exclusion; (ii) Automatic Inclusion; and (iii) Grading the rest. For exclusion, the visible rich are to be identified using local knowledge and official information. Five criteria are suggested in this regard viz., owning land double (wet) or treble (dry) the district average, possessing three or four wheeled motor vehicles, having at least one mechanized equipment, having a person drawing regular salary of over Rs. 10,000 p.m. employed in government or private sector, and paying income tax. States, if needed, may add other indicators to the criteria. Primitive tribal groups, most discriminated scheduled caste (SC) groups, single women headed households, households headed by disabled person and minors, destitute and houseless households, and households with bonded labourers are the groups suggested for automatic inclusion. The panchayat has to take due care of both immigrants and out migrants to avoid exclusion.
The rest of the poor are to be identified based on grading on a ten point scale. The households are to be surveyed on a five parameter schedule covering the caste/religious status, nature of occupation, education level of adults, prevalence of a chronic disease (TB, HIV/AIDS, Disability), and age of the head of the household. The survey is to be conducted by an earmarked government official from a block or mandal, different from one in which the village panchayat is located. A team of three residents (consisting at least one SC/ST and one women) nominated by the panchayat will accompany the government official. The list with gradation is to be authenticated by gramasabha in a transparent way. For identification, no cut-off score is suggested and the households are to be merely graded in ascending order of score. In case of similar score, the SC/ST groups may be placed above in the list preferably by the gramasabha.
Though a household will be the unit for survey, all single women, old individual above 60 years, adult with chronic diseases, and bonded labour with spouse and children are to be treated as a separate household. The list will be revised every 2 years for various changes, and the survey results will be valid for 10 years. Any household can seek correction by applying to gramasabha. The application has to be disposed off within a month’s period with a scope for appeal to the CEO of the district panchayat.
Assessment
The EG views that all basic entitlements like food, health, education and work should be provided on a universal basis, and BPL method should not curtail universal access to these services. The EG has clearly identified a strong case for enlarging the category of poor as the official poverty numbers have resulted in widespread exclusion of the poor from accessing basic entitlements. While the stand of EG on underestimation is noteworthy, the method adopted by it to arrive at poverty ratios and fixing a cap for BPL survey is highly questionable. Apparently, the EG has gone beyond its terms of reference here. Moreover, despite appreciating the need for universalisation, the EG fails to recommend universal provision even for food. Several compulsions seem to have influenced the EG forcing it to adopt a middle path between universalisation and targeting.
Moreover, the poverty cut-off arrived by EG is arbitrary and top-down. The method adopted is based more on practical than on normative considerations. The poverty line has been arrived based on the rural poverty ratio (50 per cent) the EG was comfortable with. The ex-post normative basis of 2100 calories and 12.25 kgs of cereals comes up more as a rationalization which goes against the norm that national poverty line and ratio be arrived based on weighted aggregation from state level.
The most glaring limitation is the fixing of cap for the BPL census. The existing state-wise cap imposed for BPL census by the Planning Commission based on official poverty estimation has been severely criticized. The EG has gone one step further and has even fixed district/block and panchayat level caps for poverty estimation. The BPL numbers to be arrived based on the new method have to adhere to the cap recommended. The EG believes that the states and PRIs which are responsible for conducting the BPL census are susceptible to many local biases. The imposition of the cap goes against all norms of decentralized planning and development as PRIs are reduced to mere intermediaries in identifying the poor. No scope is provided for any valid local deviation in the number of poor.
Given the varied constraints, the three-pronged approach suggested by the EG may help to an extent in reducing the exclusion and inclusion errors. The addition of automatic inclusion of the poor is the major strength. The criteria suggested for inclusion and exclusion however suffer from quite a few limitations. There is no scope for identification of households based on local knowledge and situation. Given the local peculiarities, it would be useful to identify the poor based also on some locally relevant indicators. It is also not fully clear as to how the exercise for exclusion and inclusion will be actually carried out. The EG assumes that panchayats will have full information and hence would prepare the list on their own which may not be true as many of the data are not available with the panchayats. A survey or a PRA may become necessary for this purpose. It may help even to include a voluntary exclusion category wherein the better-off households through awareness may be encouraged to file declarations for such exclusion. The land based exclusion may face difficulty due to benami holdings. About automatic inclusion, while the EG has identified many well deserving categories, the challenge will be in ensuring that they are proactively identified by the panchayts. It would be useful to actively involve some local credible NGOs and gramasabha for such inclusion. The panchayats may give suitable publicity to the inclusion criteria and invite such households to come forward.

As regards the grading, the EG’s admitted rationale is to follow a simple formula which is transparent and does not create any disincentives for the poor. In the criteria suggested there is an emphasis on identifying the chronic poor based on developmental outcomes as well as social and economic structures which influence poverty. Thus poverty is seen as a multidimensional phenomenon. As per the scores assigned, indicators focusing on caste/religion and occupational status become the key determinants of the poverty status. Again within these two indicators, certain groups like SC/ST households or agricultural labour households with or without land, given their historical status, have been assigned higher scores so that they have a greater chance of being graded high. Other indicators focusing on education, health and age serve more as supplementary variables for inclusion. The assignment of a small weightage for the minority Muslim community under the social status indicator though comes as a surprise and appears to have been influenced by the findings of the Sachar Committee.
Compared to the 2002 BPL method the scoring suggested is much easier with a fewer number of indicators (5 instead of 13) and a simple scoring scale. However, the parameters may pose many operational and other difficulties. The effort to simplify the scale may possibly ignore the wide variations that exist with regard to poverty. About occupation, there is the same mistake of assuming sole occupation by the households. The parameter would pose difficulty at the time of enumeration. With regard to education, the rationale for considering only adults is based on the assumption of perverse incentive likely to be created for the family if young children are included for the purpose. Without a strong basis it would be inappropriate to assume that poor would desist sending their children to school just to retain the BPL status.
Another concern is the inability of the proposed ten-point scale to differentiate adequately the depth of poverty. As caste and occupation are the two prominent variables, most of the poor households are likely to score between 3 and 7 on the scale. Giving preference to the real needy might pose difficulty as many are likely to have same scores. In such cases, EG has recommended giving preference to SC/ST and landless households. It would be more useful to involve gramshabha for such prioritisation.
The EG’s approach in defining a household for survey purpose also has given scope for confusion. In defining a household, the EG recommends that single women, old individuals, adults with chronic diseases and bonded labourers may be treated as a separate household. However, care has to be taken to ensure that these individuals are not forced to seek a separate livelihood for the sake of BPL by foregoing the protection of their primary household.

About the grievance mechanism, assigning the role for gramsabha here by the EG is a right step. However, the gramasabha has to be clearly vested with powers to include or exclude genuine cases. Otherwise, the grievance mechanism may lose its significance, as it has happened in the past. About the actual survey, the EG has overlooked the past experience. There have been glaring errors in the survey processes as different states had used different types of agencies. The involvement of credible NGOs with the PRIs in the survey should be commonly encouraged.

Conclusion

The new proposed BPL methodology while certainly is an improvement over the past methods, is unable to break any new grounds. Hence, it may not be able to address the major pitfalls of exclusion and inclusion error. The way forward for the BPL quagmire lies in a multi-pronged decentralised strategy. At least for few essential services like food, employment, water, and schooling universalisation is the answer until a large section of the population continues to be deprived of these basic needs. Universalisation for basic services would even obviate the need for a BPL survey. Further, we must make PRIs both responsible and accountable for identification of the poor. Instead of merely employing them as surveyors, PRIs must be empowered to take up the exercise as a part of the decentralized planning process. Let them plan, identify the poor and disburse the assistance as per need. Further, for services which are largely demanded by the poor only, self-targeting schemes like NREGS and ICDS would be of help. Even PDS can be made to become a self-targeted scheme provided it includes, even partly, items like millets consumed by the poorer sections. There can be also linkages across schemes for better targeting. Households working regularly under NREGS can be made natural claimants for assistance under schemes like PDS.

--------

This is an abridged version of a forthcoming paper by the author on New Poverty Mehods.

The author can be contacted: hss@irma.ac.in