MEASUREMENT OF POVERTY IN INDIA

Transcription

1 CHAPTER 5 MEASUREMENT OF POVERTY IN INDIA Measuring poverty excites controversy! 5.1 Introduction 5.2 Chronological Evolution of Poverty Lines Pre-independence India Poverty Estimation by Dadabhai Naoroji National Planning Committee The Bombay Plan Post-independence India 5.3 En-route Summary Pre-cursor to Official Measurement of Poverty Working Group Poverty Estimation by Dandekar and Rath Task Force on Projection of Minimum Needs and Effective Consumption Demand The Expert Group on Estimation of Proportion and Number of Poor Expert Group to Review the Methodology for Estimation of Poverty Expert Group to Review the Methodology for Measurement of Poverty

2 5.1 Introduction Measuring poverty is not a simple and straightforward task but a demanding issue fraught with difficulties. The problem arises fundamentally due to definitional contentions among scholars, experts and policy makers. This contention has led to various suggestions from experts and State representatives as to what is poverty and what should constitute as the benchmark poverty line. Also, this has given rise to a seemingly unending debate between two broad opinions those who agree with the official methodology and those who do not agree. Fault finding with the methodologies largely harps on the insufficiency of consensus as to what is poverty? Interestingly, the subsequent allegations of underestimation or overestimation of the number of poor keep the media persons busy! Poverty measurement is vital due to two reasons: One, poverty measurement serves a normative role in helping the classification of the population of the society or economy into poor and the non-poor individuals or households. Secondly, poverty measurement serves a monitoring role and therefore estimates are important inputs to designing, implementing and examining appropriate anti-poverty initiatives. The World Bank initiative on poverty reduction and equity considers numerous dimensions of well-being such as income, consumption, health, education and assets ownership, but lays emphasis largely on the income and consumption dimension and less frequently to the other dimensions. If measurement of poverty in India has been an issue involving conflict, internationally too there have been disagreements over the definition of welfare measures and therefore choice and estimation of poverty. Broadly, the two popular methods of setting the poverty line in the conventional money-metric procedure are the cost of basic needs approach and the food-energy-intake method. However, if poverty is given a multi-dimensional definition - adding attributes to be considered, the degree of complexity in measurement increases. Talking of poverty measurement in the United States, the Census Bureau makes use of a set of money income thresholds to determine poor and non-poor. This income threshold varies according to the family size and composition. The official poverty thresholds do not vary geographically, but they are periodically updated for inflation Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 67

3 using the consumer price index. The official poverty definition uses money income before taxes and does not include capital gains or noncash benefits (such as public housing and food stamps). Poverty in Bangladesh was earlier measured by direct calorie intake method. Since , Bangladesh Bureau of Statistics is using the Cost of Basic Needs method as the standard method for estimating the incidence of poverty. In this method, two poverty lines are estimated: Lower poverty line and Upper poverty line. Within these, a food poverty line and a non-food poverty line is calculated. For setting the food poverty line a basic food basket comprising of eleven food items is chosen and thereafter the quantities in the basket are scaled according to the nutritional requirement of 2,122 kcal per capita per day. The estimate of the expenses of acquiring the basket thus obtained is the food poverty line. The nonfood poverty line is calculated by estimating the cost of consuming nonfood items by the households close to the food poverty line. In India, from the cost of basic needs definition, average energy norm definition, to a more holistic view of poverty, the officially declared and the unofficially advocated poverty line has undergone important changes with the passage of time. Also, the methodologies of data collection and coverage have become more advanced and widened respectively. Poverty measurement in India has indeed served the monitoring purpose despite the complexities and numerous disagreements while the normative purpose remains questionable due to the questionable nature of the very norm that defines a person as poor or non-poor. 5.2 Chronological Evolution of Poverty Lines Pre-independence India As discussed earlier, poverty has been part and parcel of India s socio-economic history. During the British rule, massive outflow of wealth has been held responsible for the miserable state of people (Naoroji, 1876). The British system of arbitrary and repressive administration caused widespread poverty. Dadabhai Naoroji stated that, under the present system of administration, India is suffering seriously in several ways and is sinking in poverty. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 68

4 Poverty Estimation by Dadabhai Naoroji (1876) Without using the term poverty line Naoroji estimated the subsistence needs of the people and consequently arrived at a subsistence based poverty line at prices. In his eye-opening book, Poverty and Un-British rule in India, Naoroji refers to subsistence as something that is essential for meeting the basic wants of human beings to keep them in ordinary good health and decency. Thus he estimated necessary consumption, using the diet for emigrant coolies as prescribed by the Government Medical Inspector of Emigrants (on March, 26, 1870). The prescribed intake was supposed to provide the necessary inputs to the emigrant coolies during their journey ( in a state of quietude being transported to various British colonies) and therefore the energy composition was not sufficient for working individuals. He found that similar prescription was mentioned in the Administrative Report of and both of these comprised of the following Table 5.1: Estimation of necessary consumption by Naoroji (1 ounce = grams) Rice diet for one person (in ounce) Flour diet for one person (in ounce) Rice Flour Dhal 6.00 Dhal 4.00 Preserved Mutton 2.50 Preserved Mutton 2.50 Vegetables 4.27 Vegetables 4.27 Ghee 1.00 Ghee 1.00 Mustard Oil 0.50 Mustards Oil 0.50 Salt 1.00 Salt 1.00 Total Total Source: Naoroji (1876) Though in the administration report, items like tobacco, firewood, and spices were included but these were not considered by Naoroji for the calculation of the cost of basic survival. The Government Medical Inspector s diet requirements were converted into monetary terms using the prices prevailing in Ahmadabad at that time. Similarly, he also calculated the subsistence level cost of living for various regions of the country by using various other sources of diet data such as the expenses on the sustenance Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 69

5 requirements of prison inmates in a number of Indian provinces and the necessities of an ordinary agricultural worker in Bombay Presidency. Important among the assumptions that he considered for arriving at the subsistence level of expenses included the notion that one half of the children population (about 34%) did not cost anything and the remaining half cost one-half of the adult subsistence requirements. Consequently, he arrived at the subsistence costs of living that ranged between Rs. 16 to Rs. 35 per person per annum in various regions of the country. Naoroji remarked that even in times of good harvest the total produce was insufficient to meet the basic requirements of food and clothing as afforded by a criminal. There was little possibility of saving for deficiency due to crop failure. To talk of the times of bad harvest was an even sadder story. Naoroji s calculation and measurement of the poverty lines in the pre Independence India is a landmark in the history of poverty evaluation in India and in fact the first ever attempt of such a scale and commitment that defines the earliest poverty line. It is important to note that the subsistence requirements considered in the calculations by Naoroji are not sufficient for work let alone the luxuries which social and religious expectations draw from an individual National Planning Committee (1938) In 1938, Indian National Congress, then headed by Subhash Chandra Bose at its Haripura session appointed a National Planning Committee with Pundit Jawaharlal Nehru as the Chairman and K T Shah as the Secretary. For the pre-world War II period, it was noted that the minimum requirement had been estimated by economists, to range between Rs. 15 to Rs. 25 per capita per month (IIAPR, 1988) whereas the approximate estimate of the average income per capita per annum was Rs. 65 while the average income of the villager was estimated to be merely Rs. 30 per capita per annum. The Committee acknowledged the absence of sufficient food, clothing and shelter for subsistence living and set forward an aim ( fundamental aim of planning ) to ensure an adequate standard of living for the masses and get rid of the poverty of the people (Nehru, 1946). Nehru (1946) states that the National Planning Committee calculated and then aimed at a 200 to 300 per cent increase in national wealth within ten years to ensure a progressive Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 70

6 standard of living. The Committee also suggested some objective tests that included - The improvement of nutrition from the standards of an irreducible minimum requirement of proteins, carbohydrates and minerals (as well as necessary protective foods) having a calorific value of 2,400 to 2,800 units for an adult worker; improvement in clothing from the then consumption of about fifteen yards to at least thirty yards per capita per annum; housing standards to reach at least 100 square feet per capita. The report of the committee is silent on the names of economists who contributed in the calculation of the cost of irreducible minimum standard of living as mentioned above. However, an intelligent guess is presented by Srinivasan (2007) that the estimates were based on the research studies done by Gilbert Slater in 1905 (covering the prewar village status) and surveys of the Punjab Board of Socio-Economic Inquiry The Bombay Plan (1944) In 1944, the Bombay Plan was proposed by a group of industrialists charting out the developmental (and reconstruction) aspirations of the capitalist class for India after independence. This discourse that sought India s independence and set the agenda for future course of action clearly echoed the belief of wealth drain theory as put forward by Dadabhai Naoroji that British colonial rule had accentuated poverty in India. The Plan stated in its aim that in the initial 15 years of India s rebirth there was the need to achieve - a general standard of living which would leave a reasonable margin over the minimum requirements of human life (Thakurdas, et al, 1944). The authors considered food, clothing, shelter, healthcare and basic education in the minimum requirements and specifically 2800 calories of well-balanced food per day, 30 yards of clothing per year and 100 square feet of housing as the minimum need for an individual. They arrived at Rs.74 as the minimum per capita per annum monetary resource required for attaining the said minimum at the pre-war prices. Using this they arrived at the target for the per capita income and proposed that per capita real income of the time would have to be doubled to attain the targeted consumption and standard of living. Thus, in the pre independence years though the estimation of poverty was done it was largely a non-institutional attempt. The definition of poverty and therefore the poverty line was food or diet centric and the requirements beyond subsistence were not Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 71

7 emphasized. Nevertheless the importance given to the issue and the concern for removing poverty among the national leaders is commendable Post-independence India Independence seemed like a wishful thinking in the pre-independence years. But the reality dawned at the stroke of midnight of August 1947 and India became a free state with the baggage of deficiencies and inadequacies of almost all kinds. Pundit Jawaharlal Nehru in his historic speech Tryst with destiny remarked explicitly that ending of poverty and ignorance and disease and inequality of opportunity as one of the major nation building tasks facing modern India (Gupta, 2008) Pre cursor to Official Measurement of Poverty The question and need for an official poverty line was for the first time proposed as well as formally discussed among scholars and policy makers at the Indian Labour Conference in 1957 (GOI, 1979). However, Dr. Rammanohar Lohia s sharp remarks in a speech in the Lok Sabha in 1960s drew emphatic attention towards the issue of widespread poverty in India (Rath, 2011). He had written a pamphlet comparing the amount spent on the then Prime Minister (Pundit Nehru) with the amount on which the vast majority survived. He specifically stated that most Indian lived on less than 3 annas a day which was less than the expenses incurred on the upkeep of the Prime Minister s pet dog. Though Nehru raised objection by seeking the backing of the statistics provided by the Planning Commission that the average daily income of the poor was 15 annas, he took the remark by Dr. Lohia seriously. Subsequently, Pundit Nehru set up a committee to report to the Planning Commission about their findings in this regard. (Rath, 2011) The famous teen anna pandrah anna debate set the stage for defining a poverty threshold for India in the post independent years Working Group (1962) A task force known as the Committee on Distribution of Income and Levels of Living was setup to evaluate the impact of a decade long economic planning. Also, a working Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 72

8 group was set up to define a national subsistence minimum level. Constituted in 1962, the group had eminent statisticians, economists and nutritionists among its members. The team comprised of Prof. D.R. Gadgil, Dr. B.N. Ganguli, Dr. P.S. Lokanathan, Shri M.R. Masani, Shri Ashok Mehta, Shri Pitambar Pant, Dr. V.K.R.V. Rao, Shri Shriman Narayan, and Shri Anna Saheb Sahasrabuddhe. The working group made the following recommendations (GOI, 1962) - a. The national minimum for each household of five persons (four adult consumption units) should not be less than Rs. 100 per month at prices or Rs. 20 per capita. For urban areas, the figure would have to be raised to Rs.125 per household per capita or Rs. 25 per capita to cover the higher prices of the physical volume of commodities on which the national minimum is calculated. b. The national minimum excludes expenditure on health and education, both of which are expected to be provided by the State according to the Constitution and in the light of its other commitments. c. An element of subsidy in urban housing will have to be provided after taking Rs.10 per month or 10 % as the rent element payable from the proposed national minimum of Rs.100 per month. Thus the committee set Rs.20 per person per month and Rs.25 per person per month as the poverty line for rural and urban India respectively considering the amount sufficient to lead an active and healthy life. The methodology used for arriving at these figures remains unknown and there is absence of any record thereof. (Rath, 2011) A cursory comparison of the poverty line proposed by the National Planning Committee (1938) at 1938 prices as Rs per person per month with the recommendations of the Committee on Distribution of Income and Levels of Living (1962) at prices as Rs.20 per person per month, the latter is more humble, given the inflation over these years. (1938 to 1960) The benchmark minimum level of expenditure decided by the Working Group in 1962 was followed for all purposes during 1960s and 1970s even though the underlying methodology was starkly unknown. In 1962 the Perspective Planning Division of the Planning Commission prepared and released a paper Perspectives of Development: : Implications of Planning for a Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 73

9 Minimum Level of Living, under the direction of Pitambar Pant. (Srinivasan and Bardhan, 1974) The paper laid loads of emphasis on poverty removal. It stated that every person should be assured of a minimum level of income (taken as the one recommended by the working group of 1962) within a realistic time and the minimum should be revised with the passage of time as the economy progresses. Srinivasan (2007) highlights an important distinction that the paper makes between the baskets of goods that households need to acquire on account of their own earning and the one which was expected to be extended by the State as a part of its basic functions - such as health and education and the value of the former basket is taken as the poverty line. Interestingly, the poverty benchmarks suggested in the pre-independent India (Naoroji, National Planning Committee and the Bombay Plan) and the post-independence suggestion by working group (1962) were all referring to a subsistence or minimum norm that drew majorly from the survival requirements for food/energy, but none of them built a connect between food/energy requirements and the poverty line. The connection was for the first time in India successfully attempted in the empirical investigation of poverty by Dandekar and Rath in Poverty Estimation by Dandekar and Rath (1971) V.M. Dandekar and Nilkanth Rath made a seminal contribution in poverty estimation in Sukhatme (1965) using the data from the National Institute of Nutrition found the energy requirements for an average Indian, keeping in view the average age, gender and work nature, as kilocalories daily. Dandekar and Rath (1971) used this finding in their study to define poverty in India in the circumstances prevailing in those years. They considered the ability or otherwise to get two square meals daily as the deciding factor for categorizing one as poor or not. Thus via empirical research, they linked calorie intake with the poverty line. Using the data provided by Professor P V Sukhatme and they arrived at the figure of 2250 kcal. Consequently, the researchers defined poor as those who were short of income that did not allow them to access food provisioning 2250 kilocalories. This was the poverty line proposed by them common for rural and urban India. For the monetary conversion of the energy norm, the National Sample Survey (NSS) data for consumer expenditure was used to arrive at the per Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 74

10 person monthly expenditure. At prices this was obtained as Rs in rural and Rs in urban areas. In other words, Rs per month was sufficient to fetch a resident of rural India 2250 kcal and Rs per month could fetch an urbanite the minimum energy norm of 2250 kcal. These were thus the poverty lines for rural and urban India respectively. A comparison of the poverty line put forward by the working group (1962) and the poverty line estimated by Dandekar and Rath shows that the latter is a more modest figure. They too observed and noted this fact and eventually scaled it up to Rs per person per month for rural India and the urban poverty line being close to their estimate was rounded off to Rs per person per month. (Bandyopadhyay, 2010) But unlike the working group (1962), Dandekar and Rath specify the nutrition norm and the relevant set of prices to arrive at the poverty lines for rural and urban India Task Force on Projection of Minimum Needs and Effective Consumption Demand (1979) After a decade and a half (on July 30, 1977), the Perspective Planning Division of the Planning Commission set up another committee known as the Task Force on Projection of Minimum Needs and Effective Consumption Demand under the chairmanship of Dr. Y K Alagh. Over the previous years a number of academic research studies had been undertaken relating to the issues of poverty estimation at the Indian Statistical institute (Calcutta), Sardar Patel Institute of Economics and Social Research (Ahmadabad), and different insights by independent researchers had also been widely debated and discussed. (Planning Commission, 1979) With the aim to consolidate the precious ideas and information on the subject, the task force was expected to come up with a clear private consumption model for the ensuing Five Year Plan. The terms of reference of the task force were stated as - to examine the existing structural studies on consumption patterns and standards of living and the minimum needs with particular reference to the poorer sections of the population for the nation as a whole and its different regions separately by rural and urban areas; on the basis of the above studies, to forecast the national and regional structure and pattern of consumption levels and standards for the end of the Sixth Plan and subsequent perspective plan taking into consideration the basic minimum needs as well as effective consumption demand. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 75

11 Having reviewed large literature on demand and poverty estimation, in its report the task force has expressed critical opinion on the studies done in the area by independent researchers as well as the working group (1962). The task force presumes that the poverty line suggested by the working group (1962) was based on the balanced diet recommendations made by the Nutrition Advisory Committee of the Indian Council for Medical research in Also, the report of the task force expresses concern on the absence of reasoning underlying the minimum calorie requirement of 2250 kcal per person per day as suggested by Dandekar and Rath. However, Rath (2011) in retrospection has clearly stated the underlying logic - the study by P V Sukhatme. The task force estimated the poverty lines for the rural and urban areas separately and used the age-sex-activity specific calorie requirements as recommended by the Nutrition Expert Group (1968) The Group had classified the people into children (categorized by age), adolescents (categorized by age and gender) and workers (categorized by gender). The population estimates provided by an Expert Group on Population formed as a combined effort of the Perspective Planning Division (Planning Commission) and the Office of the Registrar General, Census of India in 1977 were used by the Task Force. For information on the occupational structure, the Census data for 1971 was used while the usual status participation rates for the workers was obtained using the National Sample Survey data on Employment for the 27 th Round of The weighting schema was obtained by taking the averages using the estimates for the age-sex-occupation structure of the population for Thereafter this weighting framework was applied to the category specific energy norms as suggested by the Nutrition Expert Group. Women in pregnancy and lactating phase were assigned additional calories per day. Consequently the Expert Group came up with the calorie requirement per person per day as 2435 kilocalories in rural areas and 2095 kilocalories in urban areas. (These figures were rounded off for convenience and usually literature on poverty studies quotes 2400 kilocalories and 2100 kilocalories for rural and urban India respectively as the estimates provided by Task Force of 1979) The Expert Group fully acknowledged the bias which may have crept into the estimates worked out by them - making the real requirements more or less than the estimates due to various reasons, for example disguised employment. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 76

12 The monetary equivalence for these poverty lines was obtained using the data on household consumption (both in quantitative and value terms) by the 28th round of National Sample Survey conducted in the year On the basis of the consumer behavior exhibited and captured by observation for the year , calculations led to Rs per person per month as the poverty line for rural areas at prices. Similarly Rs per person per month as the poverty line for urban areas at prices. When the prevailing prices, that is, those for were used, the poverty lines were obtained as Rs per person per month and Rs per person per month in the rural and urban areas respectively. The estimated poverty lines enabled access to clothing, shelter, education etc. besides the calorie norms in the consumption basket of an average rural or urban Indian. In other words, the focus was not on the energy requirement for survival but on the purchasing power required to meet the calorie norm intake with some provision for non-food items also. The proposed poverty line was therefore developed by combining the normative and behavioral concept of poverty. The calorie norms obtained by the Task Force (1979) were used for identifying poor and the extent of poverty in the following years. Thus the consumption basket remained the same and the monetary equivalent for any year was obtained by updating using the changes in the cost of living (due to changes in prices of various goods and services) as reflected by the Wholesale Price Index (WPI) for that year. The use of the Wholesale price index for updating the poverty line over time was later abandoned. This was done in accordance with the recommendations of the Study group on Estimation of Poverty Line constituted by the Planning Commission in 1984 which suggested the use of implicit private consumption deflator from National Accounts Statistics (of the Central Statistical Organization) for to update poverty line to be used henceforth. (The implicit private consumption deflator is obtained as the ratio of the private consumption expenditure at current and constant prices estimates for both of which are published by the NAS) In actuality, the Study Group had suggested the use of a price index approximately weighted by the consumption basket of the poor as an index for reflecting price changes relevant to the poor. This was observed to be very close to implicit private consumption deflator and hence its use. The average expenditure levels of each expenditure category in the Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 77

13 National Sample Survey data was adjusted by the ratio of private consumption expenditure in National Account Statistics to private consumption expenditure obtained from the National Sample Survey. The report of the Task Force has been thoroughly studied and criticized on a number of accounts by eminent scholars who have questioned the methods used by the committee. These include Minhas et al (1988, 89, 90), Expert Group on Estimation of Proportion and Number of Poor (Planning Commission, 1993), Dandekar (1996). These issues include the following - a. The use of a fixed basket of consumption goods across time has been questioned. The poverty line has been based on a calorie norm and the corresponding all-india consumption basket for the year The poverty line needs to be updated with the passage of time to take into account the changes which are likely to occur in price levels relevant to the consumption of the people around the poverty line and even preferences in choosing a basket of consumption goods. b. The use of the Wholesale Price Index (WPI) for updating the poverty line for subsequent years has also been debated. Whole sale prices are not reflective of the prices paid by individual purchasers who buy their requirements at retail prices and hence the Consumer Price Index seems to be a better alternative. Though a greater consensus is in favor of using an index that reflects the cost of living of the poor, construction of such an index is complicated and calls for exhaustive details about information on the consumption basket of the poor and the relevant prices. Nevertheless, use of WPI has found little favor among scholars of the subject. c. The use of a common basket of consumption goods in all regions or states has also found little acceptance. The Task Force s methodology to estimate State level poverty implicitly assumes that the age-sex and occupation distribution of population in the States follows the nationwide or all-india pattern and therefore the energy requirements are identical in all the States. The methodology also assumes that the prices and price movements of the consumption baskets are identical across various States. However, this is far from the reality. There are significant differences in the population and occupation structures across States which influence the energy or calorie requirements. Food habits are dependent on not only the local availabilities Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 78

14 but also the climatic conditions of the region. Also, cultural differences and therefore the food preferences are conspicuously distinctive in nature across Indian regions and States. Suggestions have been made regarding use of State specific poverty lines based on the State specific consumption patterns and State specific relevant prices and to conform to the standard of consistency, the national poverty line should be a weighted average of the States specific poverty lines. d. The procedure followed for adjustment to arrive at the estimates of poor population has also been criticized. The Task Force had recommended that the National Sample Survey (NSS) consumption distribution for all classes be scaled by a factor which is the ratio of private consumption expenditure as per National Account Statistics (NAS) of the CSO to private consumption expenditure obtained from the NSS. Uniform application of this factor is performed for all expenditure groups. Thus, poverty is then estimated from the adjusted distribution of population by expenditure classes. This adjustment was deemed necessary because of the difference in the NSS data and the NAS estimates for the aggregate private household consumption expenditure. The two data sets being critical components of the plan model were expected to be consistent with each other. In the difference between the NSS and NAS consumption data was 6% but it kept on increasing with time and more so when the base year revision was done by CSO. The factual difference between the two estimates is the result of several factors such as the sources and quality of data, extent of coverage, estimation methodology adopted, etc. As the gap between the two data sets widens, the adjustment factor also increases, and subsequently the poverty ratios exhibit an unrealistic decline. So, with the use of such an adjustment procedure, the prevalence of poverty changes every time the CSO makes revision. According to Expert Group (1993) the application of a uniform adjustment factor to raise the level of expenditure in all the expenditure classes is not justified in light of the fact that the difference in the two sets of estimates is much larger in respect of certain items than, in respect of others. The group though has counted the advantages of a commodity specific adjustment, the problems in respect of differences in coverage, classification, etc. between the NSS and the NAS series would not be overcome. Also, the Expert Group (1993) has suggested using NSS data completely if adjustment procedures are to be avoided. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 79

15 The Task Force (1979) estimated the poverty line at the national level for both rural and urban areas. However, the Planning Commission used the NSS consumption expenditure data for states pro-rata to NAS consumption expenditure across all expenditure classes to estimate poverty at the State level. This was not a commendable computation as noted by several scholars The Expert Group on Estimation of Proportion and Number of Poor (1993) (Popularly known as the Lakdawala Committee) With the view to consider afresh estimation of poverty, an Expert Group was constituted by the Planning Commission in September (Planning Commission, 1993) It was important to consider the issues involved in carrying forward the poverty lines set by the Task Force (1979), the contention regarding the price adjustments and incorporate the salient features of various studies that had delved into the subject of poverty estimation methodology. The terms of reference of the Expert Group were - To look into the methodology for estimation of poverty at national and state level and also to go into the question of re-defining poverty line, if necessary. The Group was formed under the chairmanship of Professor D T Lakdawala who passed away in April 1992 before the report was submitted in The team members included Prof V M Dandekar, Prof. P V Sukhatme, Dr. C Rangarajan, Prof. A Vidayanathan, Prof. R Radhakrishna, Dr. S Guhan, Prof, S D Tendulkar, Dr. S N Ray and Prof. S R Hashim. The report of the Expert Group explicitly expressed the importance of correct poverty estimation as it serves the crucial monitoring purpose - evaluating development efforts and fund allocation for poverty alleviation programmes. Though the Group considered alternative approaches to the measurement of poverty like the hunger criterion, the food share criterion, the calorie consumption criterion but decided against using these. It considered the calorie norm approach as a starting point for the poverty estimations (Rath, 2011) and suggested continuation of associating the poverty line with a fixed basket of consumption. The Expert Group appreciated the contentions on the issue that if calorie requirements form the core of poverty line specification, an allowance must be made for the Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 80

16 divergence found in inter-state or inter-regional calorie requirements due to differences in demographic and activity structures in addition to variations in preferences, climate and even prices at a point of time and over passage of time. But, due to lack of data, the Group ruled out the possibility of considering and using inter-state variations in normative calorie requirements. It was decided to apply the minimum living standard uniformly to the entire country. The commodity basket corresponding to this minimum living standard be decided at the national level and then applied to all States. This was proposed to ensure a cross-state comparison as well as an inter-temporal assessment. A highlight of the report by the Expert Group has been its take on the definition of poverty line and its interpretation. To quote from the report (GOI, 1993) - the fact that the poverty line is anchored in a norm of calorie requirement does not mean that those below the poverty line can be considered as malnourished or uniformly undernourished, everywhere or all the time. The poverty line defines on an average the level of per capita per day expenditure which meets a normative minimum standard of living, deemed reasonable. Calorie intake is but one of the ingredients, though an important one, of the minimum standard, but the poverty line makes an allowance for non-food consumption needs as well on the basis of observed consumer behavior. The Group recognizes the desirability of defining the normative standard for non-food consumption and its constituents without reference to actual behavior, but until this is done, the existing basis seems to be the most practical and reasonable. It is this consumption basket that constitutes the minimum standard for defining the boundary between the poor and the non-poor. In tune with this, the Expert Group categorized poor people as those who were unable to meet the average energy requirements and were therefore energy deficient relative to the average norms of the society. The committee recognized the incidence of malnourishment as a significant parameter for the well-being of the population but felt that its measurement, inter-region differences, preferences and energy intake and utilization would be involved in understanding the issue, which is marked by serious differences in opinion. They decided to keep away from the matter terming it not directly relevant to the measurement of poverty. Specifically, the recommendations of the Expert Group included the following- Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 81

17 a. The poverty line drawn by the Task Force (1979) as monthly per capita total expenditure of Rs and Rs for rural and urban areas respectively at prices is to be continued at the national level. (For convenience these were rounded off to Rs.49 and Rs.57) Rooted in the consumption norm of 2400 kilocalories per person per day and 2100 kilocalories per person per day for rural and urban areas respectively, these norms were to be adopted uniformly for all the states as well. b. Since a lot of study and related work had been done in defining poverty line with as the base year, its continuation was proposed for future estimation of poverty. c. The national level rural as well as urban poverty lines (as defined by the Task Force, 1979) were suggested to be disaggregated into State specific poverty lines. The Expert Group discussed the manner in which this was to be done. To begin with, the standardized consumption basket corresponding to the poverty line should be valued at the state specific prices prevailing in all the Indian states in the year , taken as the base year. In order to use the poverty line in a later year/ updating from the base year required a State specific price index which was calculated by considering a weighted average of the state specific food and non food (fuel and light; clothing and footwear; miscellaneous) price indices of the Consumer Price Index of Agricultural Laborers. The weights assigned were the respective weights of the commodities in the consumption basket of the poor at the national level. (Food items were given a weight of % and non-food items %) The Group explicitly discussed the implicit assumption for using this method as any consumer with income matching the poverty line will be capable of purchasing the standardized consumption basket which is same for all consumers and not vary with the passage of time. d. Using the updated State-specific poverty lines and the corresponding size distribution of per capita consumption expenditure of NSS, the incidence of poverty as the proportion of poor of the total population or the poverty ratio is to be thereafter computed for the rural and urban areas for every Indian state. Also, the number of poor in each State in rural and urban areas should be calculated by applying the poverty ratio to the population estimates as provided by the Registrar General of Census. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 82

18 e. The Expert Group suggested that for the rural areas it would be desirable to make use of the disaggregated commodity indices from Consumer Price Index for Agricultural laborers to update the rural poverty line and a simple average of weighted commodity indices of consumer price index for industrial workers and consumer price index of non-manual employees for updating the urban poverty line for later years. f. The national level poverty ratio should be derived as a ratio of the aggregate number of State-wise poor persons to the total population of the country - both rural and urban residents included. g. The Expert Group rejected the adjustment procedure recommended by the Task Force (1979) for maintain consistency between the aggregate consumption expenditure data sets obtained by the National Sample Survey and the National Account Statistics of the CSO. They termed the adjustment as arbitrary and recommended the use of the NSS data exclusively for estimating the rural and urban poverty rates and the same at the State level. The committee estimated the state specific poverty lines but not the national level poverty line. The all-india poverty lines for rural and urban areas were computed using the methodology suggested by the committee via interpolation of the expenditure distribution at the national level as given by the data sets of the National Sample Survey. Also, the national level poverty ratio was calculated using the average of the state specific poverty ratios. Summarizing the major changes suggested by the Expert Group (1993) a. Abandonment of NSS-NAS adjustment procedure. b. Computation of state specific poverty lines c. State specific price indices to be used for updating the poverty lines for later years. The Expert Group submitted its recommendations in July 1993 and these were deliberated upon for almost four years and in March 1997, the then Prime Minister accepted them with some modifications. Abandoning the Task Force s methodology and the adoption of the Expert Group s methodology led to a fall in the absolute poverty level for a year but when measured over time, no remarkable variation was observed in Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 83

19 the prevalence of poverty. Similar to the recommendations of the previous committee looking into the issue, (Task Force, 1979) the recommendations of the Expert Group were criticized on various accounts. Rath (2011) and Srinivasan (2007) have raised several questions on the approach adopted and the recommendation put forward by the Expert Group. These include the following a. Anchoring the base year poverty line to an average energy standard in terms of calorie requirement assumes a respect for the choice made by the consumer (given the budget constraint) and the fact that the chosen consumption bundle fetches the said average energy norm. Srinivasan (2007) has questioned the satisfaction of the second assumption in years other than the base year. This is logically discussed as the consumer may not choose the same consumption bundle as she did in the base year. Even if she does, it may not be available at the same price. Choosing a different bundle would definitely be linked to a different set of prices and also the calorie count may be different. b. The Expert Group (2009) has also questioned the use of the consumption patterns observed in to define the poverty line for rural and urban areas. These would have for sure become outdated. It is pointed that changes in the living standards, on account of economic growth and development, of the entire population including the lower strata have been observed. c. Srinivasan (1982, 1992, 2007) has shown absence of any scientific logic as per nutrition science regarding the attachment of calorie norms to officially computed poverty lines. This assertion finds foundation in the work and experiments by Prof. P V Sukhatme, who although was a member of the Expert Group (1993) but his objections (as mentioned in the supplementary note of the Expert Group s report) were not incorporated in the process of devising the methodology for estimation of all India and State level poverty lines. To quote Sukhatme from the note in the Expert Group s report - Man s capacity for work is therefore not determined by his intake but by efficiency with which he converts food energy into metabolizable energy over his homeostatic range of intake. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 84

20 d. The Expert Group s adoption of as the base year for their calculations has been termed unfortunate by Rath (2011) as the data coverage by National Sample Survey was incomplete for the said year. The consumer expenditure data was gathered for nine out of twelve months and the population sample considered during that year was relatively smaller. Thus choosing a data inadequate year was markedly puzzling. e. The methodology used for arriving at the State specific poverty lines has been one of disaggregation of the all-india data. However, since the state level data on consumption expenditures is available, this should have been the beginning and thereafter aggregation of the state level estimates should have been used to arrive at the national level estimates. The suggested direction by Rath (2011) is one of aggregation from the State level to the county level rather than the one adopted by the Expert Group. f. Rath (2011) has contended that although food baskets across states may be incomparable but the energy/calorie norms definitely are comparable. He has talked in favor of state level poverty line estimates using the same calorie norms as used at the all-india level. g. The price indices used (Consumer Price Index for Agricultural Labor for rural areas and Consumer Price Index for Industrial Workers for urban areas) for updating the poverty lines for later years were not necessarily the same prices as the ones at which the consumption expenditure data was collected by the National Sample Survey. Also, the weights allotted to the food and non-food components were not necessarily the same in the poverty line consumption baskets at national or state level. This has led to erroneous calculations of poverty lines at state levels for a number of years. (Rath, 2006, 2011) This has also been discussed by the Tendulkar Committee (GOI, 2009). h. Deaton and Tarozzi (1999) in their study have found out that that Consumer Price Index for Agricultural Labourers (that was used for the rural poverty line) had been growing too quickly and carrying forward the conclusion leads to believing that the decline in rural poverty rates had likely been understated by the official estimates of the Planning Commission. i. Srinivasan (2007) has highlighted incoherence in the existing poverty lines and has suggested broadening the concept of poverty by bringing in a new approach to anchor the poverty benchmarks to social norms. Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 85

21 Despite the criticism raised against the approach used by the Expert Group, the methodology suggested by it was used to generate the official poverty estimates Expert Group to Review the Methodology for Estimation of Poverty (2009) (Popularly known as the Tendulkar Committee) In light of the discontent with the methodology adopted to arrive at the official estimates of poverty and alternative views of many scholars and critics on the issue led the Planning Commission to set up an expert group under the chairmanship of Prof. Suresh D. Tendulkar to study and scrutinize the issue and suggest a new poverty line. In addition to the Chairman the Expert Group included Dr. R Radhakrishna, Prof. Raghav Ghaia and Dr. Suranjan Sengupta. The terms of reference for the group were - i. To examine the issues relating to the comparability of the NSS 50 th, 55 th and 61 st round, and to suggest methodologies for deriving such comparability with past and future surveys. ii. To review alternative conceptualizations of poverty and the associated technical aspects of procedures of measurement and data base for empirical estimation including procedures for updating over time and across states. iii. In the light of (ii) above, to recommend any changes in the existing procedures of official estimates of poverty. The expert group delved into all the issues in good detail and made some crucial recommendations. Though the committee recognized the multi dimensional nature of poverty, it decided to confine the study and definition of poverty in terms of private consumption. Discussing the scope of poverty, the committee acknowledged that poverty may possess material and non-material dimensions, but they would continue to view absolute consumption poverty in terms of a benchmark that conforms to a socially perceived normative minimal basket of basic human needs. As had been the practice, the use of private household consumer expenditure of Indian households as collected by the National Sample Survey Office for the purpose was to be continued. Data from the latest available consumption patterns in was applied for estimation of poverty. Although, the proposal has been to shift off the calorie norm, the proposed poverty lines have been validated by testing the adequacy of actual consumption expenditure per Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 86

22 capita near the poverty lines on food, education and health by comparing them with normative expenditures consistent with nutritional, educational and health outcomes. The Committee found that the actual private expenditures (reported by households) near the new poverty lines on these items were sufficient at the national level in both the rural and the urban areas and for most of the states. The Committee assessed the adequacy of consumption expenditure on education and health also. The Committee pointed out that the actual observed calorie intake of urban citizens near the poverty line as 1776 kilocalories per capita per day (as obtained from the NSS 61 st round), although the households were able to afford 2100 kcal per person per day. The actual intake closely matches the FAO (Food and Agricultural Organization) recommended standard of 1770 and interestingly, the actual observed intake of rural citizens around the new poverty line (1999 kcal per person per day) exceeds the FAO recommended intake. The new poverty line is near but less than the 1.25 US dollars per day (at 2005 PPP) poverty norm used by the World Bank. The Expert Group decided to adopt the MRP based estimates of consumption expenditure as the basis for future poverty lines as against the previous practice of using Uniform reference period estimates of consumption expenditure. (Till , data sets on consumption expenditure prepared by NSSO were collected by them on the basis of a uniform 30-day recall period for all items of consumption. But, onwards, the NSSO shifted to the use of a mixed recall period for collecting such information. Under the Mixed Recall Period (MRP), information on five broad item groups of household consumer expenditure with low frequency of purchase (low frequency items for short) namely, clothing, footwear, education, institutional medical care and durables is collected on a 365 day recall basis while information on consumption expenditure on all other items is collected on the basis of a 30 day recall period. Whereas, in the case of Uniform Reference Period (URP), all information on consumption expenditure is collected on a 30 day recall period basis. The price indices proposed by the Expert Group are based on the household level unit values obtained from the NSS 61 st round of consumption expenditure survey for food, Managing Poverty in India through State Interventions: Critical Assessment and Learning from Abroad 87