CURRENT ISSUE
Vol. 14, No. 2
JULY-DECEMBER, 2024
Research Articles
Tributes
Review Articles
Research Notes and Statistics
Book Review
Agrarian Novels Series
Poverty in India: The Rangarajan Method and the 2022–23 Household Consumption and Expenditure Survey
*Senior Research Assistant, Foundation for Agrarian Studies, sethuca7@gmail.com
†PhD Scholar, Centre for Development Studies
‡Office of the Vice Chairperson, Kerala State Planning Board
Abstract: This paper examines data on poverty in India from the most recent Household Consumption and Expenditure Survey, whose reference year is 2022–23. When these data were released by the Government of India, reports and studies stated that the data showed a substantial decline in poverty in India. We computed poverty levels by using the method proposed by the Expert Group to Review the Methodology for the Measurement of Poverty chaired by Dr. C. Rangarajan, which submitted its report in 2014. Our results showed significantly higher levels of poverty in 2022–23 than previously suggested.
Keywords: Household Consumption and Expenditure Survey, consumption poverty, Consumer Price Index, rural poverty, urban poverty, nutritional norms, inflation adjustment, head-count ratio, India
The Context
The release of the Household Consumption and Expenditure Survey 2022–23 (HCES 2022–23) data by the Government of India has led to fresh discussions on poverty in India. There has been no official estimate of consumption poverty in India for any year after 2011–12.
In 2012, the head-count ratio of poverty in India was estimated to be 21.9 per cent, by applying a method to estimate poverty proposed by the Expert Group to Review the Methodology for Estimation of Poverty chaired by Dr. Suresh Tendulkar in 2009 (hereafter Expert Group (2009)) to data from the Consumer Expenditure Survey 2011–12 (Press Information Bureau [PIB] 2013). When this computation of the poverty line came under criticism, the Planning Commission of the Government of India appointed an Expert Group to Review the Methodology for the Measurement of Poverty chaired by Dr. C. Rangarajan (hereafter Expert Group (2014)), to “revisit” the methodology for the measurement of poverty (Swaminathan 2010; PIB 2013).
In its report submitted in June 2014, the Expert Group (2014) proposed an alternative method of calculating the poverty line (the details of which are discussed later in this article), and estimated that the head-count ratio of poverty in India for 2011–12, using this method, was 29.5 per cent of the population of India. The Government of India did not notify its official acceptance of this estimate.
Data from Consumer Expenditure Surveys (CES) carried out by the National Sample Survey Office (NSSO), that formed the basis for various poverty estimation exercises, have not been available for over a decade. A Consumer Expenditure Survey was conducted in 2017–18, but the findings were not released citing “data quality” concerns (PIB 2019).
In this interim period, in the absence of official survey data, individual researchers’ estimates of the head-count ratio of poverty were largely, in the words of Himanshu (2022a), “shots in the dark.”
A “factsheet” from the HCES 2022–23 data was released by the Government of India in February 2024, two months before elections to the Indian Parliament. On the basis of this initial release, B. V. R. Subrahmanyam, the Chief Executive Officer of the Government of India’s central think tank, NITI Aayog, stated that the head-count ratio of poverty had come down to 5 per cent of the population (Dhoot 2024). Several others suggested that there had been a substantial decline in poverty (Anant 2024; Natti 2024; Perumal 2024; Rajora 2024). Their method was to use the Consumer Price Index to adjust a poverty line for 2011–12 for inflation and apply them to data from HCES 2022–23. Similarly, Rangarajan and Dev (2024) adjusted the Expert Group (2014) poverty line for 2011–12 using the Consumer Price Index and made a tentative estimate that 10.8 per cent of the population was below this poverty line in 2022–23.
Other scholars have questioned these claims, mainly on the grounds that the HCES 2022–23 survey method was not comparable with prior consumer expenditure survey rounds and that other evidence on the economy did not corroborate the assertion of a steep decline in poverty (Anand 2024; Himanshu 2024; Kishore and Jha 2024; Ghatak and Kumar 2024; Mehrotra and Kumar 2024).
This paper estimates a new poverty line from the HCES 2022–23 data by using the method proposed by Expert Group (2014) rather than by adjusting an earlier poverty line for inflation.
Given the differences between the survey methods followed in CES 2011–12 (the last available official consumer expenditure survey until now) and HCES 2022–23, there are problems of comparability of data between the two, and we have not attempted an intertemporal study. Nevertheless, there is merit in estimating poverty levels and a head-count ratio of poverty using HCES 2022–23 data, if only to evaluate the recent claims that this data reveals a very low head-count ratio of poverty in India.
Survey Differences and Choice of Method
The question of comparability between CES 2011–12 and HCES 2022–23 has been the subject of much discussion since the release of the new data. While both surveys aimed to capture the consumption pattern of a representative sample of the Indian population by canvassing information regarding the quantity and monetary value of expenditure for a list of items of consumption, we identified four major differences in survey method between the two rounds. First, there are differences in terms of the items for which data were collected, though most items remain the same. The HCES 2022–23 survey aggregated certain items such as different millets (ragi, jowar, bajra, among others) while it disaggregated others. Unlike CES 2011–12, the new HCES 2022–23 also collected information on items such as free rice and free sugar supplied through the Public Distribution System. Secondly, the questionnaire employed by HCES 2022–23 is more detailed and follows a different order from CES 2011–12, and uses the computer-assisted personal interviewing (CAPI) technique as opposed to the paper-based technique used in CES 2011–12. The third difference is in the number of visits per household. While CES 2011–12 had investigators visit each household once to collect all the data from that household, HCES 2022–23 involved three visits to each household to collect information regarding expenditure on food, “consumables,” and “durables” respectively. The fourth difference, which has received the most attention, is the change in sample design. HCES 2022–23 differs from CES 2011–12 in its sample design in two major ways. First, a portion of the rural sample is selected from villages within a 5 km distance from an urban area, and secondly, the selection criteria for the urban sample involves the ownership of non-commercial four-wheelers. Identifying the effects of these changes on the consumption expenditure data is beyond the scope of this paper. In our analysis, we have not reconciled these differences as we do not intend to compare the two rounds.
We have chosen the method proposed by Expert Group (2014) to estimate poverty from HCES 2022–23 data. The current official poverty line, proposed by Expert Group (2009), was calculated on the basis of data from CES collected with a “mixed reference period” (Tendulkar et al. 2009). However, HCES 2022–23 collected data using a “modified mixed reference period,” making the meaningful estimation of poverty using this method unfeasible (NSSO 2024). Consumption expenditure surveys typically ask respondents the quantity consumed and expenditure incurred for various items (such as milk, footwear, rent, travel, etc.) in the past n days. In this case, n is the reference period, which can typically be 7 days, 30 days, or 365 days. The Expert Group (2009) method used recall periods of either 365 days or 30 days, with a 365-day recall period for low-frequency items such as clothing, footwear, and educational expenses and a 30-day period for all other items. The HCES 2022–23 data contains a mix of three recall periods (for example, the recall period for milk consumption is 7 days, it is 30 days for cereals, and it is 365 days for most medical expenses). This makes it difficult to apply the Expert Group (2009) method to HCES 2022–23 data.
The Expert Group (2014), however, proposed a method to estimate poverty that used a “modified mixed reference period” (Rangarajan et al. 2014). While the poverty line derived by this method is higher than that derived via the method of the Expert Group (2009), it has its limitations (Deaton and Drèze 2014; Ramakumar 2014; Rangarajan and Dev 2015; Raveendran 2016). It has been argued (convincingly, we believe) that this method also tends to underestimate poverty.
The Expert Group (2014) method constructs a poverty line based on three components: expenditure on food, expenditure on essential non-food items, and other expenditures. The food component is based on nutritional norms. The essential non-food component is meant to be a normative measure, with the norm defined as being simply the median expenditure on these items (median expenditure could, of course, still be an insufficient level of expenditure from the point of view of need). It is also of concern that health expenses are not considered essential. The “other expenditures” component is tied to the food component. The method assumes that a person that has met their food requirement is ipso facto capable of meeting “other expenditures.” This, too, is an assumption that will not be valid for many persons.
Notwithstanding these criticisms, the Expert Group (2014) method can be considered the closest to an “official” method of calculating poverty from the new data. We use this method to calculate, in the following section, a poverty line and poverty estimates from the HCES 22–23 data.
Data, Method, and Results
We use the Household Consumption and Expenditure Survey 2022–23 (HCES 2022–23), the Periodic Labour Force Survey 2022–23 (PLFS 2022–23), and nutrition intake norms prescribed by the Indian Council of Medical Research – National Institute of Nutrition in 2020 (ICMR – NIN 2020) to generate a new poverty line using the Expert Group (2014) method. We repeat the same exercise with Consumer Expenditure Survey 2011–12 (CES 2011–12), Employment Unemployment Survey 2011–12 (EUS 2011–12), and older nutrition intake norms prescribed by ICMR – NIN (2010). The latter exercise is not so much for comparison but for assessing the deviation between our estimation and the original results of the Expert Group (2014).
The Expert Group (2014) was of the view that the consumption basket that defines the poverty line should include a food component, which addresses the question of adequate nourishment, a component that covers essential non-food items such as education, clothing, conveyance, and shelter, and a third component to address other “behaviourally determined” non-food expenditures. The method proposed by the Expert Group (2014) can be summarised as follows: first, average requirements for calories, proteins, and fats are calculated based on norms established by ICMR. These requirements are differentiated by age, gender, and activity levels for rural and urban populations to determine the normative levels of nourishment. Next, a food basket that meets these nutritional norms is defined by identifying the consumption levels of individuals within specific fractile classes. The average monthly per capita consumption expenditure (MPCE) on food for these classes is used to define the food component of the poverty line basket. Subsequently, the median expenditures on essential non-food items such as education, clothing, shelter, and conveyance are calculated. These values are treated as normative requirements for basic non-food expenses, and the expenditures by the median fractile class on these items are added to the poverty line basket. Finally, other non-food expenditures observed in the fractile classes meeting nutritional requirements are added. The sum of these three components is the new poverty line, expressed in terms of MPCE. This line is calculated separately for rural and urban areas. State-specific poverty lines are derived from these two lines using a relative Fisher Index, followed by the estimation of State-specific head-count ratios which are aggregated to arrive at the national head-count ratio for poverty (Rangarajan et al. 2014; Rangarajan and Dev 2015).
We deviate from the Expert Group (2014) method in three main ways. First, we rely on ICMR – NIN 2020 norms for nutrition intake instead of ICMR – NIN 2010 norms. Secondly, we use PLFS 2022–23 to estimate normative nutrition intake requirements, in place of a combination of Census 2011 and EUS 2011–12 used by the Expert Group (2014). This is because official age-wise population projections for rural and urban areas are not available, and a new census has not been carried out after 2011. Thirdly, we divided occupational groups into three activity levels – heavy, moderate, and sedentary – in the manner depicted in Table 1, following Alagh et al. (1979), as the exact method used by Expert Group (2014) for such a classification was not available.1
Activity Level | Occupational Sectors |
Heavy | Cultivation, Agricultural labour, Mining and quarrying, Construction |
Moderate | Livestock rearing, Forestry, Fishing, Hunting, Plantations and allied activities, Manufacturing and repairing |
Sedentary | Trade and commerce, Transport, Storage, Communication and other allied services |
Note: Non-workers are assigned the same nutritional requirements as those engaged in sedentary activity.
Source: Alagh et al. (1979, p. 6).
In order to estimate nutrient content in food items, we have used the nutrition chart prepared for CES 2011–12 by the Nutritional Intake in India Report 2011–12 (NSSO 2014).2 There are differences in food items collected in 2011–12 and 2022–23 as discussed earlier. We account for this by developing a concordance chart between items from the two periods.
The Indian Council of Medical Research prescribes normative requirements of calorie, protein, and fat for different age-sex-activity level combinations (ICMR–NIN 2020). This is given in Table 2. First, we estimated the proportion of population in these categories using PLFS 2022–23 (Table 3) and calculated the average per capita nutrition requirements. We arrived at 2,120 kcal per day, 42 gm of protein per day, and 22 gm of fat per day for rural areas; the corresponding figures for urban areas are 1,963 kcal, 45 gm of protein, and 21gm of fat per day (Table 4). Next, we divided the estimated distribution of population from HCES 22–23 into 20 fractile classes of MPCE, separately for rural and urban areas. We then calculated the average consumption of nutrition from food items for which data was captured by HCES 2022–23, for each fractile class. These values have been provided in Table 5. We aimed to find the fractile class for which the previously estimated nutrition levels are met, allowing for a 10 per cent leeway in line with Expert Group (2014) that argues such a variation will not affect nutrition adequacy.
Categories | Nutritional Norms | ||||
Age | Sex | Activity level | Energy (kcal/day) | Protein (gm/day) | Fat (gm/day) |
Less than 1 | 610 | 9.3 | 25 | ||
1–3 | 1010 | 11.3 | 25 | ||
4–6 | 1360 | 15.9 | 25 | ||
7–9 | 1700 | 23.3 | 30 | ||
10–12 | 2140 | 32.3 | 25 | ||
13–14 | Female | 2400 | 43.2 | 25 | |
13–14 | Male | 2860 | 44.9 | 25 | |
Adult | Female | Heavy | 2720 | 45.7 | 20 |
Adult | Female | Moderate | 2130 | 45.7 | 20 |
Adult | Female | Sedentary | 1660 | 45.7 | 20 |
Adult | Female | Non-Worker | 1660 | 45.7 | 20 |
Adult | Male | Heavy | 3470 | 54 | 20 |
Adult | Male | Moderate | 2710 | 54 | 20 |
Adult | Male | Sedentary | 2110 | 54 | 20 |
Adult | Male | Non-Worker | 2110 | 54 | 20 |
Elderly | Female | 1660 | 45.7 | 20 | |
Elderly | Male | 2110 | 54 | 20 |
Categories | Share of the age group in the population | |||
Age | Sex | Activity level | Rural | Urban |
Less than 1 | 1.09 | 0.88 | ||
1–3 | 4.85 | 3.77 | ||
4–6 | 7.14 | 4.07 | ||
7–9 | 5.65 | 4.06 | ||
10–12 | 6.31 | 4.92 | ||
13–14 | Female | 1.81 | 1.68 | |
13–14 | Male | 1.96 | 1.76 | |
Adult | Female | Heavy | 5.91 | 0.66 |
Adult | Female | Moderate | 2.63 | 1.98 |
Adult | Female | Sedentary | 1.39 | 4.94 |
Adult | Female | Non-Worker | 20.3 | 25.31 |
Adult | Male | Heavy | 14.45 | 4.21 |
Adult | Male | Moderate | 3.92 | 5.93 |
Adult | Male | Sedentary | 5.98 | 15.97 |
Adult | Male | Non-Worker | 6.35 | 8.35 |
Elderly | Female | 4.88 | 5.81 | |
Elderly | Male | 4.88 | 5.71 |
Source: Authors’ calculations based on NSSO (2023).
Nutritional Norm | Rural | Urban |
Energy requirement (kcal/day) | 2120 | 1963 |
Protein requirement (gm/day) | 42 | 45 |
Fat requirement (gm/day) | 22 | 21 |
90 per cent of energy requirement (kcal/day) | 1908 | 1767 |
90 per cent of protein requirement (gm/day) | 38 | 40 |
90 per cent of fat requirement (gm/day) | 20 | 19 |
Note: Expert Group (2014) argued that a deviation of 10 per cent will not affect nutrition adequacy and identified the section that met the lower bound of this range for poverty estimations.
Fractile Group of MPCE (in per cent) | Energy (kcal/day) | Protein (gm/day) | Fats (gm/day) | |||
Rural | Urban | Rural | Urban | Rural | Urban | |
0–5 | 1558 | 1601 | 41 | 44 | 30 | 38 |
5–10 | 1756 | 1761 | 47 | 49 | 37 | 45 |
10–15 | 1849 | 1854 | 50 | 51 | 41 | 50 |
15–20 | 1907 | 1907 | 52 | 53 | 44 | 52 |
20–25 | 1976 | 1961 | 53 | 54 | 47 | 55 |
25–30 | 2024 | 1999 | 55 | 55 | 48 | 58 |
30–35 | 2054 | 2024 | 56 | 56 | 50 | 59 |
35–40 | 2109 | 2079 | 58 | 58 | 52 | 62 |
40–45 | 2134 | 2120 | 58 | 59 | 54 | 63 |
45–50 | 2180 | 2163 | 60 | 60 | 56 | 66 |
50–55 | 2218 | 2188 | 61 | 61 | 57 | 68 |
55–60 | 2247 | 2246 | 62 | 62 | 59 | 70 |
60–65 | 2311 | 2267 | 64 | 63 | 62 | 71 |
65–70 | 2330 | 2349 | 64 | 65 | 63 | 75 |
70–75 | 2386 | 2384 | 66 | 66 | 65 | 76 |
75–80 | 2444 | 2450 | 68 | 68 | 68 | 80 |
80–85 | 2491 | 2541 | 69 | 70 | 71 | 84 |
85–90 | 2568 | 2675 | 71 | 74 | 74 | 89 |
90–95 | 2726 | 2828 | 76 | 78 | 81 | 95 |
95–100 | 3095 | 3488 | 86 | 93 | 97 | 115 |
Note: MPCE stands for Monthly Per Capita Expenditure.
Source: Authors’ calculations based on NSSO (2024).
We then estimated the average per capita expenditure on food items, essential non-food items (namely education, clothing, shelter,3 and conveyance), and other non-food items for each fractile class. This is shown in Table 6. The Expert Group (2014) method defines the poverty line as the sum of expenditure on essential non-food items of the median (45–50th) fractile, the expenditure on food items by the fractile that meets the nutrition norms, and the expenditure on other non-food items by the same fractile that meets the nutrition norms. This line is calculated separately for rural and urban areas.
Fractile Group of MPCE (in per cent) | Food | Essential Non-Food | Other Non-Food | |||
Rural | Urban | Rural | Urban | Rural | Urban | |
0–5 | 744 | 1023 | 178 | 300 | 450 | 678 |
5–10 | 956 | 1288 | 241 | 431 | 585 | 888 |
10–15 | 1073 | 1450 | 280 | 520 | 666 | 1022 |
15–20 | 1166 | 1589 | 311 | 621 | 727 | 1112 |
20–25 | 1248 | 1698 | 338 | 686 | 787 | 1229 |
25–30 | 1330 | 1799 | 364 | 773 | 840 | 1338 |
30–35 | 1404 | 1905 | 391 | 852 | 895 | 1442 |
35–40 | 1466 | 2015 | 422 | 949 | 959 | 1532 |
40–45 | 1534 | 2130 | 458 | 1050 | 1017 | 1622 |
45–50 | 1610 | 2235 | 480 | 1167 | 1087 | 1723 |
50–55 | 1675 | 2334 | 519 | 1284 | 1162 | 1851 |
55–60 | 1751 | 2458 | 565 | 1394 | 1237 | 2003 |
60–65 | 1847 | 2559 | 603 | 1559 | 1316 | 2160 |
65–70 | 1917 | 2726 | 667 | 1747 | 1423 | 2296 |
70–75 | 2026 | 2876 | 723 | 1947 | 1540 | 2507 |
75–80 | 2143 | 3078 | 802 | 2176 | 1683 | 2764 |
80–85 | 2276 | 3310 | 906 | 2552 | 1874 | 3066 |
85–90 | 2459 | 3686 | 1044 | 2925 | 2153 | 3625 |
90–95 | 2757 | 4211 | 1277 | 3675 | 2605 | 4513 |
95–100 | 3617 | 6226 | 2166 | 6430 | 4719 | 8166 |
Note: Essential non-food comprises education, clothing, shelter, and conveyance expenses.
Source: Authors’ calculations based on NSSO (2024).
Results
Our first result is the construction of two new poverty lines for 2022–23: Rs 2,515 per capita per month for rural areas and Rs 3,639 for urban areas. The lower bounds of nutrition norms were met by the fifth fractile class in rural areas and the third fractile class in urban areas.
Next, State-specific poverty lines were derived based on a relative Fisher Index for each State. We have used the method for calculating the Fisher Index provided by Expert Group (2009), which was also the method adopted by the Expert Group (2014). Based on the Fisher Index for each State, the all-India poverty line was adjusted to define State-specific poverty lines. The set of State-specific poverty lines for 2022–23 is shown in
Finally, we estimate the all-India head-count ratio of poverty as the weighted sum of State-specific head-count ratios based on State-specific poverty lines. This is done separately for rural and urban areas. We estimate a rural head-count ratio of 27.4 per cent, an urban head-count ratio of 23.7 per cent, and an overall head-count ratio of 26.4 per cent.
We repeated the same exercise using the Consumer Expenditure Survey 2011–12, Employment Unemployment Survey 2011–12, and ICMR – NIN 2010 norms to assess the deviation between our approach and that of the Expert Group (2014). The average per capita nutritional requirement norms we arrived at deviated slightly from those of the Expert Group (2014).4 The expenditures for the three components for each fractile class align exactly with the findings of the Expert Group (2014), allowing us to derive the same poverty line if we used the fractile classes that they identified. However, there was a difference in the calculated nutritional intakes for the fractile classes. While we were able to match our figures with those provided in the NSSO report titled Nutritional Intake in India, 2011–12, based on the same data, these values deviated from the calculations made by the Expert Group (2014). This suggests that the method of calculating nutritional intake employed by the Expert Group (2014) deviates from the method employed by NSSO. As a result, our analysis indicates that the urban nutritional norms are met by a higher fractile class than what was identified by the Expert Group (2014). This resulted in a higher poverty line for urban areas, and in turn, a higher HCR of poverty in 2011–12 as well. The corresponding tables have been provided as
Source of Estimation | Poverty Line | Head-Count Ratio | |||
Rural | Urban | Rural | Urban | Overall | |
Expert Group estimates, 2011–12 | 972 | 1407 | 30.9 | 26.4 | 29.5 |
Authors’ estimates, 2011–12 | 972 | 1502 | 31.3 | 30.8 | 31.2 |
Inflation-adjusted estimates, 2022–23 | 1837 | 2603 | 12.3 | 8.0 | 10.8 |
Authors’ estimates, 2022–23 | 2515 | 3639 | 27.4 | 23.7 | 26.4 |
Source: Expert Group (2014) and inflation-adjusted estimates from Rangarajan and Dev (2024).
Although our method largely followed the original method of the Expert Group (2014), we computed a head-count ratio of 31.2 per cent in 2011–12, as against the 29.5 per cent estimated by the Expert Group (2014).
Discussion and Concluding Remarks
This paper applies the method proposed by the Expert Group to Review the Methodology for the Measurement of Poverty, chaired by Dr. C. Rangarajan (Expert Group 2014), to data from Household Consumption and Expenditure Survey 2022–23 (HCES 22–23) in order to estimate a poverty line and the head-count ratio of poverty from these data.
Our results indicate that more than a quarter of all households in India have a monthly per capita expenditure that is below the poverty line in 2022–23. The head-count ratio of rural poverty (27.4 per cent) is higher than the head-count ratio of urban poverty (23.7 per cent).
Further enquiry into the reasons for high poverty levels in 2022–23 is the subject of our current research and will be dealt with in a subsequent paper. For the present, we note that the per capita energy consumption across quartiles of monthly per capita expenditure stagnated between CES 2011–12 and HCES 2022–23, and, in fact, declined by 2.6 per cent for the poorest quartile in rural India.5
The method of adjusting a prior poverty line using Consumer Price Index is inaccurate for at least two major reasons. First, the Consumer Price Index is calculated using outdated base weights for items in the consumption baskets. In the absence of new consumption expenditure data, these weights have not been updated for more than a decade. The weights assigned to items in the basket represent the estimated consumption pattern, which is likely to change over such a long period of time (Ramakumar 2014). A second and more important reason is that the Consumer Price Index, as apparent here, is not an instrument with which to track poverty. The consumption pattern of and prices experienced by the people below the poverty line differ from the consumption pattern of and prices experienced by the people above the poverty line.
Thus, our estimates are higher than the provisional head-count ratio of 10.8 for 2022–23 reported by Rangarajan and Dev (2024), derived by adjusting the 2011–12 poverty line of the Expert Group (2014) using the Consumer Price Index. It is also higher than estimates of poverty head-count ratios in other reports (Dhoot 2024; Perumal 2024; Natti 2024; Rajora 2024). These were also obtained by adjusting the official poverty line (taken from the report of the Expert Group (2009) chaired by Dr. Suresh Tendulkar) to current data.
Our results are also to be read in the context of evidence from research on rural wages, incomes of agricultural households, and the informal sector, which suggest that there has not been a substantial growth in incomes for the rural poor. Data from the Situation Assessment Surveys of Agricultural Households of 2012–13 and 2018–19 suggest that the average monthly incomes for agricultural households grew at 2.44 per cent per annum between these years, from Rs 8,843 to Rs 10,218 at constant prices (Bakshi 2021). Analysing data from two sources of wage rates from the Government of India – the Wage Rates in Rural India and the Periodic Labour Force Surveys – Das and Usami (2023) find that real wage rates in India stagnated between 2014–15 and 2022–23. Analysis of data from NSSO’s new Annual Survey of Unincorporated Sector Enterprises indicates a struggling informal sector with declining number of enterprises and stagnating wages (Das and Drèze 2024; Mohanan and Kundu 2024). Additionally, wages in the lower rung of the formal economy, such as daily earnings of factory floor workers, are observed by Singh (2024) to have grown only by 0.6 per cent per annum between 2002–03 and 2021–22 at constant prices, based on various rounds of the Annual Survey of Industries.
Our calculations show that more than a quarter of India’s population falls below the poverty line constructed using the method of the Expert Group (2014). We note that the method that we use is one that is likely to underestimate poverty rather than overestimate it (Ramakumar 2014). Consumption poverty remains an urgent and important problem in India.
Acknowledgements: We thank Sai Chandan Kotu and Arindam Das for participating in discussions on the paper with us and for verifying our calculations. We thank Madhura Swaminathan and an anonymous referee for their suggestions.
Notes
1 This has also been observed by Raveendran (2016).
2 This chart is prepared based on Indian Food Consumption Tables published by the Indian Council of Medical Research – National Institute of Nutrition. Five major Food Consumption Tables have been used in India; these were published in 1937, 1951, 1971, 1989, and 2017 respectively. The data in the chart we use are based on the 1989 tables, but remains the most recent one available.
4 The Expert Group (2014) used a combination of Census 2011 and EUS 2011–12 data for this. We have used only EUS 2011–12 to make it consistent with our approach for 2022–23.
5
References
Alagh, Y. K., Coondoo, D., Gupta, D. B., Iyengar, N. S., Jain, L. R., Murty, G. V. S. N., Radhakrishna, R., Tendulkar, S. D., and Majumdar, K. C. (1979), Report of the Task Force on Projections of Minimum Needs and Effective Consumption Demands, Government of India Planning Commission, available at https://dspace.gipe.ac.in/xmlui/bitstream/handle/10973/51430/GIPE-186074-Contents.pdf, viewed on September 10, 2024. | |
Anand, I. (2024), “What Does the Data from the Household Consumer Expenditure Survey 2022–23 Tell Us?” The India Forum, July 9, available at https://www.theindiaforum.in/public-policy/household-consumption-expenditure-survey-2022-23, viewed on July 10, 2024. | |
Anant, T. C. A. (2024), “India’s Poverty Debate Needs to Move On: Let’s Adopt New Norms, Mint, May 15, available at https://www.livemint.com/opinion/online-views/indias-poverty-debate-needs-to-move-on-let-s-adopt-new-norms-11715704815771.html, viewed on September 10, 2024. | |
Bakshi, A. (2021), “Situation Assessment Survey of Agricultural Households 2019: A Statistical Note,” Review of Agrarian Studies, vol. 11, no 2, pp. 119–28, available at https://doi.org/10.25003/RAS.11.02.0009, viewed on September 10, 2024. | |
Das, A., and Drèze, J. (2024), The Problem of India’s Stagnant Real Wages, Ideas For India, July 26, available at http://www.ideasforindia.in/topics/poverty-inequality/the-problem-of-india-s-stagnant-real-wages.html, viewed on September 10, 2024. | |
Das, A., and Usami, Y. (2023), “Downturn in Wages in Rural India,” Review of Agrarian Studies, vol. 13, no. 2, pp. 4–28, available at https://doi.org/10.25003/RAS.13.02.0002, viewed on September 10, 2024. | |
Deaton, A., and Drèze, J. (2014), Squaring the Poverty Circle, Ideas For India, July 30, available at http://www.ideasforindia.in/topics/poverty-inequality/squaring-the-poverty-circle.html, viewed on September 10, 2024. | |
Dhoot, V. (2024), “Poverty Levels Below 5%, Claims NITI Aayog Chief,” The Hindu, February 25, available at https://www.thehindu.com/news/national/poverty-levels-now-below-5-destitution-almost-extinct-niti-aayog-ceo/article67885895.ece, viewed on September 10, 2024. | |
Ghatak, M., and Kumar, R. (2024), “Determining How Many Indians Are Poor Today,” Ideas for India, May 29, available at https://www.ideasforindia.in/topics/poverty-inequality/determining-how-many-indians-are-poor-today.html, viewed on September 10, 2024. | |
National Sample Survey Office (NSSO) (2013a), Employment and Unemployment, July 2011-June 2012, NSS 68th Round (National Sample Survey DDI-IND-MOSPI-NSSO-68-10-2013), Ministry of Statistics and Programme Implementation, Government of India, available at https://microdata.gov.in/nada43/index.php/catalog/127/, viewed on September 10, 2024. | |
Natti, S. (2024), “Has Poverty Level Fallen Below 5%?”, The New Indian Express, February 27, available at https://www.newindianexpress.com/business/2024/Feb/27/has-poverty-level-fallen-below-5, viewed on September 10, 2024. | |
NSSO (2013b), Household Consumer Expenditure, NSS 68th Round Sch1.0 Type 2: July 2011–June 2012 (National Sample Survey DDI-IND-MOSPI-NSSO-68Rnd-Sch1.0-July2011-June2012), Ministry of Statistics and Programme Implementation, Government of India, available at https://microdata.gov.in/nada43/index.php/catalog/1, viewed on September 10, 2024. | |
NSSO (2014), Nutrition Intake in India, 2011–12 (560(68/1.0/3)), Planning Commission, Government of India. | |
NSSO (2023), Periodic Labour Force Survey (PLFS), July 2022-June 2023 (National Sample Survey DDI-IND-CSO-PLFS-2022-23), Ministry of Statistics and Programme Implementation, Government of India, available at https://microdata.gov.in/nada43/index.php/catalog/179, viewed on September 10, 2024. | |
NSSO (2024), Survey on Household Consumption Expenditure: 2022-23 (591), Ministry of Statistics and Programme Implementation, Government of India. | |
Press Information Bureau (PIB) (2013), Poverty Estimates, 2011–12, press note, Government of India, available at https://www.niti.gov.in/sites/default/files/2020-05/press-note-poverty-2011-12-23-08-16.pdf, viewed on September 10, 2024. | |
PIB (2019), Household Consumer Expenditure Survey, Nov 15, available at https://pib.gov.in/Pressreleaseshare.aspx?PRID=1591792, viewed on September 10, 2024. | |
Himanshu (2022), “Poverty Estimates Are a Shot in the Dark,” The Indian Express, Apr 12, available at https://indianexpress.com/article/opinion/columns/poverty-estimates-are-a-shot-in-the-dark-7865012/, viewed on September 10, 2024. | |
Himanshu (2024), “The Household Consumption Survey Results Raise Some Important Question,” Mint, Mar 22, available at https://www.livemint.com/opinion/online-views/the-household-consumption-survey-results-raise-some-important-question-11711031712486.html, viewed on September 10, 2024. | |
Indian Council for Medical Research – National Institute of Nutrition (ICMR – NIN) (2010), Dietary Guidelines for Indians—A Manual, Indian Council of Medical Research (ICMR), available at https://www.nin.res.in/downloads/DietaryGuidelinesforNINwebsite.pdf, viewed on September 10, 2024. | |
Indian Council for Medical Research – National Institute of Nutrition (ICMR – NIN) (2020), Revised Short Summary Report 2024, ICMR-NIN Expert Group on Nutrient Requirements for Indians, Recommended Dietary Allowances (RDA) and Estimated Average Requirements (EAR)—2020, ICMR, available at https://www.nin.res.in/RDA_short_Report_2024.html, viewed on September 10, 2024. | |
Kishore, R., and Jha, A. (2024), “First Consumption Data Released After 2011-2012,” Hindustan Times, Feb 25, available at https://www.hindustantimes.com/india-news/first-consumption-data-released-after-20112012-101708801139276.html, viewed on September 10, 2024. | |
Mehrotra, S., and Kumar, R. R. (2024), “Why the 2023 Consumption Survey Is Not Comparable with Previous Rounds,” The Wire, Feb 27, available at https://thewire.in/economy/why-the-2023-consumption-survey-is-not-comparable-with-previous-rounds, viewed on September 10, 2024. | |
Mohanan, P. C., and Kundu, A. (2024), “Growth or Decline? Understanding Non-agricultural Informal Enterprises,” Business Standard, available at https://www.business-standard.com/economy/news/growth-or-decline-understanding-non-agricultural-informal-enterprises-124071400346_1.html, viewed on September 10, 2024. | |
Perumal J. P. (2024), “Has Poverty Really Dropped to 5% in India?” The Hindu, Mar 14, available at https://www.thehindu.com/opinion/op-ed/has-poverty-really-dropped-to-5-in-india/article67950618.ece, viewed on September 10, 2024. | |
Rajora, S. (2024), “Economists Divided over Poverty Decline Claims by SBI, Niti Aayog,” Business Standard, Feb 28, available at https://www.business-standard.com/india-news/economists-divided-over-poverty-decline-claims-by-sbi-niti-aayog-124022801156_1.html, viewed on September 10, 2024. | |
Ramakumar, R. (2014), “On the Rangarajan Report on Poverty,” People’s Democracy, July 13, available at https://peoplesdemocracy.in/2014/0713_pd/rangarajan-report-poverty, viewed on September 10, 2024. | |
Rangarajan, C., and Dev, S. M. (2015), “Counting the Poor: Measurement and Other Issues,” Economic and Political Weekly, vol. 50, no. 2, pp. 70–4. | |
Rangarajan, C., and Dev, S. M. (2024), “With New Consumption Survey, the Need for New Indices,” The Indian Express, Mar 12, available at https://indianexpress.com/article/opinion/columns/moving-to-a-better-count-9208676/, viewed on September 10, 2024. | |
Rangarajan, C., Dev, S. M., Sundaram, K., Vyas, M., and Datta, K. L. (2014), Report of the Expert Group to Review the Methodology for Measurement of Poverty, Government of India Planning Commission, available at https://forms.iimk.ac.in/libportal/reports/232858161-Planning-Commission-report-on-poverty-estimates.pdf, viewed on September 10, 2024. | |
Raveendran, G. (2016), “A Review of Rangarajan Committee Report on Poverty Estimation,” Indian Journal of Human Development, vol. 10, no. 1, pp. 85–96, available at https://doi.org/10.1177/0973703016648033. | |
Singh, K. (2024), How Much Do India’s Factory Workers Earn on Average? CEDA, available at https://ceda.ashoka.edu.in/how-much-do-indias-factory-workers-earn-on-average/, viewed on September 10, 2024. | |
Swaminathan, M. (2010), “The New Poverty Line: A Methodology Deeply Flawed,” Indian Journal of Human Development, vol. 4, no. 1, pp. 121–25, available at https://doi.org/10.1177/0973703020100107. | |
Tendulkar, S. D., Radhakrishna, R., and Sengupta, S. (2009), Report of the Expert Group to Review the Methodology for Estimation of Poverty, Government of India Planning Commission. |
Appendix
State/Union Territory | Rural | Urban |
All India | 2515 | 3639 |
Andaman and Nicobar Islands (U. T.) | 3848 | 4992 |
Andhra Pradesh | 2609 | 3541 |
Arunachal Pradesh | 3247 | 4248 |
Assam | 2849 | 3933 |
Bihar | 2616 | 3539 |
Chandigarh (U. T.) | 2939 | 3999 |
Chhattisgarh | 2382 | 3312 |
Dadra and Nagar Haveli and Daman and Diu | 2603 | 3290 |
Delhi | 3181 | 3964 |
Goa | 3104 | 4100 |
Gujarat | 2654 | 3913 |
Haryana | 2799 | 3935 |
Himachal Pradesh | 2575 | 3681 |
Jammu and Kashmir | 2457 | 3397 |
Jharkhand | 2263 | 3391 |
Karnataka | 2627 | 3656 |
Kerala | 2694 | 3650 |
Ladakh (U. T.) | 2812 | 3910 |
Lakshadweep (U. T.) | 3107 | 4078 |
Madhya Pradesh | 2295 | 3425 |
Maharashtra | 2665 | 3932 |
Manipur | 3098 | 4067 |
Meghalaya | 2756 | 3807 |
Mizoram | 3225 | 4245 |
Nagaland | 2980 | 4238 |
Odisha | 2288 | 3324 |
Puducherry (U. T.) | 2962 | 3655 |
Punjab | 2763 | 3653 |
Rajasthan | 2614 | 3577 |
Sikkim | 3244 | 4506 |
Tamil Nadu | 2815 | 3759 |
Telangana | 2833 | 3823 |
Tripura | 2893 | 3923 |
Uttar Pradesh | 2443 | 3701 |
Uttarakhand | 2735 | 3707 |
West Bengal | 2511 | 3576 |
Source: Authors’ calculations from NSSO (2024).
Categories | Estimated Population Share | ICMR Nutritional Norms (2020) | |||||
Age | Sex | Activity level | Rural | Urban | Energy (kcal/day) | Protein (gm/day) | Fat (gm/day) |
Less than 1 | 1.3 | 1.1 | 585 | 10.2 | 19 | ||
1–3 | 5.49 | 4.59 | 1060 | 16.7 | 27 | ||
4–6 | 6.57 | 5.02 | 1350 | 20.1 | 25 | ||
7–9 | 6.29 | 5.19 | 1690 | 29.5 | 30 | ||
10–12 | 7.52 | 6.22 | 2100 | 40 | 35 | ||
13–14 | Female | 1.97 | 1.78 | 2330 | 51.9 | 40 | |
13–14 | Male | 2.31 | 2.07 | 2750 | 54.3 | 45 | |
Adult | Female | Heavy | 5.7 | 0.6 | 2850 | 55 | 30 |
Adult | Female | Moderate | 1.31 | 1.68 | 2230 | 55 | 25 |
Adult | Female | Sedentary | 22.06 | 26.08 | 1900 | 55 | 20 |
Adult | Female | Non-Worker | 0.84 | 3.43 | 1900 | 55 | 20 |
Adult | Male | Heavy | 16.77 | 4.08 | 3490 | 60 | 40 |
Adult | Male | Moderate | 2.97 | 6.5 | 2730 | 60 | 30 |
Adult | Male | Sedentary | 5.83 | 7.58 | 2320 | 60 | 25 |
Adult | Male | Non-Worker | 4.85 | 16.03 | 2320 | 60 | 25 |
Elderly | Female | 4.12 | 4.12 | 1900 | 55 | 20 | |
Elderly | Male | 4.11 | 3.93 | 2320 | 60 | 25 |
Source: Authors’ calculations based on NSSO (2023).
Nutritional Norm | Rural | Urban |
Energy requirement (kcal/day) | 2243 | 2092 |
Protein requirement (gm/day) | 49 | 51 |
Fat requirement (gm/day) | 28 | 26 |
90 per cent of energy requirement (kcal/day) | 2018 | 1883 |
90 per cent of protein requirement (gm/day) | 44 | 46 |
90 per cent of fat requirement (gm/day) | 26 | 23 |
Note: Expert Group (2014) argued that a deviation of 10 per cent will not affect nutrition adequacy and identified the section that met the lower bound of this range for poverty estimations.
Source: Appendix Table 2a.
Fractile Group of MPCE (in per cent) | Energy (kcal/day) | Protein (gm/day) | Fats (gm/day) | |||
Rural | Urban | Rural | Urban | Rural | Urban | |
0–5 | 1634 | 1638 | 43 | 44 | 21 | 27 |
5–10 | 1815 | 1756 | 48 | 48 | 26 | 34 |
10–15 | 1904 | 1838 | 51 | 50 | 29 | 38 |
15–20 | 1964 | 1872 | 52 | 51 | 31 | 41 |
20–25 | 1979 | 1915 | 53 | 53 | 33 | 43 |
25–30 | 2039 | 1969 | 55 | 54 | 35 | 46 |
30–35 | 2080 | 2033 | 56 | 55 | 37 | 49 |
35–40 | 2087 | 2050 | 56 | 57 | 39 | 51 |
40–45 | 2147 | 2104 | 58 | 57 | 41 | 54 |
45–50 | 2168 | 2130 | 59 | 58 | 43 | 55 |
50–55 | 2220 | 2167 | 60 | 59 | 45 | 57 |
55–60 | 2236 | 2231 | 61 | 61 | 46 | 60 |
60–65 | 2268 | 2244 | 62 | 62 | 49 | 63 |
65–70 | 2313 | 2286 | 63 | 63 | 51 | 64 |
70–75 | 2362 | 2389 | 64 | 65 | 53 | 68 |
75–80 | 2436 | 2431 | 67 | 67 | 56 | 72 |
80–85 | 2526 | 2491 | 70 | 68 | 60 | 72 |
85–90 | 2554 | 2586 | 71 | 71 | 63 | 79 |
90–95 | 2667 | 2805 | 74 | 77 | 69 | 87 |
95–100 | 3263 | 3190 | 91 | 86 | 92 | 100 |
Source: Authors’ calculations based on NSSO (2013b).
Fractile Group of MPCE (in per cent) | Food | Essential Non-Food | Other Non-Food | |||
Rural | Urban | Rural | Urban | Rural | Urban | |
0–5 | 316 | 415 | 54 | 83 | 152 | 203 |
5–10 | 401 | 533 | 71 | 112 | 195 | 264 |
10–15 | 452 | 600 | 80 | 144 | 218 | 312 |
15–20 | 493 | 656 | 88 | 181 | 235 | 344 |
20–25 | 516 | 713 | 99 | 204 | 261 | 383 |
25–30 | 554 | 769 | 102 | 242 | 277 | 415 |
30–35 | 586 | 822 | 111 | 271 | 293 | 465 |
35–40 | 612 | 889 | 120 | 310 | 314 | 492 |
40–45 | 640 | 919 | 132 | 364 | 333 | 538 |
45–50 | 678 | 977 | 141 | 407 | 347 | 571 |
50–55 | 710 | 1019 | 149 | 447 | 372 | 631 |
55–60 | 733 | 1096 | 166 | 507 | 402 | 661 |
60–65 | 775 | 1156 | 182 | 549 | 425 | 739 |
65–70 | 814 | 1210 | 195 | 634 | 461 | 808 |
70–75 | 861 | 1302 | 218 | 726 | 500 | 883 |
75–80 | 922 | 1384 | 241 | 831 | 550 | 1001 |
80–85 | 1002 | 1504 | 279 | 968 | 607 | 1129 |
85–90 | 1077 | 1650 | 340 | 1161 | 710 | 1375 |
90–95 | 1216 | 1946 | 442 | 1592 | 898 | 1812 |
95–100 | 1771 | 2859 | 834 | 3434 | 1877 | 3989 |
Source: Authors’ calculations based on NSSO (2013b).
State/Union Territory | Rural | Urban |
All India | 972 | 1502 |
Andaman and Nicobar Islands (U. T.) | 1229 | 1845 |
Andhra Pradesh | 1036 | 1475 |
Arunachal Pradesh | 1132 | 1517 |
Assam | 1023 | 1548 |
Bihar | 976 | 1340 |
Chandigarh (U. T.) | 1209 | 1577 |
Chhattisgarh | 897 | 1314 |
Dadra and Nagar Haveli | 1005 | 1669 |
Daman and Diu | 1206 | 1552 |
Delhi | 1353 | 1647 |
Goa | 1166 | 1560 |
Gujarat | 1134 | 1667 |
Haryana | 1137 | 1637 |
Himachal Pradesh | 985 | 1472 |
Jammu and Kashmir | 982 | 1432 |
Jharkhand | 940 | 1393 |
Karnataka | 921 | 1491 |
Kerala | 1031 | 1435 |
Lakshadweep (U. T.) | 1166 | 1467 |
Madhya Pradesh | 946 | 1455 |
Maharashtra | 1084 | 1707 |
Manipur | 1285 | 1732 |
Meghalaya | 1124 | 1624 |
Mizoram | 1159 | 1735 |
Nagaland | 1279 | 1768 |
Odisha | 878 | 1327 |
Puducherry (U. T.) | 1070 | 1421 |
Punjab | 1148 | 1604 |
Rajasthan | 1059 | 1532 |
Sikkim | 1090 | 1595 |
Tamil Nadu | 989 | 1391 |
Tripura | 912 | 1441 |
Uttar Pradesh | 918 | 1446 |
Uttarakhand | 984 | 1520 |
West Bengal | 971 | 1501 |
Source: Authors’ calculations based on NSSO (2013b).
Quartiles of MPCE (in per cent) | Rural | Urban | ||||
2011–12 | 2022–23 | % Change | 2011–12 | 2022–23 | % Change | |
0–25 | 1859 | 1811 | –2.6% | 1804 | 1821 | 1.0% |
25–50 | 2104 | 2105 | 0.0% | 2057 | 2085 | 1.4% |
50–75 | 2279 | 2305 | 1.1% | 2264 | 2300 | 1.6% |
75–100 | 2690 | 2676 | –0.5% | 2701 | 2823 | 4.5% |
Note: MPCE stands for Monthly Per Capita Expenditure.
Source: Authors’ calculations based on NSSO (2024) and NSSO (2013b).
Quartiles of MPCE (in per cent) | Rural | Urban | ||||
2011–12 | 2022–23 | % Change | 2011–12 | 2022–23 | % Change | |
0–25 | 290 | 913 | 214% | 446 | 1498 | 236% |
25–50 | 434 | 1383 | 219% | 815 | 2490 | 206% |
50–75 | 614 | 1951 | 218% | 1317 | 3749 | 185% |
75–100 | 1355 | 3846 | 184% | 3458 | 7978 | 131% |
Note: MPCE stands for Monthly Per Capita Expenditure.
Source: Authors’ calculations based on NSSO (2024) and NSSO (2013b).
Quartiles of MPCE (in per cent) | Rural | Urban | ||||
2011–12 | 2022–23 | % Change | 2011–12 | 2022–23 | % Change | |
0–25 | 0.015 | 0.022 | 45.5% | 0.020 | 0.026 | 29.9% |
25–50 | 0.019 | 0.025 | 32.7% | 0.025 | 0.030 | 18.8% |
50–75 | 0.021 | 0.027 | 24.2% | 0.028 | 0.032 | 14.3% |
75–100 | 0.025 | 0.029 | 15.5% | 0.030 | 0.033 | 9.0% |
Note: MPCE stands for Monthly Per Capita Expenditure.
Source: Authors’ calculations based on NSSO (2024) and NSSO (2013b).