Poverty Trends and Measures

Abstract

It is undeniable that there has been a reduction in the proportion of India’s population in poverty over the last five decades during which it has been measured and tracked in the country. However, the high incidence of poverty, the large number of those who are poor, combined with the multiple deprivations that the poor experience, makes this the most important development challenge that faces us.

With necessary permissions, this chapter draws on Chap. 2 of the India Chronic Poverty Report (2011); Mehta and Bhide (2010); Mehta (2013); and Mehta (2017).

1 Introduction

It is undeniable that there has been a reduction in the proportion of India’s population in poverty over the last five decades during which it has been measured and tracked in the country. However, the high incidence of poverty, the large number of those who are poor, combined with the multiple deprivations that the poor experience, makes this the most important development challenge that faces us.

Research, carried out by the scholars and other professionals, has highlighted which segments of the population are more vulnerable to poverty – seen across social groups, geographical locations, gender or the age groups. Governments and others have attempted to address these issues over the years through a plethora of programmes and policies, moving from targeted and single objective approaches to more universal and ‘convergence-oriented’ approaches. The global efforts to reduce poverty have been articulated in the Millennium Development Goals and more recently the Sustainable Development Goals.

The persistence of poverty in certain specific socio-economic and regional groups of population has also meant that it is not only one generation of household members that suffer from poverty, but that children in these poor households may also grow up as poor. This intergenerational persistence of deprivations makes the goal of poverty reduction ever more imperative.

In this chapter, we provide a brief review of the estimates of poverty to highlight the scale and complexity of the challenge of poverty reduction before the country.

In Sect. 2 of this chapter, we discuss the trends in the poverty head count ratio as well as poverty lines and methods used to measure it. Section 3 draws attention to the fact that poverty is concentrated in certain parts of India, among particular social groups and among certain occupations. An understanding of the geographical, sociological and economic concentrations of poverty is important for an assessment of the type of policy approaches that can reach the poor. Section 4 of the chapter identifies the factors that explain the incidence of poverty. Section 5 outlines the importance of initial conditions in explaining the limited impact of growth on poverty reduction. Section 6 draws attention to the bands of vulnerability presented by the National Commission for Enterprises in the Unorganised Sector (NCEUS). Section 7 concludes the chapter.

2 Trends in Poverty: Incidence and Determinants

The head count ratio (HCR), or percentage of population below the poverty line, is an important indicator of the extent of poverty in a country. Before a problem can be addressed, the size of the problem must be known. The HCR is used to track progress in reducing poverty as well as design programmes to alleviate it. How many Indians are poor? What is the percentage of India’s population that is below the poverty line? The answer to these questions depends on how poverty is measured.

Traditionally, discussion regarding the extent of poverty and poverty trends in India is based on estimates of the head count ratio determined on the basis of household sample surveys conducted on a quinquennial basis by the National Sample Survey Organisation (NSSO). The official poverty estimates, patterns and trends in poverty are determined on the basis of analysis of data on household consumption expenditures on which poverty lines are juxtaposed to separate the poor from the non-poor and determine the extent of poverty. Detailed data on consumer expenditure from nationally representative samples are available from the early 1970s to the present at intervals of approximately 5 years.¹

Ten large sample consumer surveys have been conducted by the NSS on a quinquennial basis since 1973–1974. Poverty measured in terms of HCR registered a sustained decline over the 30-year period from 1973–74 to 2004–05, during which it halved from 54.9 to 27.5%. It declined further to 21.9% in 2011–12 (Table 1). However, it is important to note that the estimates of poverty presented in Table 1 are based on poverty lines computed by using three different methods. Estimates for 1973–74 to 2004–05 are based on the recommendations of the Lakdawala Committee (Planning Commission 1979); for 1993–94 to 2011–12 on the Tendulkar Committee (Planning Commission 2009); and for 2009–10 and 2011–12 on the Rangarajan Committee (Planning Commission 2014). These are discussed in Sect. 1 below.

Table 1 Trends in poverty

Whether 27.5% of India’s population was in poverty in 2004–05 or whether the estimate was 37.2% depends on the poverty line that is applied. If the poverty line is computed on the basis of the method suggested by the Lakdawala Committee, then 27.5% of the population was in poverty in 2004–05. If, however, the methodology suggested by the Tendulkar Committee is used to compute the poverty line, 37.2% Indians were living in poverty in this year. The basis for computing the poverty lines that lead to these different estimates of the percentage of the population that is in poverty is outlined in the section below.

2.1 The Poverty Line

Planning Commission constituted a Task Force to estimate poverty in 1977. On the basis of a systematic study of nutritional requirements, the Task Force submitted a report in which it recommended separate national-level poverty lines for rural and urban areas.

Based on observed consumer behaviour in 1973–74, it estimated that, on average, consumer expenditure of Rs 49.63 per capita per month was associated with a calorie intake of 2400 per capita per day in rural areas. Correspondingly, Rs 56.76 per capita per month was associated with a calorie intake of 2100 per capita per day in urban areas. The poverty line for subsequent years was estimated by adjusting the poverty line for the base year of 1973–74 for inflation (Planning Commission 1979).

In 1989, the Planning Commission constituted an Expert Group to review the methodology used for the assessment of poverty (Planning Commission 1993). It recommended:

• Continuation of the calorie-based consumption expenditure as a cut-off to determine the proportion of population below the poverty line;

• Disaggregation of national poverty lines into state poverty lines and then updating them using the Consumer Price Index (CPI) for industrial workers in urban areas and the CPI for agricultural labour in rural areas;

• Discontinuation of adjustment for the difference between NSS estimates of mean consumption expenditure and the National Accounts Statistics (NAS) estimate, because of increasing divergence between the two and because the adjustment increased the level of consumption expenditure for all households and decreased the estimated rate of poverty as compared with unadjusted data.

However, questions have remained regarding the adequacy of the poverty line that has been adopted. A large body of the literature questions the accuracy of official estimates of poverty on various counts: consumption patterns underlying the rural and urban poverty line basket (PLB) remain tied down to those observed in 1973–74; changes in the consumption pattern of the poor are not reflected in the poverty lines; use of the CPI for agricultural labourers understates the price rise for the rural population and hence the extent of rural poverty relative to urban poverty; the state is assumed to provide basic social services of health and education [although private expenditure on education and health was covered in the base year of 1973–74, no account has been taken either of the increase in the proportion of this in total expenditure over time or of its proper representation in available price indices (Planning Commission 2009)]. There are issues of consistency between the national accounts and sample survey data; deviation of the official poverty lines from their original definition based on minimum calorie norms; unrealistically large ratios of official rural–urban poverty lines in subsequent years compared with the initial 1973–74 ratio; lack of comparability of estimates from the 1999–00 survey of consumer expenditure and subsequent surveys due to changes in method, etc.²

Sen (2005) argued that these poverty lines are ‘not arbitrary figures, but have been derived from age-sex-occupation-specific nutritional norms by using the all-India demographic data from the 1971 Census…based explicitly on estimates of the normative nutritional requirement of the average person in the rural and urban areas of the country separately’.

Table 2 presents Planning Commission estimates of poverty lines separately for rural and urban areas over the period from 1973–74 to 2011–12. Poverty lines for the period 1973–74 to 1993–94 and 2004–05 are computed on the basis of the Lakdawala Committee method; for 2004–05, 2009–10 and 2011–12 on the basis of the Tendulkar Committee method; and for 2009–10 and 2011–12 also based on the Rangarajan Committee method. These poverty lines have been used to determine incidence of poverty at different points of time since 1973–74.

Table 2 Rural and urban poverty lines 1973–74 to 2011–12: consumption expenditure per capita per month (Rs) in current prices

Two fundamental problems have been raised in the literature with regard to the poverty lines. First is the lack of correspondence of poverty lines to consumption of 2400 kcal in rural and 2100 kcal in urban areas (Mehta and Venkatraman 2000; Sen 2005; Srinivasan 2007; Patnaik 2007, 2010). Second, while dietary requirements are calculated on a ‘scientific’ basis according to bodily needs, this is not applied to the non-food component of the poverty threshold so there is no guarantee of meeting basic non-food needs (Saith 2005). Serious concerns have been expressed especially regarding state budgetary allocations to and provisioning of health care (NRHM 2005). Ill health, and the need to spend large amounts on health care, exacerbates the suffering of those who are already poor and leads those who are not poor to poverty.³

Further, if nutrition is the underlying criterion of the poverty line, then it is shown from the same survey data which is used to define poverty line that even the population well above the poverty line may not be consuming the minimal requirement of calories.

In view of all of the above critiques and differences in views, the Planning Commission set up an Expert Group headed by Professor S. D. Tendulkar to re-examine the issue and suggest a new poverty line and poverty estimates. The Tendulkar Committee Report (Planning Commission 2009) recommended four major changes in the methodology for estimating poverty incidence:

• 1. A conscious move away from the calorie anchor while testing for the adequacy of actual food expenditure near the poverty line to ensure certain aggregate nutritional outcomes;

• 2. Using the same consumption basket for the rural poor as for the urban poor, but applying prices prevailing in rural areas to estimate the poverty line for rural areas. This exercise was done for each state, and estimates of the poor were then built to the national level for rural and urban areas;

• 3. A price adjustment procedure based predominantly on the same data set that underlies the poverty estimation and hence corrects for problems associated with externally generated and population segment-specific price indices with outdated price and weight bases used so far in official poverty estimations;

• 4. Explicit provision in price indices for private expenditure on health and education, which has been rising over time, and testing for their adequacy to ensure certain desirable educational and health outcomes.

The Lakdawala Committee-based poverty line was set at Rs 356.30 per capita per month for rural areas and Rs 538.60 per capita per month for urban areas for 2004–05. The Tendulkar Committee raised the poverty line for 2004–05 from:

• 1. Rs 356.30 per capita per month based on previous assessment to Rs 446.68 or by Rs 90 per capita per month (rural);

• 2. Rs 538.60 per capita per month to Rs 578.80 or by Rs 40 per capita per month (urban).

The HCR correspondingly increased from 27.5 to 37.2% in 2004–05. In other words, small increases in the poverty line led to a massive increase in the estimated population in poverty of almost 10 percentage points, with 407.2 million people living below this subsistence-level poverty threshold.

The Tendulkar Committee changed the basis for computing the poverty line. From the cut-off level of expenditure at which households are generally able to obtain food items that provide a certain number of calories of nutrition, the cut-off level of expenditure has now been delinked from the calorific intake of 2400 kcal in rural and 2100 kcal in urban areas. A miniscule amount has been added to accommodate consumption of certain basic services such as health, education and housing. The distinction between rural and urban requirements in computing the poverty line has also narrowed. There was dissatisfaction with the Tendulkar Committee methodology as well. Hence, the Rangarajan Committee was constituted in June 2012 and it redefined the poverty line.

Application of the Tendulkar Committee method to estimate poverty results in 29.8% of India’s population or 355 million people living in poverty in 2009–10. In comparison, estimation of poverty based on application of the Rangarajan Committee method results in 38.2% of the population or 455 million Indians being in poverty in 2009–10 and 30% of the population or 363 million poor in 2011–12. It can therefore be concluded that small increases in this subsistence-level poverty line lead to massive increases in the proportion of the population that is estimated to be in poverty.

The trends show that when the methodology for defining the poverty line changed, the poverty lines, as well as the difference between the rural–urban poverty lines, changed considerably. For instance, application of the methodology recommended by the Tendulkar Committee to the poverty line for 2004–05 led to a 25% increase in the poverty line for rural areas and a 7.5% increase in the poverty line for urban areas, relative to those defined as per the methodology proposed by Lakdawala Expert Group. The Rangarajan Expert Group’s approach led to an increase in the poverty line defined by Tendulkar Expert Group by 19% and 39.3% in 2009–10 for the rural and urban areas (Tables 3 and 4).

Table 3 Poverty lines adopted for measuring HCR in different periods

Table 4 Percentage change in the poverty line per year over the period from the previous survey year to the reporting year

As shown in Table 1, significant changes in poverty lines have led to a sharp increase in the estimated percentage of the population that is classified as poor. However, the reduction in HCR using different approaches to define poverty line has been of similar order of magnitude when such poverty lines are available for the same time periods. For instance, the Rangarajan Expert Group methodology indicates a reduction in HCR by 8.7 percentage points between 2004–05 and 2011–12. The Tendulkar approach indicates a decline in HCR by 7.9 percentage points. Between 1993–94 and 2004–05, using Lakdawala methodology, the incidence of poverty for the aggregate population fell by 8.5 percentage points and for the same period HCR fell by 8 percentage points using the Tendulkar methodology. Hasan et al. (2013, p. 2) point out that the poverty rates are measured by expenditure, and expenditures have gone up for every percentile of the population when adjusted for price changes over time and across space.

The decline in HCR between 2004 and 2011 reported by the World Bank (Table 5) is comparable to the decline seen in the estimates by the Tendulkar approach. However, the decline of almost 10 percentage points between 2009 and 2011 is greater than that based on the estimates by Tendulkar (7.9%) and Rangarajan (6.7%) approaches. Hasan et al. (2013, p. 3) note the same pattern.

A Task Force on the Elimination of Poverty in India was set up by NITI Aayog on 16 March 2015 under the Chairmanship of Dr. Arvind Panagariya, Vice Chairman, NITI Aayog. The Task Force prepared a paper titled ‘Eliminating Poverty: Creating Jobs and Strengthening Social Programs’ (NITI Aayog 2016). The Task Force concluded that tracking poverty over time and space was the principal objective behind the measurement of poverty. It suggested the consideration of four options for tracking extreme poverty:

• (i) Continue with the Tendulkar poverty line;

• (ii) Switch to the Rangarajan or other higher rural and urban poverty lines;

• (iii) Track progress over time of the bottom 30% of the population;

• (iv) Track progress along specific components of poverty such as nutrition, housing, drinking water, sanitation, electricity and connectivity.

However, a decision in this regard is yet to be taken. The latest official estimate of poverty that is available is for 2011–12 based on the Tendulkar method. Additionally, estimates have been provided by the Committee headed by Prof. Rangarajan.

2.2 Global Standards for Measuring Extreme Poverty

The World Bank estimates of poverty were initially set at PPP $1 a day in 1990 and were the average of the poverty lines of the poorest 15 countries. This became the ‘standard for measuring extreme poverty in the world’ and the basis of MDG 1 to halve poverty by 2015 (Ravallion et al. 2008).

This was revised or adjusted each time a new set of PPPs were produced by the International Comparison Program – PPP $1.08 in 1993, PPP $1.25 a day in 2005 and PPP $1.90 in 2015.

Ravallion (2010) responding to Deaton (2010) points out that in 2005, $1.25 a day was the average line of the poorest 15 countries. Further, $1.00 a day at 2005 prices was very close to India’s official poverty line and that ‘India’s official line is low by developing country standards’, and it is a ‘frugal line’.

The decline in the HCR based on this ‘frugal line’ that is representative of ‘extreme’ poverty for India is shown in Table 5. It is interesting to note that while the decline is 7.7 percentage points over 11 years between 1993 and 2004, it is 7.1 percentage points over 5 years between 2004 and 2009 and a massive 9.9 percentage points over only 2 years between 2009 and 2011.

Table 5 India poverty head count ratio at $1.90 a day (2011 PPP) (% of population)

Commenting on the PPP $1.90 per day that has been adopted by the SDGs Hickel (2015) points out that

$1.90 is not enough for basic human survival…. this amount of money is inadequate to achieve even the most basic nutrition. The US Department of Agriculture calculates that in 2011 the very minimum necessary to buy sufficient food was $5.04 per day. And that’s not taking account of other requirements for survival, such as shelter and clothing.

He further points out that

The World Bank picked the $1.90 line because it’s the average of the national poverty lines of the very poorest countries in the world, like Chad and Burundi. …The bank itself admits that poverty in Latin America, for example, should be measured at about $6 a day. And yet for some reason it persists with the $1.90 line.

If we want to stick with a single international line, we might use the “ethical poverty line” devised by Peter Edward of Newcastle University. He calculates that in order to achieve normal human life expectancy of just over 70 years, people need roughly 2.7–3.9 times the existing poverty line. In the past, that was $5 a day. Using the bank’s new calculations, it’s about $7.40 a day. As it happens, this number is close to the average of national poverty lines in the global south.

Yet the SDGs use the $1.90 line to measure poverty even though it is an ‘implausibly low threshold’, perhaps because a ‘more honest approach would force us to face up to the fact that the global economy simply is not working for the majority of humanity’.

The World Bank is now reporting poverty rates for all countries using two new international poverty lines: a lower middle-income international poverty line, set at $3.20/day, and an upper middle-income international poverty line, set at $5.50/day.

Regardless of what poverty line is used, it is clear that poverty remains a massive problem in the Indian context. If poverty lines are raised to realistic – instead of subsistence – levels, the percentage of poor will be much larger than the present estimates. A significant proportion of those who are above the poverty line are vulnerable to decline into poverty. The first step in tackling poverty then is that we acknowledge the extent of poverty and measure it. The potential for divergence of attention from addressing the poverty challenge because of the rise in the overall economic growth remains significant.

3 Social, Spatial and Occupational Concentration of Poverty

Estimates of incidence of poverty across social groups by the Planning Commission (2013a, b) also reflect large variations. The pattern of changes between 2004–05 and 2011–12 shows that the decline in poverty was faster in the case of Scheduled Castes than for Scheduled Tribes and Other Backward Castes. The pace of decline in HCR for all these three social groups exceeded that for the rural population as a whole and that for each of these groups in urban areas also. The incidence of poverty remains higher for ST, SC and OBC than for the ‘other category’ in both rural and urban areas in 2004–05 and 2011–12 (Table 6).

Table 6 Incidence of poverty across social groups: HCR

Findings of a panel study by Thorat et al. (2017) are presented in Table 7. These are for around the same period as the estimates provided in the Planning Commission report (Table 6). However, the reduction is somewhat greater for ST in this study and the incidence of poverty remains the highest for ST among the major social groups.

Table 7 Incidence of poverty across social groups from a panel survey: HCR

• Spatial Concentration of Poverty ⁴

Geographical factors are important, and the chronically poor are likely to be concentrated in the poorest states in India, which may also ironically be abundant in natural resources and in districts where multidimensional deprivation is significant. Poverty persists in almost all states. However, the proportion of the poor who suffer long-duration poverty and intergenerational transmission is likely to be significantly higher in those parts of the country that have consistently suffered greater incidence of severe poverty and multidimensional deprivation over many years. Poor states, in terms of per capita state domestic product, have in general remained poor compared with others – and the inequalities among the states have certainly not diminished over time (Bandyopadhyay 2001; Shepherd et al. 2004).

As can be seen from Table 8, states/UTs with 30% or more of their population in poverty in 2011–12 are Chhattisgarh, Dadra and Nagar Haveli, Jharkhand, Manipur, Arunachal Pradesh, Bihar, Orissa, Assam and Madhya Pradesh. Uttar Pradesh has 29.4% of its population in poverty.

Table 8 State population below the poverty line, 2011–12

The regional concentration of poverty has also been recognised in various policies since the early days of planning. For example, the Second Five-Year Plan (1956–1960) articulated balanced regional development as a key goal for India’s development effort (Bhide and Srinivasan 2004). Drèze and Srinivasan (1996), in World Bank (1997), illustrate variations in poverty incidence within a state but, even so, high levels of poverty have persisted in Bihar, Orissa, Madhya Pradesh and Uttar Pradesh.

More than one-third of India’s poor are located in Uttar Pradesh and Bihar (Table 9), in 2011–12. Just four states – Uttar Pradesh, Bihar, Madhya Pradesh and Maharashtra – account for more than half (52%) of India’s poor. Panda (2008) points to the contiguous nature of ‘high poverty states’.

Table 9 Spatial concentration of India’s poor

Those who suffer poverty are deprived not just in terms of calorific intake and income, but in many dimensions. Table 10 presents HDI estimates for India and states/union territories (UTs) for 2006.

Table 10 HDI and GDI scores for states/UTs, 2006

The scores on the Human Development Index and Gender Development Index also follow the pattern of the incidence of poverty. The lowest scores on HDI and GDI were achieved by Bihar, Uttar Pradesh, Madhya Pradesh, Orissa, Rajasthan, Chhattisgarh and Jharkhand. Most of these are states in which there is high income poverty.

On the basis of incidence of poverty and certain other development parameters, the Planning Commission set up an Expert Committee in 1997 to identify the 100 most backward and poorest districts in the country. The broad parameters adopted to analyse the causes of backwardness included indicators of deprivation (poverty ratio) and social and economic infrastructure. Social infrastructure in rural areas was viewed in terms of: (1) availability of safe drinking water; (2) basic health facilities; and (3) housing facilities.

The Committee found that Bihar and Jharkhand had more than a third, or 38, of the poorest districts; Madhya Pradesh and Chhattisgarh 19; Uttar Pradesh and Uttarakhand 17; Orissa 4; and Rajasthan 2. It is important to note, though, that some of these districts were also found in Maharashtra (10), West Bengal (4), Karnataka (1), Haryana (1), Himachal Pradesh (1), Dadra and Nagar Haveli (1) and Sikkim (2).

While spatial inequalities exist at all levels of disaggregation, the extent of these varies with choice of indicator and the geographical space over which comparisons are made. Multidimensional deprivation was estimated for about 379 districts in 15 large states of India based on data for the early 1990s, using variables for which data were available at the district level and that reflect long-duration deprivation (Mehta 2003; Mehta et al. 2004). For example, persistent spatial variations in the IMR could be considered to reflect persistent deprivation in the means of accessing health care. This could be a result of several factors, such as inability to obtain medical care because of lack of income; lack of available healthcare facilities in the vicinity; poor quality drinking water, resulting in waterborne diseases that cause mortality; lack of roads and public transport that enable quick transportation to hospitals in case of emergency; or all of the above. Similarly, illiteracy could be considered a persistent denial of access to information, knowledge and voice. Low levels of agricultural productivity may reflect a poor resource base; low yields owing to lack of access to irrigation and other inputs; poor quality of soil resulting from erosion; or lack of access to resources for investment because of lack of collateral or adverse climatic or market conditions. Poor quality of infrastructure reflects persistent denial of opportunities for income generation and growth.

While different lists of backward districts are available, Table 11 lists the 52 districts that suffer from the highest levels of persistent deprivation that are common to HDI and Augmented HDI (AHDI) methods of estimation. All the districts are from six of the seven states identified as having extensive rural poverty (Nath 2006).

Table 11 Fifty-two most deprived districts

Multiple deprivations that historically marginalised groups suffer make it harder for them to escape poverty, as different forms of poverty tend to be mutually reinforcing. Regions that are particularly likely to have large numbers of chronically poor people include tribal and forested (or degraded forest) regions, much of which are in the central and eastern ‘poverty heartlands’ and in semi-arid areas (Mehta and Shah 2001).

Bagchi (2018) mentions a recent study by Singh, Arora and Siddiqui⁵ in which they find a huge improvement in India’s multidimensional poverty between 2005–06 and 2015–16, due to the performance of southern states, i.e. Kerala, Tamil Nadu, Karnataka, Telangana and Andhra Pradesh. However, multidimensional poverty remains high in Bihar, Jharkhand, Uttar Pradesh, Rajasthan and Odisha (29%).

NITI Aayog has identified 115 aspirational districts based on 49 indicators across five sectors that include health and nutrition, education, agriculture and water resources, financial inclusion and skill development, and basic infrastructure. These districts are being encouraged to catch up with the best district in their state and aspire to become the best district in the country. States with the highest number of such districts are Jharkhand 19 districts, Bihar 13, Chhattisgarh 10 and Madhya Pradesh, Odisha and Uttar Pradesh 8 districts each (Table 12).

Table 12 State-wise distribution of selected aspirational districts

• The BPL Census

For implementation purposes, however, other measures, such as a census of household regarding the ownership of assets by the household or economic activities of household members, have been used to identify the poor. Hence, the Ministry of Rural Development (MoRD) conducted a Below Poverty Line (BPL) Census in 1992, 1997 and 2002 in association with states/union territories, to identify rural households that need assistance through various ministry programmes.

The MoRD BPL Census 2002 was based on an indicator-based scoring approach to classify households as poor and non-poor. The scorecard had 13 questions on various aspects, like size of landholding, type of house, availability of clothing, ownership of consumer durables, food security, access to sanitation, education attainment, migration. Each question had five scores, from zero to four, and the household was given a total score out of a maximum possible of 52. The BPL status of each household is on MoRD’s website. As was the case with methods used prior to the BPL Census 2002, the scorecard method used in this Census too was critiqued on a large number of grounds.

Vardhan (2010) re-administered the BPL scorecard to all households in two villages in February 2010 to determine the change in household status over time (Box 1)

Box 1: Below Poverty Line Census Results, 2002 and 2010

The MoRD BPL Census 2002 scorecard was re-administered to all households in two villages in February 2010. The data on all 13 dimensions were analysed for households in Juvvalapalem and Thippalakatta villages in Guntur district, Andhra Pradesh, for two points in time, 2002 and 2010. It was possible to identify houses that had exited poverty, those that had entered it and those that had persisted in it, and to analyse the factors leading to such movements. Using a cut-off score of 20 for declaring a household poor, 21% of households were found to be chronically poor. Some salient observations were:

• Many households around the poverty line had been vulnerable to shocks and influenced by enablers in moving above and below the line. Their entry or exit from poverty cannot be said to be relatively permanent. These are transient poor and vulnerable households, excluded from chronic poverty calculations.

• Children seem to be most impacted by the economic movements of a household. All households that had exited poverty showed an improvement in children’s access to education without them having to contribute to family income; the reverse was true for households that entered poverty.

• Sanitation and access to health care are important in entry, exit and persistence of poverty. Among households without access to sanitation facilities, poor households formed a disproportionately large group.

• A total of 59% of households that had exited poverty showed an increase in the score on migration. Linkages with the urban economy might be driving the escape from poverty in rural India.

• Other factors for exit from poverty are enablers (like access to credit, favourable agro-climatic conditions, alternative asset base) and more secure livelihoods (in terms of reduced market risks or more days of work).

• Low literacy/educational attainment is connected to persistence of poverty. A total of 89% of households that had remained poor showed no change in educational attainment status (qualification of the most literate adult).

• Other factors for persistence are unsecure livelihoods and poor asset base of households. This indicates that self- and wage employment programmes will help chronically poor households.

• Shocks related to health and agro-climatic conditions are the most common reasons for entry into poverty. Poor public healthcare delivery and inefficient implementation of women and child welfare programmes are detrimental.

• Being non-poor is associated with multidimensional wellness. Each parameter contributed almost equally between 5 and 10% to the total score. For poor households, the contributions of each of the parameters varied between 2 and 24%.

Source Vardhan (2010).

In May 2011, Union Cabinet approved BPL Census along with Caste Census for both rural and urban areas. The Saxena Committee (2011) and Hashim Committee (2012) reports provide details regarding the methodology suggested for identifying the poor for rural and urban areas, respectively. The methodology developed used a mechanism for automatic inclusion and automatic exclusion based on the criteria given below. The remaining households are graded to identify the poorest among them. SECC 2011 shows that three-fourths (74%) of the total households are located in rural areas.

• SECC 2011: Rural Areas

• Automatic exclusion: A household is automatically excluded from the BPL category if it has any of the following: motorised 2 or 3 or 4 – wheeler/fishing boat; mechanised 3 or 4 – wheeler agricultural equipment, Kisan Credit Card with a credit limit of over Rs 50,000; any household member is a government employee; non-agricultural enterprise registered with government; a member of the household earns above Rs 10,000 per month; income tax payee or a professional tax payee; house with three or more rooms with pucca walls and roof; ownership of a refrigerator, landline phone, more than 2.5 acres of irrigated land with 1 irrigation equipment, 5 acres or more of irrigated land for two or more crop seasons or 7.5 acres of land or more with at least one irrigation equipment.

• Automatic inclusion : A household is automatically included among the BPL category if it lacks shelter; is destitute or lives on alms; depends on manual scavenging; belongs to a Primitive Tribal Group; and is a legally released bonded labourer.

The remaining households are graded to determine the poorest on the basis of seven deprivation criteria. The deprivation criteria are households with only one room, kuccha walls and kuccha roof; no adult members between ages of 16 and 59; female-headed households with no adult male member between 16 and 59; households with a disabled member and no able-bodied member; SC/ST households; households with no literate adult above 25 years; and landless households deriving a major part of their income from manual casual labour. As many as 8.73 crore out of 17.97 crore rural households, or 48.58% households report at least one deprivation (Table 13).⁶

Table 13 Rural households suffering from each of the seven specific deprivations

Casual agricultural labour is the largest group that is stuck in poverty (Bhide and Mehta 2003; Mehta et al. 2011). This pattern has been corroborated by the data collected by the Socio-Economic Caste Census. Landless households dependent on manual casual labour constitute 30% of all households. This category of households additionally suffers from several of the seven deprivations mentioned above. These are ‘working poor’ for whom the state has not been able to meet its requirement to secure the right to an adequate means of livelihood.

The results of SECC 2011 are being used by the Ministry of Rural Development while implementing programmes in rural areas.

• SECC 2011: Urban Areas

The methodology suggested for identifying the poorest in urban areas is similar to that for rural areas (see Planning Commission 2012). The indicators suggested for determining exclusion, inclusion and scoring criteria are listed below.⁷

• Automatic exclusion: If the number of dwelling rooms exclusively in possession of the household is 4 or above (dwelling rooms as specified in the Report) that household will be excluded. Secondly, the household possessing any one of the assets, i.e. ‘4 wheeler motorised vehicle’, ‘AC set’ and ‘computer or laptop with Internet’, will also be excluded. Besides the households possessing any three of the following four assets, i.e. refrigerator, telephone (landline), washing machine, two wheeler motorised vehicle will also be excluded.

• Automatic inclusion : Households facing various kinds of deprivations and vulnerabilities, viz. residential, social and occupational vulnerabilities, would be automatically included in the BPL list.

• i. Residential vulnerability: If the household is ‘houseless’ as defined in the Report or the household has a house with roof and wall made of plastic/polythene or the household having only one room or less with the material of wall being grass, thatch, bamboo, mud, un-burnt brick or wood and the material of roof being grass, thatch, bamboo, wood or mud, then that will be automatically included.

• ii. Occupational vulnerability: The household having no income from any source; any household member (including children) engaged in a vulnerable occupation like beggar/rag picker, domestic worker (who are actually paid wages) and sweeper/sanitation worker/mali); and all earning adult members in a household are daily wagers or irregular wagers; then, that household should be automatically included.

• iii. Social vulnerability: If there is no member of the household aged 18 years and above (child-headed household) or there is no able-bodied person aged between 18 and 60 years in the household or all earning adult members in a household are either disabled, chronically ill or aged more than 65 years, then that household should be automatically included.

• Scoring index : In the third and final stage, the remaining households will be assigned scores from 0 to 12 based on various indicators of residential, social and occupational vulnerabilities. Those households with scores from 1 to 12 are to be considered eligible for inclusion in the BPL list in the increasing order of the intensity of their deprivations meaning thereby that those with higher scores are more deprived.

• International Measures of Multidimensional Deprivation

The Oxford Poverty and Human Development Initiative ‘counts the different types of deprivation that individuals experience at the same time, such as a lack of education or employment, or poor health or living standards’. Based on this, a multidimensional index of poverty (MPI) is constructed. The 2017 MPI shows that the per cent of India’s population in multidimensional poverty declined from 53.7% in 2005–06 to 41.3% in 2011–12 though the data are collected from different surveys. However, the per cent of India’s population in multidimensional poverty is far higher than that in Brazil (5.3% in 2014), China (4% in 2014), South Africa (9.2% in 2014–15) and Sri Lanka (5.3% in 2003) (Table 14).

Table 14 Multidimensional poverty index (MPI) and population in multidimensional poverty

4 Factors Affecting the Incidence of Poverty

As pointed out earlier, the regional and household social characteristics significantly differentiate the incidence of poverty. The persistence of this pattern reflects the vulnerability of these segments of the society to the poverty trap.

The results of the panel or longitudinal surveys of rural households reported in Bhide and Mehta (2011) and Dhamija and Bhide (2011) show that Scheduled Caste and Scheduled Tribe status, large household size and large proportion of children among household members were consistently associated with severe poverty (households with consumption expenditure 25% or more below the poverty line) in each of three rounds of the survey in 1970–71, 1981–82 and 1998–99 in rural India.

The findings of the above-mentioned studies show that access to land, livestock and village-level infrastructure and urban linkage (relatively large urban population of the district), irrigation and village size are statistically significant and have a negative impact on the incidence of poverty in two out of three rounds of the survey.

Using the NSS data for 2004–05 and 2011–12 and following the Tendulkar approach to the poverty line, Chatterjee et al. (2016) highlight the fact that measured by the traditional consumption expenditure norm to define poverty line, there was a sharp reduction in the percentage of poor in the population and a decline in the absolute number of poor between the two rounds by about 130 million. The paper also points out that using an internationally comparable poverty line of 1.9 PPP dollars (2011 base year) per capita per day, the data show a sharp decline in poverty during the period. Nevertheless, the paper notes that India continues to have a proportionately larger number of the world’s poor – 26% – and the largest number in absolute terms. Further, India’s success in reducing the proportion of poor by a consumption measure does not translate into improvements in the other dimensions of poverty – under-5 mortality and under nutrition. The under-5 mortality rate in India was higher than Nepal, Bangladesh and Vietnam. We may note that India remains among the low-ranked countries in a range of human development indicators. In the recent 2017 Global Hunger Index of IFPRI, India ranks among the lowest in Asia. The faster pace of poverty reduction is a significant achievement, in comparison with the past record but much remains to be done to improve the living conditions of the poor. Chatterjee et al. note that ‘many households that escaped poverty after 2005 still had consumption levels that were precariously close to the poverty line in 2012’.

Besides the vulnerability of social and regional groupings also noted above, the study by Chatterjee et al. (2016) indicates that:

• ‘In the case of households belonging to Scheduled Tribes and Scheduled Castes, the non-monetary deprivation of well-being such as health and education status was also greater than other caste groups.

• Poor in the low-income states, who were sizable in number, also were faced with the prospects of limited opportunities for income mobility as well.

• Growth of agriculture which was an important driver of poverty reduction is not any different from the growth impact of the other sectors on poverty reduction.

• Cities, more than the sectors, drive poverty reduction, but in this sense it is the non-farm sectors that provide higher income levels for the households than agriculture, which is the major source of livelihood for the rural areas.

• Jobs, more than income transfers, mattered for the households who escaped poverty’.

The analysis presented by Chatterjee et al. (2016) points to the key role urbanisation has played in reducing poverty during the period 2005–2012. Urban growth has provided increasing number of jobs, including for the poor. The sharp reduction in the incidence of poverty between 2005 and 2012 has been attributed primarily to (1) increase in wage rates for the unskilled workers, which emerged as demand for unskilled labour increased as construction activity in both rural and urban areas expanded (2) rise in agricultural commodity prices which in turn led to an increase in demand for workers in farming and (3) withdrawal of women from labour market for a variety of reasons.

Urbanisation and economic growth are seen to be the key drivers of poverty reduction during 2005–12. However, the study does caution on the continued vulnerabilities of a significant part of the population to any income shocks.

5 Growth and Incidence of Poverty

The limitations of the ‘trickle-down effects’ of the growth process translating into benefits for the poor have been widely recognised. Beginning from the seminal work of Ahluwalia (1978) in which he drew attention to the poverty-reducing effect of agricultural growth, further research by a number of scholars has pointed to the complex set of factors that affect poverty trends and patterns. Economic growth at a broad level, or even sectoral growth, while important, may not necessarily lead to poverty reduction. The work of Rao et al. (1986) and Sen (1996) highlights the many factors that influence the pattern of poverty and poverty reduction with implications for the linkages between growth and poverty reduction.

The impact of growth on poverty reduction, even in the case of agricultural growth, has been shown to be nuanced. Rao et al. (1986) draw attention to the positive effects of irrigation and rural electrification but note that the effect of roads and fertiliser use is not conclusive. Productivity of agricultural output per hectare is associated with lower HCR, but productivity per person is associated with higher HCR. Higher agricultural prices are seen to be associated with higher HCR. The vulnerability of agricultural labour to poverty has also been highlighted.

Examining the period 1960–61 to 1993–94, Sen (1996) finds that the initial conditions with respect to irrigation, female literacy and infant mortality are significant factors influencing poverty reduction. He also finds that the effect of relative prices (agriculture relative to overall) on poverty is more important than the overall inflation rate. Hasan et al. (2013, p. 12) also note that one of the explanations for lower impact of economic growth on poverty reduction witnessed in India when compared to other countries such as China, Vietnam and Indonesia is the set of initial conditions of human development. Additionally, the structure of economic growth in India has been such that employment growth has occurred in sectors where productivity growth rate is slower.

The need for more purposeful policies for poverty reduction, rather than relying merely on economic growth, has been articulated by a wide range of studies [Gaiha (1989, 1995), Jha (2000), Krishna et al. (2005), Bhalla and Hazell (2003)]. Hasan et al. (2013) point to the role of the structural changes that lead to poverty reduction. Faster output growth in sectors that employ more labour has a greater impact on poverty reduction than otherwise.

6 Bands of Vulnerability or Poverty

The National Commission for Enterprises in the Unorganised Sectors (NCEUS) presents the concept of a poverty band, so as to be able to explore the realities hidden behind the concept of a poverty line (Table 15). It estimates that 76.7% of the Indian population in 2004–05 lived on average per capita daily expenditure of just Rs 16, with the maximum expenditure just Rs 24. Moreover, of the total population, 36% were in the ‘vulnerable’ category. A single exogenous shock (such as death or disability of a breadwinner, serious sickness of a child or others or marriage expenses) could pull them back into the official ‘poverty’ group (Kannan 2010).

Table 15 Poverty status and poverty band, 2004–05

This picture, constructed on the basis of DPCE, reveals that every fifth Indian had only Rs 12 or less to spend each day in 2004–05. Further, three of four persons were in the poor and vulnerable categories in terms of daily consumer expenditure. High- and middle-income categories held 4 and 19.3% of the population, respectively. While estimates of this pattern are not available for more recent surveys, the estimated HCR from the 2011–12 expenditure survey remains at least 20% or 260 million population.

7 Conclusions

In this chapter, we have presented estimates of poverty over time and patterns of incidence of poverty across socio-economic groups of population and geographical or regional variations in the incidence of poverty.

While poverty rates have declined over the years, continued persistence of high rates, especially for some regions and social groups, is indicative of the intractability or stubbornness of poverty in such cases and the lack of responsiveness to overall or average economic growth and development.

The multidimensional nature of poverty has been highlighted in the studies which have examined the deprivations of many minimal sets of necessities experienced by the poor as also of those who may be slightly above the ‘poverty line’ that is adopted to estimate the size of the population which is poor. Deprivation of health, education and shelter has been both an effect and a cause of economic or consumption poverty.

The studies examining the correlates of poverty point to a number of household-level social and economic characteristics and a few features of the village or district that are closely associated with the economic status of the households in rural India. Continued prevalence of this pattern over time also suggests that the dynamics of economic changes has not been sufficiently in favour of the disadvantaged to change the structure of poverty. The interaction between economic growth at the macro-level and incidence of poverty has also been a subject of a large number of studies.

Many of those living in poverty today will remain poor over time and may pass their poverty to their children. This, combined with the size of the population which is poor, demands that we address the poverty challenge on priority. Accurately estimating the number of poor is important but is not enough. We also need to understand how many people are stuck in poverty and why? What are the factors that explain the persistence of poverty? How can these be addressed? How many people have moved out of poverty? What enabled them to move out of poverty? Did they manage to stay out of poverty? How many people who were not poor have become poor? What are the shocks that they suffered that led to their impoverishment? How can these be prevented from pushing people into poverty? These issues are discussed in the next chapter.