ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846
-A A +A

Multidimensional Poverty in Tripura

An Inter-temporal Analysis

Niranjan Debnath (n2n.nath@gmail.com) is a research scholar and Salim Shah (salimshah.tu@gmail.com) is an assistant professor at the Department of Economics, Tripura University, Agartala.

The changes in multidimensional poverty are analysed for the states of the north-eastern region in general, and Tripura in particular, from 2000 to 2015. Special emphasis is laid on the household characteristics in Tripura and the rural–urban, gender, social, and religion subgroups. The value of the multidimensional poverty index fell by 4.6% during the study period and the proportion of the multidimensionally poor by 1.7% per annum. The reduction in the MPI and the incidence of poverty was significantly higher in the rural areas of Tripura considering the social and religious subgroups.

The measurement of poverty based on monetary attributes, such as consumption or income, has been criticised heavily. Critics claim that monetary poverty measures are not sufficient to explain the multifaceted nature of the deprivations of well-being. A realistic poverty measure should contain basic human needs, such as minimum levels of educational attainment, basic attainment of health, and living conditions (Tsui 2002). Poverty is characterised by multidimensional deprivations and is best explained by a multidimensional approach (Bourguignon and Chakravarty 2003). There has been a conceptual revolution in measuring poverty during the past few decades and a focus on developing a more comprehensive multidimensional framework.

Several methodologies have been developed for identifying the multidimensionally poor: counting approaches, multiple correspondence analysis, latent variable techniques, and fuzzy set theory. Researchers have proposed multidimensional poverty measures based on the capability approach with a normative framework. One of them is the Alkire and Foster (2007) methodology based on the counting approach (Atkinson 2003) along with the axiomatic approach following the Foster–Greer–Thorbecke (FGT) class of unidimensional measures. In this approach, the axioms of subgroup decomposability and dimensional breakdown allow for decomposing the aggregate value into population subgroups, such as social and religious groups, to identify the major drivers of poverty (Alkire and Foster 2011). The decomposability of the poverty measure appears to be an important component for the policymakers to frame effective multisectoral planning across states and districts.

Traditionally, poverty measurement in India has revolved around the ability to spend on goods and services that was measured from the consumption side (monetary attribute only) rather than by the capability to achieve well-being and functioning in society (Sen 1999). However, there have been methodological revisions and debates on the multifaceted nature of poverty and the need for inclusive growth (GoI 2009, 2014; Sen and Himanshu 2004; Deaton and Drèze 2009; Subramanian 2011; Ahluwalia 2011). Many empirical studies show that a significant percentage of the multidimensionally poor are not income-poor and that a significant percentage of the income-poor are not multidimensionally poor (Laderchi et al 2003; Wang et al 2016). The Government of India (GoI) moved from an income approach to a multidimensional approach for identifying poor families and improving the provision of government services in 2002 and began constructing a new “index of multiple deprivations” in 2008.

There are some empirical studies on multidimensional poverty measures in India, but these are confined mainly to the major states. Some recent studies demonstrate that poverty has fallen between the 1990s and early 2000s (Jayaraj and Subramanian 2010; Alkire and Seth 2008, 2013; Mishra and Ray 2013). By using the measure proposed by Chakravarty and D’Ambrosio (2006), Jayaraj and Subramanian (2010) and Mishra and Ray (2013) show that multidimensional poverty has not fallen uniformly across different population subgroups at the national level. Alkire and Seth (2013), find that population subgroups (state, caste, or religion), which were the poorest in 1999, could reduce poverty the least in the following seven-year period. Seth and Alkire (2014) emphasise on the importance of inequality measures to improve the Alkire–Foster class of poverty measures.

The two rounds of Demographic Health Surveys (DHS) with the representative years 1998–99 and 2005–06 have been used for an inter-temporal analysis of poverty in India. Seth and Alkire (2014) find that poverty in terms of multidimensional poverty index (MPI), as well as the intensity of poverty, fell between 1999 and 2006: the national MPI declined from 0.300 to 0.251 and the intensity of poverty from 52.9% to 51.7%. Although the MPI and its components have been reduced for each of the caste groups—Scheduled Castes (SCs), Scheduled Tribes (STs), Other Backward Classes (OBCs), and General—the inequality among the poor appears to be higher for the poorest subgroup, the STs. Inequality among the poor in India as a whole has fallen from 0.100 to 0.097, but no such reduction is noticed for most of the subgroups. The inequality across the poor has fallen within SCs and OBCs, but not within STs or the General category. Inequality has risen among the poor within the poorest subgroup (STs), though it has fallen intra-group overall and between groups.

Most studies of poverty and multidimensional poverty are descriptive in nature, and they focus mainly on a state as the unit of analysis (Nayyar 2005; Kurian 2000; Roy and Haldar 2008). But, the north-eastern region (NER) is diverse: ethnic groups are numerous; and languages, religions, forms of governance, topography, and natural climate are varied. The NER is also economically backward. Educational performance is poor, health service facilities inadequate, road connectivity between and within states insufficient, and basic amenities lacking. Despite the provision for the Non-Lapsable Central Pool of Resources and many other development initiatives, economic development is uneven throughout the region and within states, and variations are enormous among and within states in terms of the reduction in poverty and inequality and the growth rate of per capita net state domestic product (NSDP) (Shah and Debnath 2015; Debnath and Shah 2015; Dehury and Mohanty 2015; Alkire and Seth 2013). A unidimensional measure of poverty cannot capture all the dimensions of deprivations of the region—or any of its states, such as Tripura, one characterised by immense ethnic, cultural, administrative, and economic diversity—or the actual causes of backwardness; these can be adequately addressed only by multidimensional measurements of poverty based on the capability approach.

One recent comprehensive study (Alkire and Seth 2013) on a multidimensional approach to poverty measurement related to non-monetary variables examines the change in multidimensional poverty in India and its states, including the north-eastern states, for the period from 1999 to 2006 using the Alkire–Foster methodology. Based on the India Human Development Survey (IHDS) 2011–12, and using the household as the unit of analysis, Dehury and Mohanty (2015) estimate and decompose multidimensional poverty—considering health, education, standard of living, and the environment of households—for the regions of India, including the North East, for 2011–12. By adopting the Alkire–Foster methodology for eight indicators of the above dimensions, Dehury and Mohanty find that 43% of India’s populations are multidimensionally poor, and that the regional variations are large.

Following the Alkire–Foster poverty measure, and using stratified random sampling methods, Debnath and Shah (2015) estimate an MPI for the gram panchayats and village committees in the Kathalia rural development block (consisting of both gram panchayats under panchayati raj institutions and village committees under the Tripura Tribal Areas Autonomous District Council) in the Sepahijala district of Tripura through a primary survey of 300 households. They find diversity in multidimensional deprivation between the two administrative regions within the same block for 2014, where the habitats of village committees are more deprived than that of the gram panchayats. Shah and Debnath (2015) also cover the theoretical aspects and methodological issues in multidimensional poverty measurements across the social and religious groups in the context of Tripura. These studies constitute the primary source of inspiration for the present study to examine the changes in multidimensional non-monetary poverty at the household level for a longer period for Tripura.

This paper takes up a multidimensional poverty measurement of Tripura using non-monetary dimensions to track poverty levels and examine the variations across social, religion, rural–urban, and household characteristics. This inter-temporal study covers the period from 2000 to 2015 and is the first in measuring multidimensional poverty in the NER. It aims to estimate multidimensional poverty for Tripura at three points of time to find out the changes in the situation of poverty and to decompose the poverty estimated across the population subgroups to examine the change in their poverty over time.

Methodological Issues

This assessment of multidimensional poverty was based on the counting approach developed by Alkire and Foster (2007, 2011) owing to its advantages of dual cut-offs (deprivation cut-off and poverty cut-off), joint distribution, and decomposability. The present study concentrates on the general form of the social indicators with their ordinal meaning. The in-depth explanation of the methods is available in Alkire and Foster (2011). Since 2010, the United Nations Development Programme has been using the Alkire–Foster methodology to calculate the MPI for several countries.

Let us consider a group of individuals from which poor persons are to be identified using dual cut-off in terms of deprivation cut-off across indicators and poverty cut-off to identifying the poor in a multidimensional space of achievements. Prior to the application of these cut-offs, a set of 10 indicators (d) was identified such that each of the indicators is widely accepted as
essential for human well-being. These indicators were classified into three broad dimensions (T), which were equally weighted.

Similarly, the indicators were in turn to assign equal weights within each of the dimensions such that the weight attached to indicator j, with j = (1,2,3,….,d) is wj = (1/T) × (1/d). The first cut-off relates to the deprivation cut-offs for each of the 10 indicators as defined in Table 1. The cut-off point is a normative minimum level that an individual i requires to achieve to be defined as non-deprived.

Let Z be a row vector of indicator-specific deprivation cut-offs zj with j(= 1,2,3,...,d). An individual is defined as deprived if their achievement is less than the specified cut-off. For example, the ith individual, deprived in jth indicator be assigned the value xij = 1, otherwise xij = 0 in the achievement matrix. Let C be a column vector representing the weighted sum of the deprivation score of the ith individual. Let k (0 k ≤ 1) be the threshold (or poverty cut-offs) to identify an individual as poor. However, the value of k can be chosen based on previous studies, or what society would reasonably consider, or on the specific policy goals of a country or state.

In the present case, an individual is considered to be multidimensionally poor if the weighted sum of the deprivation score exceeds at least one-third, or ci ≥ k (= 1/3), i = (1,2,3..., n) as the poverty cut-offs defined in an international MPI. Now the multidimensional headcount ratio H (incidence of poverty) is defined as the ratio of total number of multidimensionally poor (qk) to the total population (n), or H = (qk/n) where qk= Ʃni=1 pi(k) such that pi(k) = 1 when ci ≥ k(=1/3), otherwise pi(k) = 0. The share of possible deprivations suffered by ithpoor can be
obtained by ci (k)=[pi(k)]. Ʃ dj=1 wj xij] and the average deprivation share across the poor (intensity of poverty) obtained by A = (1/qk ). Ʃki=1 ci (k). The MPI defined as the product of incidence and intensity of poverty. So, the value of MPI of a society is obtained by MPI = H × A.

However, the major advantage of the measure is its decomposability across population subgroups and across indicators. Thus, the MPI is decomposable by the population subgroups because the measure is expressible as the weighted sum of individual poverty; it is also decomposable by its indicators because the measure is expressible as the weighted sum of the censored deprivations by indicators to obtain contribution of each indicator to the MPI.

Data for Analysis

This study uses the second, third, and fourth rounds of the National Family Health Survey (NFHS) database, but not its first round because it did not cover nutritional information, a major concern for poverty. The surveys used a multistage, stratified random sampling method to collect samples. The NFHS-2 data sets, collected in 2000, contain information on 1,290 sample households of Tripura; the NFHS-3, collected in 2006, comprises 1,695 sample households of the state; and the NFHS-4, collected in 2015, contains 4,576 sample households altogether. This study considers the years 2000, 2006, and 2015 and is thus inter-temporal in nature. However, to make the comparisons of multidimensional poverty across three time periods robust, the study goes through certain adjustments on some of the indicators of the MPI following Alkire and Seth (2013) (Table 1).

Of the 10 indicators listed in the table, six are identical to the international MPI, but due to the differences in the NFHS data sets four indicators—nutrition, mortality, school attendance, and flooring material—have been modified to set deprivation cut-offs. The nutrition indicator in the NFHS-4 data set has been adjusted to be compatible with the information available in the NFHS-2 and NFHS-3. The international MPI considers undernourishment of both women and children, but the present study only considers child undernourishment for comparing all the three time points under study. In the mortality indicator, the present study considers under-five mortality instead of child mortality at any age as in the international MPI in order to match the information available in different data sets of NFHS.

Similarly, the indicator of school attendance is not directly comparable across the NFHSs owing to their methodological differences (the period of survey does not coincide with any particular academic year). The present study considers whether a school-aged child (7–14) is attending the school during the survey period instead of considering a particular academic year. If a household has no child in the schoolgoing age, the household is treated as non-deprived. The indicator “flooring material” has been considered as one of the indicators of the standard of living dimension in the international MPI. This paper considers “housing” as an indicator instead of flooring to make the different time points of survey comparable. For instance, the NFHS-2, 3, and 4 collected information on the type of houses in common.

Performance of Tripura

This study attempts to determine the changes in the overall circumstances of the poor in Tripura during the past 15 years; it analyses the comparable indicators to do so. Considering multidimensional poverty by indicator, it is found that Tripura’s performance has improved in each of the indicators. Table 2 represents the proportion of the deprived population by each selected indicator for the reference years 2000, 2006, and 2015 along with their lower and upper bounds for 95% confidence intervals. The last two columns of Table 2 display the absolute and percentage changes in deprivation from 2000 to 2015 along with their level of statistical significance. This study considers that a reduction in deprivation is statistically significant if the lower bound of an estimate of 2000 is higher than the respective upper bound of the 2015 estimate.

Volatility is observed in the nutrition and mortality indicators of the health dimension. From 2000 to 2006, deprivation increased 15.2% for nutrition mortality and 26.7% for child mortality. These may be the results of the agrarian crisis, rapid rise in food prices, and increasing unemployment. Yet, another factor may have been government policy, such as the weakening of the public distribution system (pds), which led the average calorie intake to decline in rural India (Patnaik 2008, 2013). Tripura—which is mostly rural and largely dependent on government support—is also affected. Rising malnutrition and poor health services are highly associated with mortality, leading to increased child mortality in Tripura. Household deprivation in terms of nutrition and mortality, like other indicators, declined between 2006 and 2015, but not the deprivation in nutrition and health in 2015 over 2000, due to maybe the weakening of the PDS, hike in food prices, and growing unemployment. Other indicators—like sanitation, cooking fuel, electricity, and housing—are part of the ongoing government developmental programmes and, in most cases, do not depend on the capacity of households. Therefore, the problem lies with the demand side, not the supply side.

Table 2 reveals that the largest absolute reductions have taken place in some of the standard-of-living indicators. During the 15-year period of reference from 2000 to 2015, the highest reduction (of more than 60%) was among those deprived in housing and sanitation. Deprivations in electricity, drinking water, and assets have also been reduced by more than 20 percentage points, and the percentage of people using firewood, crop residue, kerosene, or cow dung cake for cooking has gone down by 15.7 percentage points.

In Tripura, the percentage of people deprived of cooking fuel fell from 86% in 2000 to 70.3% in 2015. However, the reduction of deprivation has been slow in education—from 25.7% in 2000 to 11.2% in 2015—and appears to be statistically insignificant (3.8 percentage points) in the case of child attendance. The rate of reduction in the health dimension during the reference period also reveals to be statistically insignificant. However, deprivation in the nutrition indicator has reduced by 5.2 percentage points from 15.2% in 2006 to 10.0% in 2015 and that of mortality indicator has reduced by 18.1 percentage points from 26.7% in 2006 to 8.6% in 2015.

Regarding the relative changes of deprivation from 2000 to 2015, the relative changes in the education dimension are 56.4% for years of schooling and 88.3% for child attendance. The relative changes in the standard-of-living dimension are 18.3% for cooking fuel and 80.6% for the sanitation indicator. The rate of reduction in the sanitation, electricity, housing, and drinking water indicators are higher than that in the years of schooling in both the absolute and relative sense. The proportion of people deprived in both of the indicators of the education dimension—years of schooling and child attendance—has undergone a decline, as have the standard-of-living indicators drinking water, cooking fuel, electricity, and assets.

In the case of Tripura, the picture of changes in multidimensional poverty in terms of MPI and multidimensional headcount ratio (H) and average deprivation (intensity) among the poor (A) using a poverty cut-off of one-third of all weighted indicators (k = 1/3) is demonstrated in Table 3. It reveals that the value of MPI decreased at the rate of 4.6% per annum from 0.163 in 2000 to 0.05 in 2015 and that it is statistically significant at 0.05 levels. Considering H, the component of MPI, the higher rate of reduction in MPI is due to mainly a statistically significant decline in the incidence of poverty, H. Although the fall in intensity of poverty, A, is also statistically significant, the magnitude of decline is much lower than that of incidence. Overall, Tripura could reduce the proportion of the multidimensionally poor at the rate of 1.7% per annum by 24.9 percentage points from 37.2% in 2000 to 12.3% in 2015. The maximum decline, from 33.5% to 12.3%, was from 2006 to 2015, a total reduction of 21.2 percentage points of multidimensionally poor people, at the rate of 2.12 percentage points per annum.

The reduction in deprivation is not identical for the 10 indicators; therefore, it is important to explore the changes in the composition of poverty. Considering the depth of poverty, it is necessary to find out the contribution of each of the 10 indicators to overall poverty and the changing pattern of their contribution over time. Table 4 illustrates the censored headcount ratio (CHCR), percentage of multidimensional poor deprived in individual indicators, and the contribution of each of the indicators to the overall poverty. The CHCR of an indicator represents the proportion of population to the total multidimensionally poor and to those deprived in that indicator. The percentage of the multidimensional poor deprived in individual indicators considers the proportion of the MPI poor deprived in a particular indicator to the total number of MPI poor. The weighted average of the CHCR of all indicators is identical with the value of MPI.

Table 4 shows that the reductions in the CHCR are statistically significant for all the indicators of education and standard-of-living dimension during the 15-year reference period. The reductions in the CHCR do not necessarily replicate the pattern of reduction in the uncensored (raw) deprivations reported in Table 2. The reductions in child attendance are almost uniform in 2000 and 2015. However, the reduction in the CHCR is higher for years of schooling, cooking fuel, and assets, and it is lower for the rest excepting nutrition and mortality. In the case of nutrition and mortality indicators of the health dimension, no specific pattern has been followed.

Between 2006 and 2015, there has been a significant reduction in deprivation of the nutrition and mortality indicators (Table 2), though the reduction between 2000 and 2015 is not significant. The reduction in the CHCR for nutrition between 2006 and 2015 is a statistically significant 6.4 percentage points, relatively higher than the reduction in the uncensored headcount ratio. The reduction in the CHCR for mortality is 14.3 percentage points, slightly higher than the reduction in the uncensored headcount ratio. The relative changes in the corresponding CHCRs are slightly higher in education and standard-of-living indicators than the uncensored headcount ratios, except for sanitation between 2000 and 2015. The relative reductions in the CHCR are highest in housing and child attendance during the 15-year period, followed by electricity, sanitation, drinking water, years of schooling, and so on.

Over the reference period, Tripura could reduce multidimensional poverty significantly due to mainly the reduction in the incidence of poverty (H) along with the higher reductions in the CHCRs of the standard-of-living indicators (15–35 percentage points). Between 2000 and 2015, the CHCRs of education indicators declined by 3.9 percentage points in child attendance and 18.3 percentage points in years of schooling; between 2006 and 2015, the CHCRs of the health indicators declined by 6.4 percentage points in nutrition and 4.3 percentage points in mortality.

The interpretation of the CHCR with respect to the percentage of poor people will help examine the pattern of changes in deprivation among the multidimensionally poor in each indicator. For example, 69% of the multidimensionally poor in 2000 were deprived in the years of schooling indicator and in 2015, 59.3% remain as deprived. Likewise, around 34% of MPI poor are deprived in the nutrition and mortality indicators in 2015. In the case of the standard-of-living indicator, more than 90% of MPI poor are deprived in cooking fuel and asset holding indicators in 2015, whereas 58.1% of the poor are deprived in the sanitation indicator and 48.5% in the drinking water indicator. The contribution of individual indicators to overall poverty over time is highest for years of schooling followed by mortality, cooking fuel, nutrition, and assets. Thus, these indicators appear to be the important policy variables for reducing poverty in the state of Tripura.

Performance across the Population Subgroups

Tripura reduced poverty significantly between 2000 and 2015, but was the reduction in poverty uniform across the major population subgroups (rural and urban; SCs, STs, OBCs, and General; and Hindus, Muslims, Christians, and Others [Buddhists and Jains])?

Rural–urban: Both the MPI and the incidence of poverty (H) for rural and urban areas register statistically significant reductions between 2000 and 2015 (Table 5, p 57). There was a satisfactory reduction of poverty in terms of the headcount ratio (H) for both rural and urban areas, and the reduction was higher for rural areas of Tripura, but the reduction in the intensity of poverty (A) was not as satisfactory anywhere. The reduction of poverty in urban Tripura may have been relatively lower due to the increasing population pressure for rural–urban migration. During the reference period, the population share of urban Tripura increased by 9.6 percentage points from 18.8% in 2000 to 28.4% in 2015.

The intensity of poverty (A) fell by 3 percentage points in rural Tripura, but by 4.4 percentage points in urban Tripura between 2000 and 2015, probably owing to a greater share of amenities (such as health, education and other facilities).

Social and religious groups: This study categorises the population into SCs, STs, OBCs, and General. The General category includes all households except those residing in households self-identified as SCs, STs, and OBCs. Table 6 outlines the reduction of poverty across social and religious groups in Tripura over the reference period. There have been statistically significant reductions in multidimensional poverty in terms of both MPI and H for each of the four subgroups, but the reduction has not been uniform. The largest reduction in the proportion of multidimensionally poor is observed among the SCs (36.3 percentage points) from 45.2% in 2000 to 8.9% in 2015; the smallest reduction is among the General category. Besides, in 2015, the highest MPI and the incidence of poverty (H) are found among the STs and the lowest among the OBCs.

Again, to examine how poverty changed across religious groups over time, the present study classifies the population into Hindus, Muslims, Christians, and Others. According to the NFHS-4 of 2015, 83% of the population in Tripura are Hindu and 8.6% Muslim. Both the MPI and incidence of poverty (H) fell across the religious groups and the fall is statistically significant, but the proportion of the multidimensionally poor appears to be highest (31.6%) among the Others in 2015, followed by Muslims (21.7%), Christians (13.5%), and Hindus (10.4%). The average deprivation or intensity of poverty (A) was almost the same for all religion subgroups.

Household characteristics: To understand the variations in poverty reduction, this study decomposes the population in household subgroups by the household characteristics like gender of the household head and household size. The population share of female-headed households in Tripura increased by 3 percentage points from 7.7% in 2000 to 10.7% in 2015 (Table 7). Poverty fell among both male and female-headed households. However, the MPI for male-headed households decreased at a statistically significant level (0.115 points) in comparison to female-headed households, and the incidence of poverty (H) for male-headed households declined by 25.5 percentage points, from 37.3% in 2000 to 11.8% in 2015.

The proportion of people living in households with five or fewer members has increased from 49.9% in 2000 to over 74% in 2015. Multidimensional poverty in terms of both MPI and incidence of poverty is higher in the case of households having six or more family members as per 2015 estimates. Poverty fell significantly across all household subgroups, but the reduction has been higher among households of five or fewer members; and poverty fell more among the least-member households than among the more-member (eight or more) households.

Conclusions

During the reference period, the incidence of poverty (H) fell in Tripura, and the reduction in the CHCRs of the standard-of-living indicators was higher; that is why poverty fell significantly in the state. This reduction of poverty in terms of MPI and incidence is higher in rural areas, and the relatively lower reduction of poverty in urban areas maybe due to the increasing population pressure from rural–urban migration. The reduction in the intensity of poverty (A) is lower in rural areas than in urban areas because the share of amenities (such as health, education and other facilities) in urban areas is greater.

For the social groups, the largest reduction in the proportion of the multidimensionally poor is observed among the SCs, and the highest MPI and the incidence of poverty (H) are found among the STs. Regarding religious subgroups, the proportion of multidimensionally poor appears to be the highest among the others, followed by Muslims, Christians, and Hindus. Poverty reduction is higher among male-headed households and households with smaller family than their counterparts. The policy formulation for inclusive development of the state of Tripura should adequately address all such varied perspectives of social welfare.

References

Ahluwalia, Montek S (2011): “Prospects and Policy Challenges in the Twelfth Plan,” Economic & Political Weekly, Vol 46, No 21, pp 88–105.

Alkire, Sabina and James Foster (2007): “Counting and Multidimensional Poverty Measures,” OPHI Working Paper 07, University of Oxford.

— (2011): “Counting and Multidimensional Poverty Measurement,” Journal of Public Economics, Vol 95, No 7, pp 476–87.

Alkire, Sabina and Suman Seth (2008): “Determining BPL Status: Some Methodological Improvements,” Indian Journal of Human Development, Vol 2, No 2, pp 407–24.

— (2013): “Multidimensional Poverty Reduction in India between 1999 and 2006: Where and How?” OPHI Working Paper No 60, University of Oxford.

Atkinson, Anthony B (2003): “Multidimensional Deprivation: Contrasting Social Welfare and Counting Approaches,” Journal of Economic Inequality, Vol 1, No 1, pp 51–65.

Bourguignon, Francois and Satya R Chakravarty (2003): “The Measurement of Multidimensional Poverty,” Journal of Economic Inequality, Vol 1, No 1, pp 25–49.

Chakravarty, Satya R and Conchita D’Ambrosio (2006): “The Measurament of Social Exclusion,” Review of Income and Wealth, Vol 52, No 3,
pp 377–98.

Deaton, Angus and Jean Drèze (2009): “Food and Nutrition in India: Facts and Interpretations,” Economic & Political Weekly, Vol 44, No 7, pp 42–65.

Debnath, Niranjan and Salim Shah (2015): “Regional Disparity in Terms of Economic Well-being Using Multidimensional Poverty Index: A Case Study in Tripura,” Vidyasagar University Journal of Commerce, Vol 20, pp 136–50.

Dehury, Bidyadhar and Sanjay K Mohanty (2015): “Regional Estimates of Multidimensional Poverty in India,” Economics: The Open-access, Open-assessment E-Journal, Vol 9, pp 1–35.

Government of India (2009): “Report of the Expert Group to Review the Methodology for Estimation of Poverty,” Planning Commission, New Delhi.

— (2014): “Report of the Expert Group to Review the Methodology for Measurement of Poverty,” Planning Commission, New Delhi.

Jayaraj, D and S Subramanian (2010): “A Chakravarty-D Ambrosio View of Multidimensional Deprivation: Some Estimates for India,” Economic & Political Weekly, Vol 45, No 6, pp 53–65.

Kurian, N J (2000): “Widening Regional Disparities in India: Some Indicators,” Economic & Political Weekly, Vol 35, No 7, pp 538–50.

Laderchi, Caterina R, Ruhi Saith and Frances Stewart (2003): “Does it Matter that We Do Not Agree on the Definition of Poverty? A Comparison of Four Approaches,” Oxford Development Studies, Vol 31, No 3, pp 243–74.

Mishra, Ankita and Ranjan Ray (2013): “Multidimensional Deprivation in India during and after the Reforms: Do the Household Expenditure and the Family Health Surveys Present Consistent Evidence?,” Social Indicators Research, Vol 110, No 2, pp 791–818.

Nayyar, Gaurav (2005): “Growth and Poverty in Rural India: An Analysis of Inter-state Differences,” Economic & Political Weekly, Vol 40, No 16, pp 1631–39.

Patnaik, Utsa (2008): “Theorising Poverty and Food Security in the Era of Economic Reforms,” Globalisation and the Washington Consensus: Its Influence on Democracy and Development in the South, Gladys Lechini (ed), Buenos Aires: Consejo Latinoamericano de Ciencias Sociales, pp 161–200.

— (2013): “Poverty Trends in India 2004–05 to 2009–10: Updating Poverty Estimates and Comparing Official Figures,” Economic & Political Weekly, Vol 48, No 40, pp 43–58.

Roy, Kajari and Sushil K Haldar (2008): “Estimation of Multidimensional Poverty in India: Trends and Regional Variations,” Indian Journal of Regional Science, Vol XXXX, No 2, pp 7–20.

Sen, Amartya (1999): Commodities and Capabilities, India: Oxford University Press.

Sen, Abhijit and Himanshu (2004): “Poverty and Inequality in India–II Widening Disparities during the 1990s,” Economic & Political Weekly, Vol 39, No 39, pp 4361–75.

Seth, Suman and Sabina Alkire (2014): “Measuring and Decomposing Inequality among the Multidimensionally Poor Using Ordinal Data: A Counting Approach,” OPHI Working Papers 68, University of Oxford.

Shah, Salim and Niranjan Debnath (2015): Regional and Social Disparity a Special Focus on Tripura, Kolkata: Progressive Publishers.

Subramanian, S (2011): “The Poverty Line: Getting It Wrong Again,” Economic & Political Weekly, Vol 46, No 48, pp 37–42.

Tsui, Kai-yuen (2002): “Multidimensional Poverty Indices,” Social Choice and Welfare, Vol 19, pp 69–93.

Wang, Xiaolin, Hexia Feng, Qingjie Xia and Sabina Alkire (2016): “On the Relationship between Income Poverty and Multidimensional Poverty in China,” OPHI Working Paper 101, University of Oxford.

Updated On : 13th Jan, 2020

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top