Research & Publications
Research & Publications
NEW POVERTY LINE: A CRITIQUE
H.S.Shylendra
The Expert Group (EG) headed by S. Tendulkar appointed to review the alternative conceptualisation of poverty and methods and procedures for its estimation submitted its report in November 2009. The EG has come up with a new methodology to arrive at state-wise rural/urban poverty lines to estimate poverty ratios. The present official poverty line has been arrived based on the per capita calorie consumption requirement (2100 calorie for urban and 2400 for rural per day) converted into monetary consumer expenditure measured at 1973-74 market prices. The EG has identified three major limitations of the official poverty line. The first major weakness is that the consumption basket underlying the poverty line arrived in 1973-74 has become too outdated. Second, the price indices which are used for adjustments have failed to capture fully the price rises resulting in underestimation of poverty. Lastly, the existing poverty lines have not explicitly accounted for expenditure on health and education as they were expected to be provided by the state.
The EG claims that it has arrived at a new poverty methodology which reflects marked improvements on four major counts. The improvements are: (i) unlike in the past the new poverty lines are detached from their underlying calorie norm; (ii) second, a uniform poverty line basket (PLB) is adopted for both rural and urban population based on the current consumption expenditure data; (iii) third, there is a construction of new price indices which are primarily based on the unit prices of consumption expenditure survey; and (iv) lastly, there is an attempt to explicitly incorporate private expenditure on health and education in the price indices to account for their increasing role in determining household health and education status.
Uniform Poverty Line Basket (PLB):
In order not to discriminate between the urban and rural population, the EG has recommended adopting a uniform PLB for the both areas instead of a separate one as in the past. The new PLB recommended is the Mixed Reference Period (MRP) equivalent per capita total consumer expenditure (PCTE) corresponding to all-India urban poverty ratio of 25.7 per cent in 2004-05. Keeping the urban HCR as the basis, the EG takes the corresponding MRP urban per capita total expenditure (PCTE) as the new reference poverty line both for rural and urban areas. The main rationale given by the EG for benchmarking this PCTE is that it helps the new method to be situated in some existing practice. Further, the official urban poverty ratio for 2004-05 is considered to be less controversial than the corresponding rural poverty ratio. Several advantages are claimed for the new PLB by the EG. The MRP data on which the PLB is based is considered more accurate and reliable than the Uniform Recall Period (URP) data. The new PLB identified can help capture a more recent pattern of consumer behavior.
Delinking from Calorie Norm:
The new method gives up the explicit linkage between the calorie intake norm and the official poverty line. The EG, has preferred the new PLB over the existing calorie based poverty line as it believes that there is growing evidence about the declining calorie intake overtime in the country. In addition, the actual calorie intake measured based on NSSO surveys is found not well correlated with the nutritional outcomes observed from other surveys. Despite the delinking, the EG says that the actual per capita calorie intake of those near the new poverty line is found largely matching with the revised norms (1800 calories) advocated by the Food and Agricultural Organisation (FAO) for India.
New Price Indices: For adjusting the poverty line for changes in the prices, the EG has proposed construction of new price indices which are based on the unit values of consumption items derived from the 61st round of NSS consumer expenditure survey. For health and education, the unit values have been derived from the other related NSSO surveys. The EG argues that the existing price indices viz., the consumer price index for agricultural labourers (CPIAL) and consumer price index for industrial workers(CPIIW) are outdated and have failed to capture the extent of price rise. The main advantage claimed for new indices is that they are based on the same set of data from which poverty lines are derived. Also they make allowances for the variations in consumption pattern across the states besides allowing for calculation of separate poverty lines and poverty ratios for North-Eastern states.
New Estimates: Based on the uniform PLB for urban and rural areas and using the new price indices, the EG has arrived at poverty lines for all states and has re-estimated the poverty ratios for 2004-05. For comparative purpose, a preliminary estimate has been made by the EG for 1993-94 by reworking the poverty lines. The new methodology yields a higher poverty ratio for rural and all-India level both during 2004-05 and 1993-94 (Table 1).
Table 1: New Poverty Head Count Ratios (HCR %)
1993-94 2004-05
Rural 50.1 (37.2) 41.8 (28.3)
Urban 31.8 (32.6) 25.7 (25.7)
Total 45.3 (36.0) 37.2 (27.5)
Figures in brackets are the existing official estimates.
For checking the consistency of the new poverty lines the EG has carried out validation checks by comparing the adequacy of the actual private per capita expenditures near PLB levels on food, education and health services with the normative levels matching the nutritional, education and health outcomes. The EG has found that the new poverty lines satisfy the adequacy tests for normative level of expenditures on food, health and education.
Assessment of the New Method:
The new methodology seen especially from a decentralised development perspective is top-down in nature devoid of any clear normative basis. Despite the mandate to look at an alternative conceptualization of poverty, the EG has failed to explore such a possibility. Having dodged the opportunity, the EG has tried merely to base its notion of poverty in the existing official practice and find reasons for its rationalization. The EG has continued with the narrow perception of poverty. Despite admitting that poverty is a multi-dimensional phenomenon, the EG has preferred a liberal stand of going by individual preferences to look at poverty over considering any social norm. It has also tried to be more practical by confining its focus only on private consumption expenditure. The improvement claimed by the EG in incorporating expenditures on health and education as such does not emerge out of any normative framework. This attempt in a way rationalizes the neo-liberal approach wherein the poor have to shell out more and more on the basic social needs. The earlier understanding that state would take care of these basic needs for the poor is diluted.
Having failed to look at poverty from any alternative perspective, the EG goes by a highly arbitrary approach to arrive at new poverty line. The major basis of EG for its PLB is the magnitude of the 2004-05 urban head count ratio (25.7 percent). Having a clear knowledge that the official methods in general grossly underestimate poverty, the EG still has taken recourse use the official poverty ratio as the basis. However, no clear evidence is presented in the report of EG to show as to how and why this ratio is less controversial than the rural ratio. Given the general statistical and sampling errors which afflict the official methods, the urban poverty ratio could not have remained isolated from the ill effects of such errors. A new poverty line is thus arrived based on a possibly erroneous HCR rather than going by a normative framework.
In adopting a uniform PLB there is an implicit notion in EG’s method that we no longer need to make any distinction between rural and urban when it comes to poverty and deprivations. The EG seems to be under the impression that it is being more magnanimous in suggesting a higher level of poverty line for rural India. It would have been more meaningful if the PLB for the rural area is worked out based on a normative consideration. Equating rural and urban areas for PLB implicitly goes with an ethnocentric view of inferring as to what is desirable for rural society, especially for the poor. What is the advantage or rationale of equating rural and urban is never explained or made out.
A major strength claimed for the new PBL is the delinking of measuring poverty based on calorie intake norm. The EG claims that it is a conscious decision as there is evidence of declining calorie intake over time and calorie intakes measured from consumer expenditure surveys have not shown good correlations with nutritional outcomes. The EG’s stand here appears more of an ex-post rationalisation than based on adequate reasoning. The first reason for rationalization is the EG’s failure to take a normative view to define a PLB. Also given the current low level of calorie intake reported for households below poverty line, one may not find good correlations with the nutritional outcomes. It may be hence the low levels of calories actually consumed that may be the problem in finding such correlations and not the normative calorie level suggested per se. It is quite obvious now that the poverty lines arrived over different surveys based on price adjustments have actually got diverged from the calorie norm initially specified.
The EG in way has been compelled to give up the calorie norm as the previously defined levels were found unacceptable to it. Instead of arriving at a normative level as per the current need, the EG has decide to delink the calorie norm from the PLB. This convenient stand of the EG becomes evident from the fact that after having arrived at the new PLB delinked form calorie norm, the EG goes on to test its adequacy for the new calorie norm as suggested by FAO. The EG has argued that there is a clear downward shift in calorie Engel curve. However, the report lacks adequate evidence to substantiate the argument that the observed intakes are the result of the declining need by the rural poor. The external validation checks carried to test the consistency of new poverty lines in relation to nutritional, education and health outcomes also suffer from certain limitations. For validation, the normative levels of outcomes have been arrived after the estimation of the poverty lines. Thus, normative values have been used for validation and not for defining poverty. Even the normative levels of expenditure on food, health, and education have been arrived based on actual expenditure incurred by the poor on these items corresponding to the average levels of outcomes in a state for the poverty group.
Lastly, there is a need to test the new price indices constructed for their accuracy and authenticity. As the price data is derived from the unit values of the NSSO survey, one has to examine the reliability of these unit values. As admitted by the EG itself the proposed price adjustments have made the new methodology become very complicated. The new method has rendered the new poverty ratios non-comparable to past estimates needing fresh calculations. The validity for applying a new methodology for past consumer expenditure data needs to be fully established. A glaring lacunae that cannot be missed is the effort of the EG to retain the state population-weighted urban poverty at 25.7 per cent. The EG has allowed for variation in the ratios across states only as long as the all-India average remains the same.
To conclude, the new methodology suggested by EG breaks no new ground. The top-down and arbitrary nature of official methodology continues. The proxy of using consumer expenditure has not only given scope for inaccurate assessment of poverty but also misunderstanding of poverty. The nature of poverty has changed both absolutely and relatively. It has deepened in the absolute sense due to problems like widespread malnutrition. It has worsened in the relative sense because of the growing social and economic inequality. It is high time that the top-down statistical exercises are radically altered. For this we need to look at poverty from a wider angle so as to capture both absolute and relative deprivations. Instead of a narrow calorie based norm we could possibly move towards measuring developmental outcomes which reflect wide variety of deprivations across different sections. This could include measuring conditions like malnutrition, absence of chronic diseases, level and quality of education, housing and sanitation conditions, ownership of assets, and nature and magnitude of employment.
-------------------------------------------------------------