Multi-dimensional Poverty Among the Handloom Weavers in Odisha,India: Prevalence and Determinants

Abstract

This study focuses to examine the prevalence and determinants of poverty among handloom weavers in a non-monetary multi-dimensional approach. Drawing from a field study in four villages of Odisha which consists of 1,853 individual members in 435 weaver households, the Alkire–Foster methodology has been used for the estimation of multi-dimensional poverty. Three dimensions namely health, education and standard of living are considered to construct the multi-dimensional poverty index. The results show that around a quarter of the sample handloom weavers are multi-dimensional poor and are deprived in 40% of the indicators. Apart from this, another 29% are vulnerable to poverty. Nutrition under the health dimension and asset ownership, sanitation and cooking fuel under standard of living dimension have contributed the most to the overall MPI score. The results of regression analysis infer that household size, income, education, skill and social category are the significant factors in affecting the magnitude of poverty. Results point to a pressing need for extensive implementation and wider coverage of existing government schemes to the target population.

Keywords

Multi-dimensional poverty Alkire–Foster methods handloom weavers MPI Odisha

Introduction

Traditionally the concept of poverty has been defined as the inability of a person or household to meet their basic needs leading to deprivation in overall well-being (World Bank, 2000). Therefore, poverty is a kind of quantification of income or consumption required to meet certain basic necessities of human beings including both food and non-food demands (Haughton & Khandker, 2009). However, poverty is not the bare deficiency of income to meet basic requirements rather deprivation in basic human capabilities (Sen, 1999). This means that looking poverty in monetary terms cannot capture the non-monetary aspects of destitution being experienced by an individual or a family. Therefore, poverty measurement in monetary approach in terms of income or consumption has been severely criticised in recent times. Poverty is distinguished as multi-dimensional deprivations, which should be aimed at fulfilling multiple basic needs of human beings such as a minimum level of education, basic health care facility and an improved living condition (Bourguignon & Chakravarty, 2003; Tsui, 2002). Therefore, the Multi-dimensional Poverty Index (MPI) offers a reliable framework that can complement global income poverty estimates (Alkire & Santos, 2014).

The concept of MPI was globally recognised when the flagship Human Development Report got published by Oxford Poverty and Human Development Initiative (OPHI) jointly with United Nations Development Programme in 2010. In this report, OPHI, based on the methodology suggested by Alkire and Santos (2010), calculated global multi-dimensional poverty for more than 100 countries by considering three dimensions that is health, education and living standard. However, Alkire and Foster (2011) improved this methodology based on Foster–Greer–Thorbecke indices which got wide popularity due to its property of multi-dimensional monotonicity. It has a range of technical and practical advantages that makes it favourable for the estimation of non-monetary poverty. The Global MPI 2021 says that India is in the 66th position out of 109 countries with an MPI score of 0.118 (NITI Aayog, 2021). According to the baseline report of the national MPI 2021, multi-dimensional poverty is mostly concentrated in rural areas in India. Other studies also infer similar observations (Das et al., 2021; Das & Paria, 2021; Tripathi & Yenneti, 2020). Again, the incidence of multi-dimensional poverty is observed mostly in the eastern states of India (Dehury & Mohanty, 2015; Tripathi & Yenneti, 2020).

Poverty has been seen as a rural phenomenon as around 70% of India’s population live in rural areas (Census, 2011). Most of the rural population are engaged in the informal sector where the wage rate is comparatively low and hence are deprived of basic amenities. The higher incidence of poverty in rural areas can be explained by the larger presence of the informal sector (Kathuria & Raj, 2016). These activities include cultivation, household occupation, petty street vendors and casual labour and one of such informal activities in India is handloom weaving. Handloom weaving is an age-old occupation that has its relevance to being a distinctive feature of the cultural heritage of India. With 3.5 million workers engaged in this sector, it is the second most employers in the informal sector after cultivation (Ministry of Textiles, Government of India, 2019b). Contributing 95% to the export of the world’s woven fabrics, this sector is important to the Indian economy in terms of output, export and employment generation (Ministry of Textiles, Government of India, 2019a).

Weaving in the hand-operated loom is a traditional ancient activity carried out mostly in the rural areas in Odisha, the eastern state of India. Due to unsustainable agriculture and industrial backwardness in the state which hampers employment generation, the handloom industry acts as an important source of income and livelihood for the masses in Odisha. It has the potential to employ unskilled and semiskilled rural populations. According to the fourth National Handloom Census 2019–2020, there are 1.18 lakh handloom workers in Odisha spread over in 63,223 weaver households. From among the total handloom workers, more than 97% are in the rural areas. Due to its excellent artwork in ikat (both single and double) with natural motives and attractive colour combinations, the handloom products of Odisha are highly appreciated all over the world (Meher & Rout, 2020). Despite its wide acceptance and high demand in the country and overseas, the handloom weavers in Odisha earn a meagre income which poses a threat to their sustainable livelihood (Meher, 1995). Around 99% of the weaver households earn below ₹10,000 income per month with most of the handloom weavers (55.7%) having primary or below the primary level of education. Also, half of them have a kutcha house to work and reside (Ministry of Textiles, Government of India, 2019b).

The present study was conducted at a major pocket of handloom workers in the state that is, Bargarh district which is located in the western most part of Odisha. According to the fourth National Handloom Census, 2019–2020, there are 16,110 handloom weavers with 21,168 allied workers in Bargarh district spread over 17,569 households. This constitutes a considerable number of populations in the district and highest number of handloom workers in the state. However, the District Development and Diversity Index constructed by Shariff (2015), places Bargarh district in the 435th position among 599 districts in India for whom the index was prepared. The report categorises Indian districts in five groups with respect to development criteria such as most-developed, developed, developing, under-developed and least-developed where it identifies Bargarh as an under-developed district. From among the four most suitable indicators taken in this study, Bargarh is found out as least developed in economic index, under developed in education index, less developed in health index and under developed in material well-being index. This shows the intensity of deprivation being suffered by the households in the district.

The National Handloom Census conducted by the Government of India does not provide data on other indicators of MPI apart from monthly income earned by the handloom weavers, their level of education and housing status. According to the baseline report of national MPI 2021, 29.35% population in Odisha are multi-dimensional poor which is above the national average of 25.01%. Given this backdrop with earlier studies on the living and working condition of the handloom weavers in the state and the development parameters of the district, the authors suspect multiple deprivations being experienced by the handloom weavers. Therefore, it is pertinent to study the non-monetary aspects of deprivations being experienced by the handloom weavers in Odisha. Hence, this article attempts to investigate the prevalence and determinants of multi-dimensional poverty among them. The novelty of this article lies in capturing the multiple dimensions of non-monetary aspects of poverty among the handloom weavers in Odisha which up to the authors’ knowledge no studies have done so far. The evidence from primary data collected through field surveys would provide insight into the multi-dimensional deprivations being experienced by the handloom weavers.

The remaining parts of the article are planned as follows. The second section presents the review of literature. The third section explains the data and methodology used for the study. Empirical findings and discussion are provided in the fourth section. Finally, the fifth section concludes the study with conclusion and policy implications.

Literature Review

Various studies have been conducted since 2010 after the publication of 1st MPI report globally to study the dynamics of MPI and its determinants at various national, state and regional levels. These studies are based on both primary and secondary data where secondary information is mostly taken from the country-specific demographic surveys. Globally, many studies have been conducted to find out the incidence and intensity of multi-dimensional poverty. According to Alkire and Santos (2010), sub-Saharan Africa contained the highest multi-dimensional poverty headcount whereas most of the people living in poverty were found in South Asia. After these two, come the Arab States and East Asia and the Pacific who are multi-dimensional poor both in incidence and intensity. Investigating multi-dimensional poverty in Sindh province of Pakistan, Khan et al. (2014) found out that the magnitude of multi-dimensional poverty was much higher than the average poverty in monetary terms. The level of poverty was worse in the rural areas as compared to the urban areas. Higher levels of deprivation in education, health and housing facilities contributed more to the magnitude. Vulnerability was more widespread than poverty despite the prevalence of very strong traditional social support networks and semi-subsistence lifestyles in Melanesia (Feeny & Mcdonald, 2016). Around 50% of the population in Shan and 75% in Chin of Myanmar were multi-dimensional poor with an average intensity of 38% in Shan and 44% in Chin. Deprivation in education and deprivation in health contributed the highest to multi-dimensional poverty in Shan and Chin, respectively. Households in the rural areas, with lower educational attainment, consumption poor and those lived in Chin were significantly higher multi-dimensional poor (Mohanty et al., 2018).

Studies on MPI at the national level are diverse and hence some of the relevant ones are presented here. According to Alkire and Seth (2008), up to 12% of the sample poor population and 33% of the acute poor in India could be misidentified as non-poor by the pseudo-BPL method and hence income poverty measure cannot alone describe the poverty issue. In India, 43% of the population were multi-dimensional poor with large variations across regions. Economic dimension accounted for 22% of the deprivation followed by the health dimension (36%), the household environment (31%) and the education dimension (11%). Around 45% of the population in India had a concentration of more than 50% of the multi-dimensional poor spread over in the states of Bihar, Jharkhand, Chhattisgarh, Odisha, Madhya Pradesh, Uttar Pradesh and West Bengal (Dehury & Mohanty, 2015). The problem of multi-dimensional poverty in Uttar Pradesh was at par with that of India. The major factors responsible for it were low level of income and education, problematic occupational pattern, unsafe fuel used for cooking, unhygienic sanitation facilities and lack of proper nutritional diet (Unjum & Mishra, 2017).

Multi-dimensional poverty headcount in West Bengal was higher than that of the national level but the intensity of poverty was similar. Degree of poverty varied widely across the social castes where SCs were most deprived and OBCs were least. Use of conventional cooking fuel and lack of improved sanitation facility were the major indicators contributing highly to the district MPI (Bagli & Tewari, 2019). The incidence of poverty declined significantly from 2000 to 2015 with a higher decrease in MPI and incidence in the rural areas compared to the urban areas in Tripura. In the rural areas, the reduction in intensity of poverty was lower than the urban areas. The incidence of poverty and MPI was highest among the STs and other social groups. There was a larger reduction of poverty among the SCs, households headed by male and smaller family size (Debnath & Shah, 2020). There was a steep reduction in rural poverty in comparison to urban poverty in India. The highest contributor to poverty was low level of education followed by insufficient income and poor standard of living. The MPI analysis at the state level revealed that there was a higher concentration of poverty headcount in the states of Arunachal Pradesh, Jharkhand, Rajasthan, Uttar Pradesh, Bihar, Odisha and Chhattisgarh (Tripathi & Yenneti, 2020).

One-fourth of the total sample households in the districts of North India were in the clutch of poverty whereas more than 50% were vulnerable to it. The dimensions such as education, standard of living and economic and social security contributed significantly to the multi-dimensional poverty and vulnerability. The key indicators of deprivation under these dimensions were cooking fuel, assets ownership, sanitation, social security measures and informal jobs (Mishra et al., 2020). The concentration of multi-dimensional poverty was highest among rural areas in the eastern region whereas it was lowest in the northern region across social sub-groups in India. There was regional concentration of MPI particularly in the central and eastern regions which increased by four times in 2015–2016 from 2005 to 2006 (Das et al., 2021). Around 33% of the population in urban India was multi-dimensional poor and about 16% were vulnerable to poverty. Higher concentration of multi-dimensional poverty was found among over-crowed households, households headed by female, widowed and scheduled tribes (Mohanty & Vasistha, 2021).

The above literature shows that most of the studies across the globe are based on secondary data obtained from country-specific demographic surveys. In India, most of the studies used National Family Health Survey (NFHS) data to analyse multi-dimensional poverty which does not have direct economic variables. Primary studies based on population and occupational sub-groups such as SC, ST, OBC, formal and inform sector workers are few. Studies focusing on Odisha state are rare which has been identified as a multi-dimensional poor state by other studies. Particularly the handloom weavers have not been considered as the focus of the study either at the national or at the state level. Therefore, considering the present status of the handloom weavers in Odisha, this study attempts to answer the questions; (a) what is the incidence and intensity of multi-dimensional poverty prevalent among the handloom weavers and (b) what are the factors significantly contributing to the multi-dimensional poverty? It is hypothesised that the handloom weavers in Odisha are experiencing multiple deprivation in terms of poverty and the major factors contributing to their deprivation are both economic and social characteristics of the handloom weavers’ household. Suitable policy measures are also suggested to overcome these problems.

Materials and Methods

Data

This study is developed on the basis of primary data collected through field survey and supplemented by secondary data published in government reports and research articles. Following Yamane (1967) and his simplified formula¹ to determine sample size, a total of 435 handloom weaving households in Bargarh district of Odisha state have been selected for the study. The study area has been selected based on multi-stage sampling methods where in the first stage, Bargarh district is chosen based on highest concentration of handloom weavers in the state. In the subsequent stages, the blocks and villages are selected following the same criteria that is intense clustering of the handloom weavers. Within the villages, the active handloom weaving households have been chosen using random sampling technique without replacement. In this way two blocks namely Bijepur and Sohela and four villages (two villages in each block) namely Laumunda and Bijepur village in Bijepur block and Sarkanda and Jhar village in Sohela block have been selected for the study. Proportional sampling method (taking into account total weavers’ population) has been employed to determine the sample households in each block and village.² This gave rise to selection of 278 households in Bijepur block (173 households in Laumunda and 105 households in Bijepur) and 157 households in Sohela block (97 households in Sarkanda and 60 households in Jhar).

Using pre-tested structured interview schedule, the head of the household or main worker who was having most knowledge about the social and economic situation of the family has been interviewed for detailed information. The comprehensive survey gathered information from the respondents on major aspects of the household such as health, education and living standard which is necessary for the present study. This study was conducted during the COVID recovery period that is, in the months of August, September and October 2021 when COVID-19 norms were relaxed in the state. Primary data was processed in Ms-Excel for the calculation of multi-dimensional poverty and STATA was used to run the regression analysis. Table 1 shows the demographic profile of the sample respondents. It reveals that the average age of the sample handloom weavers is 41.23. On an average 4.26 persons are residing in a household from among which 2.81 persons are workers and 1.44 persons are dependents. Almost all the sample respondents are male and married. Most of them are educated up to secondary level and majority of them belong to OBC category (see Table 1).

Table 1.

Demographic Characteristics of the Sample Respondents.

	Mean	SD	Range
Age (in years)	41.23	10.71	18–71
Household size	4.26	1.41	2–14
Number of workers	2.81	1.06	2–8
Number of dependents	1.44	1.08	0–6
	Frequency (N = 435)		Percentage
Gender
Male	432		99.3
Female	3		0.7
Relationship status
Married	396		91.0
Unmarried	37		8.5
Divorced	1		0.2
Widowed	1		0.2
Education
≤ Primary	172		39.5
> Primary ≤ Secondary	232		53.3
≥ Higher secondary	31		7.2
Social category
SC	22		5.1
ST	30		6.9
OBC	383		88.0

Source: Compiled from primary data, 2021.

Methods

There are several methods for measuring multi-dimensional poverty depending upon availability of data. The standard methodology given by Alkire and Foster (2011) has been followed widely by scholars to estimate multi-dimensional poverty. This methodology provides the flexibility to consider dimensions and indicators based on information availability. Therefore, the present study also follows the same method as it can be used for decomposition of population sub-groups and also by deprivation indicators.

Estimation of MPI

The Alkire–Foster methodology employs a dual cut-off level where it picks out the poor in each deprived indicator in the first stage and then the second cut-off aggregates the weighted sum of deprivation in separate indicators and dimensions. Three dimensions are taken into account for the calculation of global MPI such as health, education and standard of living. It assigns equal weights to each dimension and within a dimension; equal weights are assigned to each indicator. Generally, three indices are estimated in the AF methodology; incidence and intensity of poverty expressed as H and A, respectively, and the MPI. They are described as follows:

The incidence of poverty also known as the headcount ratio (H) is the fraction of population that is multi-dimensional poor in the total sample population and denoted as

H = q / n,

where q represents total number of multi-dimensional poor and n represents total population.

The intensity of poverty (A) is the average weighted deprivations of indicators experienced by the multi-dimensional poor and is denoted as

A = \frac{1}{q} \sum_{i = 1}^{q} c i

where $\sum_{i = 1}^{q} c i$ is the sum of weighted deprivation experienced by the individuals and q represents total number of multi-dimensional poor.

The MPI is the product of H and A and is denoted as

M P I = H * A

The contribution of each indicator to overall MPI can be counted as

w_{i} C H_{i} / M P I_{t o t a l} * 100

where w_i represents weight of the ith indicator and CH_i represents censored headcount ratio of the ith indicator.

As the AF methodology provides the flexibility to consider context-specific dimensions, indicators and weights, it is the focus of estimation procedures. This study follows the estimates of global MPI and hence three indicators such as health, education and standard of living have been considered to calculate multi-dimensional poverty among the handloom weavers in Odisha. The details of dimensions, indicators and respective weights have been given in Table 2. Equal weights have been assigned to each dimension under consideration and within each dimension; equal weights are assigned to each indicator. A household is termed as multi-dimensional poor if its deprivation score (c) is greater than or equal to one-third of the total deprivation scores that is, one (see Table 2).

Table 2.

Weights, Deprivation Cut-offs, Indicators and Dimensions for Calculating MPI.

Dimensions (weights)	Indicators (weights)	Deprivation cut-offs: A household is deprived if…..
Health (1/3)	Nutrition (1/6)	Any household members for whom the nutritional information is malnourished, as measured for the adults by their body mass index and for the children by the height-for-age z-score.
Health (1/3)	Child mortality (1/6)	Any child below five years of age has died in the household in the last five years preceding the survey.
Education (1/3)	Years of schooling (1/6)	Five years of schooling is not completed by any members of the household.
Education (1/3)	School attendance (1/6)	Any school-aged child is not attending school up to the age at which he/she should complete class eight.
Standard of living (1/3)	Electricity (1/18)	The household does not have electricity connection.
	Drinking water (1/18)	The household does not have access to safe drinking water or the source of safe drinking water is more than 30 minutes’ roundtrip walk from home.
	Sanitation (1/18)	The household does not have improved toilet facility or it is improved but shared with other household members.
	Housing (1/18)	The family members reside in a kutcha house.
	Cooking fuel (1/18)	The household cooks using wood, coal, cow dung, crop residue etc.
	Assets ownership (1/18)	The household has not more than one of these assets: TV, radio, bicycle, telephone, computer, refrigerator, motorbike and does not have a car or truck.

Source: Compiled from the report of Global Multi-dimensional Poverty Index, 2021.

Regression Analysis

The magnitude or severity of multi-dimensional poverty depends on many socio-economic components of the household. Therefore, in order to understand how poverty has been affected by these factors, a multiple regression analysis has been employed. As multi-dimensional poverty has been measured in terms of degree of deprivation of the households, the dependent variable in this case is households’ deprivation score. As the deprivation score of a particular household increases, the severity of poverty escalates and vice versa. The independent variables and the regression model are specified as follows:

\begin{array}{l} D S_{i} = α + β_{1} (M I H_{i}) + β_{2} (E H H_{i}) + β_{3} (H S_{i}) + β_{4} (A H H_{i}) + β_{5} (O B C D_{i}) + \\ \begin{matrix}  \end{matrix} β_{6} (S C D_{i}) + β_{7} (S T D_{i}) + β_{8} (S H G D_{i}) + β_{9} (S D_{i}) + β_{10} (L O D_{i}) + e_{i} \end{array}

where α denotes the intercept, β is the coefficients to be estimated and i is the ith household. However, DS, MIH, EHH, HS and AHH refer to deprivation score of the household, monthly income of the household in rupees, education of the main worker/head of the household in terms of schooling years, household size and the age of the household head, respectively. Moreover, OBCD is the OBC dummy; it becomes one if the household is OBC or else zero, SCD is the SC dummy; it becomes one if the household is SC or else zero, STD is the ST dummy; it becomes one if the household is ST or else zero, SHGD is the self-help group (SGH) dummy; it becomes one if the household members (female) participate in self-help groups or else zero, SD is the skill dummy; it becomes one if any of the household members are skilled³ or else zero, and LOD is the land ownership dummy; it becomes one if the household poses any land or else zero. Finally, e denotes the error terms.

In poverty measurement, all the methods draw a line of estimation to identify the poor. Generally, that population which touches the cut-off line or lies above it is considered as the poor. However, there exist a large number of people just below the cut-off line who are left out to be included in poor category. This population is called as the vulnerable group who will be having high chance of getting into poverty in near future if any adverse situation arises. Therefore, while studying and estimating poverty it is necessary to keep an eye on these people who are vulnerable to poverty through suitable policy measures. In order to study the magnitude of multi-dimensional poverty, the sample weaver households are classified into four categories as per their deprivation score. Group 1 with a deprivation score of greater than zero and less than or equal to 0.20 is termed as multi-dimensional non-poor, group 2 having privation score of greater than 0.20 and less than equal to 0.33 is labelled as vulnerable to poverty. Similarly, with the deprivation score of greater than 0.33 and less than equal to 0.50, group 3 is normal multi-dimensional poor and group 4 is severe multi-dimensional poor having deprivation score of greater than 0.50.

Empirical Results and Discussion

Prevalence of Multi-dimensional Poverty

This section is devoted to estimation of multi-dimensional poverty among the handloom weavers and its prevalence across the sample villages which comprise of 1,853 individual members in 435 sample households. As the headcount ratio, intensity and MPI is calculated at the individual level, results show that the headcount ratio of 1,853 sample individuals is 0.234 with the intensity score of 0.408 and MPI value of 0.095. This indicates that the handloom weavers are in a better condition than both the state MPI score of 0.136 and national MPI score of 0.118 (NITI Aayog, 2021). The indicators wise censored and uncensored headcount ratio is presented in Table 3. It shows that most of the people are deprived in asset ownership followed by sanitation and nutrition. In case of possession of household assets, most of the respondents revealed that they received those assets as dowry during their marriage and very less has acquired it from their own pocket. A country wide campaign viz. Swachh Bharat Abhiyan was initiated in 2014 by the Indian Government for eliminating open defecation to control the communicable diseases and to provide a better living standard to the masses. However, it is observed from the field study that this initiative has not been able to achieve its targeted goal due to huge corruption in implementation. Though government provided fund for construction of toilets free of cost, the respondents expressed that the government officials built those toilets which is very small in size and of very poor quality and hence they are not using it. It is also noticed that many of them are having improved toilet facilities in their houses but still they prefer for open defecation. It shows the lack of interest of the people to use toilet and hence need mass awareness for the same (see Table 3).

Table 3.

Indicators Wise Censored and Uncensored Headcount Ratio.

Indicators	Uncensored H (in %)	Uncensored H (in %)
Nutrition	43.33	20.29
Child mortality	14.78	10.63
Years of schooling	4.85	3.93
School attendance	1.29	0.80
Electricity	0	0
Drinking water	7.44	3.40
Sanitation	58.2	18.67
Housing	21.91	8.52
Cooking fuel	33.40	15.60
Asset ownership	61.62	20.50

Source: Primary survey, 2021.

The deprivation in nutrition is mostly found among the children and teens which is a major policy concern and questions the implementation of government’s free mid-day-meal schemes and others for its targeted objectives. Few people are deprived in education dimension which is a good sign of proper working of government’s free and compulsory education facility. The sample individuals are aware of the benefits of getting education and hence they are educating their children. No one is deprived of electricity in the study area which is a better indicator of improved well-being among the masses. However, the respondents revealed that there is frequent power cut and they receive continuous power supply for hardly 14–16 hours a day. This creates huge disturbance in doing waving work and the situation is pathetic during summer.

Analysing multi-dimensional poverty based on different population sub-groups will be helpful to identify the region and indicator specific concentration of poor and accordingly policy measures can be administered for their upliftment. Table 4 shows the decomposition of incidence of poverty in terms of headcount ratio, intensity and MPI in different villages of the study area. It can be seen that the incidence of poverty is highest in Jhar followed by Bijepur, Sarkanda and Laumunda. As Jhar in Sohela block is a small village having higher numbers of SC and ST population, this might be the possible reason for highest incidence of poverty. Bijepur in Bijepur block itself is a notified area council, being urbanised to a small extent. Still a considerable number of poor handloom weavers can be attributed to various reasons and needs further investigation to find out the cause.

The intensity of poverty exhibits how poorer the poor are. In other words, it elucidates the extent to which the poor people are deprived in different indicators. Laumunda and Jhar villages are equally deprived in 41% of the total indicators whereas Sarkand and Bijepur are deprived in 39% and 22% of the indicators respectively. So, Jhar village is poorest among the four having highest incidence and intensity of poverty and hence resulting in a higher MPI score of 0.205. Though Bijepur is having second highest incidence of poverty, the poor are deprived in lowest numbers of indicators due to urbanisation where people enjoy many urban facilities. Laumunda and Sarkanda are having almost equal intensity score which is a considerable one and hence needs early intervention in this regard (see Table 4).

Table 4.

Village Wise Decomposition of Incidence, Intensity and MPI.

Village Name	Incidence	Intensity	MPI
Laumunda	0.17	0.41	0.085
Bijepur	0.39	0.22	0.085
Sarkanda	0.20	0.39	0.078
Jhar	0.50	0.41	0.205

Source: Primary Survey, 2021.

The village-wise deprivation of sample population in different indicators is illustrated in Table 5. It unveils that Jhar village is having highest deprivation in almost all the indicators which is observed from its top score in incidence and intensity of poverty. However, no one is deprived of school attendance in this village. The handloom weavers in Sarkanda are mostly deprived in asset ownership which is highest among the four villages. Around half of the sample population is deprived in nutrition indicator and a considerable number of people (31.73%) are deprived in sanitation facility. Bijepur exhibits equal proportion of deprivation in sanitation and asset ownership followed by nutritional deprivation. Most of the sample individuals in Laumunda are deprived in sanitation facility followed by asset ownership and cooking fuel. It shows lowest deprivation in nutrition indicator among the four villages (see Table 5).

Table 5.

Village Wise Deprivation of Sample Population in Different Indicators.

Indicators	Village Name
Indicators	Laumunda (%)	Bijepur (%)	Sarkanda (%)	Jhar (%)
Nutrition	26.91	53.69	49.51	60.72
Child mortality	12.69	14.37	16.58	18.62
Years of schooling	05.02	05.07	02.88	07.28
School attendance	0.41	02.95	01.68	0
Electricity	0	0	0	0
Drinking water	07.67	02.74	01.68	25.50
Sanitation	67.50	56.87	31.73	78.94
Housing	26.63	17.12	12.09	34.00
Cooking fuel	31.93	27.90	23.07	65.58
Asset ownership	52.71	56.02	70.19	83.80

Source: Primary Survey, 2021.

Contribution of different dimensions and indicators to overall MPI makes us to understand in which dimensions and indicators policy measures should be focused on. Table 6 represents all the dimensions and indicators of multi-dimensional poverty and their contribution to MPI. It shows that health dimension is having highest and education dimension is having lowest contribution to MPI. Within health dimension, nutrition contributes the most to MPI followed by child mortality. Though standard of living dimension has second highest contribution to MPI, asset ownership, sanitation and cooking fuel indicators within this dimension handout the most to overall MPI. This finding corroborates mostly with the findings by Dehury and Mohanty (2015), Bagli and Tewari (2019) and Mishra et al. (2020, 2021) (see Table 6).

Classification of households as per their deprivation score is necessary to identify the vulnerable population which are generally left out of policy domain constituted for the official poor population. Therefore, this study classifies the total sample weaver households in four poverty groups as per their deprivation score. Table 7 conveys that around 50% of the sample households are multi-dimensional non-poor and around one-fourth of them are multi-dimensional poor from among which a few falls in severe poor category. A significant number of households (29%) are vulnerable to poverty which is clearly noticed from the table. This group has a greater chance of getting into poverty with any unfavourable situation being experienced in the near future. Therefore, while framing targeted policies to uplift the multi-dimensional poor, it should also be taken into account that the vulnerable people would not fall into poverty subsequently (see Table 7).

Table 6.

Contribution of Each Dimension and Indicator to Overall MPI.

Dimensions	Contribution to MPI (in %)	Indicators	Contribution to MPI (in %)
Health	53.7	Nutrition	35.3
Health	53.7	Child mortality	18.4
Education	08.1	Years of schooling	06.7
Education	08.1	School attendance	01.4
Standard of living	38.2	Electricity	00.0
		Drinking water	01.9
		Sanitation	10.7
		Housing	04.9
		Cooking fuel	09.0
		Assets ownership	11.7

Source: Primary Survey, 2021.

Table 7.

Classification of Households as Per Deprivation Score.

Poverty Groups	Deprivation Range	Frequency (%)
1. Multi-dimensional non-poor	0 < C ≤ 0.20	209 (48.0)
2. Vulnerable to poverty	0.20 < C ≤ 0.33	126 (29.0)
3. Normal multi-dimensional poor	0.33 < C ≤ 0.50	91 (20.9)
4. Severe multi-dimensional poor	0.50 < C ≤ 1	9 (2.1)

Source: Primary Survey, 2021.

Determinants of Multi-dimensional Poverty

Poverty acts as a curse upon humanity and hence no one wants to live in poverty. However, there are several socio-economic and demographic factors responsible for poverty among the people. In order to examine the determinants of multi-dimensional poverty, a multiple regression analysis has been carried out with the household deprivation score being the dependent variable. The severity of poverty depends on the deprivation score having a positive relationship between them. The independent variables taken here are household size, age of the household head, years of schooling of household head, monthly income of the household from weaving and allied activities, whether household members are skilled or not, whether the female members participate in the SHG or not, land ownership status and social category of the households. The household size is a major determinant of poverty because as the size increases, more people comprise of children and older ones will be dependent on the working members. This will result in inadequate availability of food, education and other well beings.

The head of the household is generally considered as the main bread earner of the family. Therefore, as the age of household heads increases, they become more experienced and productive resulting in higher income. However, they gradually become unproductive after crossing a certain age limit which differs from person to person. Monthly income of the household is negatively related with the severity of poverty. As income increases predicament declines and vice versa. Poverty decreases with the increase in education as people acquire more skills and hence higher income and other well beings. In this study, it is assumed that as weavers become skilled ones who know and do the ikat work in their family they earn a higher income compared to those who do not. The SHG programme is constituted by the government to empower the women by making them financially independent and hence helping them to overcome poverty. The land ownership by the households fetches additional income resulting in financial stability and other well beings. It is a general consensus that the SC and ST population constitute a major part of the poor population in India due to being marginalised in every aspect of life. In this regression analysis ST category is taken as the reference group.

Table 8 infers the regression results where deprivation score of the households is regressed with the above-mentioned explanatory variables. After calculating the regression results, the model went through other diagnostic tests such as autocorrelation, multi-collinearity and heteroscedasticity. We did not find presence of any of these problems and hence the regression model appeared significant with an F value of 11.006. A total of 18.9% of the variation in the dependent variable is explained by the independent variables. Though the R-squared value is statistically less, we can consider the model as the data is primary and cross sectional in nature. It is observed from the regression results in Table 8 that household size, monthly income, education, skill and OBC category emerged to be the significant factors determining deprivation score among the handloom weaving households. One unit increase in household size leads to 0.130 units increase in deprivation score. Likewise, every unit increase in monthly income and years of schooling reduces the household deprivation score by 0.287 and 0.159 units, respectively. With the increase in skill level, severity of poverty decreases by 0.101 units. When the household is OBC, deprivation score declines by 0.172 units (see Table 8).

Table 8.

Determinants of Multi-dimensional Poverty with Regression Coefficients.

Explanatory Variables	Coefficients	t value
Household size	0.130*	2.511
Age of household head	−0.074	−1.450
Monthly income	−0.287*	−4.907
Years of schooling	−0.159*	−3.189
Skill dummy	−0.101**	−2.018
SHG dummy	0.031	0.665
Land ownership dummy	−0.057	−1.162
SC dummy	0.004	0.066
OBC dummy	−0.172*	−2.712
F-statistic	11.006 (P value: .000)
R-squared	0.189
Number of observations	435

Symbols * and ** indicate significance at 1% and 5% significance level.

Poverty estimation in monetary approach has been severely criticised by the scholars for not being able to reflect the non-monetary aspects of deprivation being experienced by the households. Therefore, the multi-dimensional poverty measurement has been widely prevalent in the present time which estimates poverty among the masses beyond monetary sense and focuses on other household deprivations. Though there are studies on multi-dimensional poverty both at the international and national level, there are few studies available which examine poverty based on population sub-groups across states in India. Analysis of multi-dimensional poverty among the informal handloom weavers in the Odisha state is carried out for the first time in this study. Primary data collected from 435 active handloom weaving households which comprise of 1,853 individual members in four villages of Bargarh district in Odisha shows that 23% of the sample households are multi-dimensional poor. This figure is below the national average of 29.35% and state average of 25.01%. Multi-dimensional poverty measured at the individual level in terms of incidence, intensity and MPI stands at 0.23, 0.40 and 0.095, respectively. The MPI score among the handloom weavers is lower than the 2021 national MPI score of 0.118 and state MPI score of 0.136. However, deprivation in the indicators such as asset ownership followed by sanitation, nutrition and cooking fuel is mostly prevailing among the sample population. It is contented to know that no one in the study area is deprived of electricity connection.

Village wise decomposition shows that Jhar village is the poorest one with highest incidence, intensity and MPI score. Incidence of poverty is also considerable in Bijepur and intensity of poverty is almost equally severe in both Sarkanda and Laumunda village. Village wise deprivation of sample population in different indicators reveals that Jhar village is the most deprived in several indicators. Apart from Jhar village, majority of the population is deprived in asset ownership in Sarkand village, whereas deprivation in sanitation is highest in Laumunda. Most of the people are nutritionally deprived in Bijepur and greater deprivation in cooking fuel is found in Laumunda. Health dimension contributed the most to overall MPI followed by standard of living. Though 23% of the sample population is found out to be multi-dimensional poor, another 29% are vulnerable to poverty. The results of regression analysis indicate that the variables like household size, monthly income of the household, education level of the household head, skill of the household members and OBC category significantly affected the severity of poverty measured in terms of deprivation score.

Conclusion and Policy Implications

This study observed that the handloom weavers in the Bargarh district of Odisha are in a better condition than the state and national status measured in terms of multi-dimensional poverty. However, there are some key indicators where the handloom weavers are experiencing their worst. This requires immediate intervention in terms of suitable policy measures for its implementation in order to uplift the poor handloom weavers from multi-dimensional poverty. The nutrition indicator in health dimension has highest contribution to MPI which should be taken very seriously as Odisha has past evidences of acute child malnutrition and death. The NFHS-5 (International Institute for Population Sciences, 2021) also shows that 31% of the children in Odisha are stunted. Therefore, the nutritional schemes by the government such as Integrated Child Development Service and Mid-Day-Meal scheme should be intensively implemented with wide coverage. The quality of food provided under these schemes should be properly inspected by the authority to substantiate that the children are being supplied with adequate nutrition. Under the standard of living dimension, asset ownership, sanitation and cooking fuel indicators have highest contribution to MPI where most of the households are deprived. A better income and awareness among the people to have household assets of their own and not to depend on dowry can solve this problem to a greater extent. Some assets such as fan, cooler and TV can be provided to the weavers under different government schemes constituted for the handloom weavers.

Deprivation in sanitation and cooking fuel needs extensive implementation of the existing government schemes. Stringent regulation against the corrupted implementing officials and mass awareness among the people for using toilets would be beneficial in this regard. Many of the households in the study area received free gas cylinder and stove under central governments’ Ujjwala scheme. However, during the survey they revealed that due to lack of money and soaring price of gas cylinder, they are not being able to refill it and hence they are bound to use conventional source of energy for cooking. In this case, targeted policy needs to be framed to provide at least one gas cylinder per month in subsidised price to the actual needy households. As a whole, there are some loopholes in implementation of the existing government schemes which if corrected on time will result in a better health and living standard for the handloom weavers in Odisha. It requires proper coordination and wholeheartedness between the central and state government to work together to achieve broader social and economic interest.

Nonetheless, considering the limitations and scope for further investigation, this study can be generalised for handloom weavers along with other informal sector workers in and outside the state given similar socio-economic status. This study is also limited by the dimensions and indicators that are on par with the international standard. Therefore, further study in this area to identify the most suitable deprivation indicators would be helpful to investigate other lacunas in the development path of the informal handloom weavers in India.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

References

Alkire

, & Foster

(2011). Counting and multidimensional poverty measures. Journal of Public Economics, 95(7/8), 476–487.

Alkire

, & Santos

M. E.

(2010). Acute multidimensional poverty: A new index for developing countries [Oxford Poverty & Human Development Initiative (OPHI) Working Paper No. 38 United Nations Development Programme Human Development Report Office Background Paper No. 2010/11].

Alkire

, & Santos

M. E.

(2014). Measuring acute poverty in the developing world: Robustness and scope of the multidimensional poverty index. World Development, 59, 251–274.

Alkire

, & Seth

(2008). Measuring multidimensional poverty in India: A new proposal. University of Oxford.

Bagli

, & Tewari

(2019). Multidimensional poverty: An exploratory study in Purulia district, West Bengal. Economic Affairs, 64(3), 517–527.

Bourguignon

, & Chakravarty

S. R.

(2003). The measurement of multidimensional poverty. Journal of Economic Inequality, 1(1), 25–49.

Census. (2011). Census of India 2011—Provisional population totals. Office of the Registrar General and Census Commissioner.

Das

, Ghosh

, & Paria

(2021). Multidimensional poverty in India: A study on regional disparities. GeoJournal, 87, 3987–4006.

Debnath

, & Shah

(2020). Multidimensional poverty in Tripura. Economic & Political Weekly, 55(2), 53.

10.

Dehury

, & Mohanty

S. K.

(2015). Regional estimates of multidimensional poverty in India. Economics, 9(1), 2015–2036.

11.

Feeny

, & McDonald

(2016). Vulnerability to multidimensional poverty: Findings from households in Melanesia. The Journal of Development Studies, 52(3), 447–464.

12.

Haughton

, & Khandker

S. R.

(2009). Handbook on poverty and inequality. World Bank.

13.

International Institute for Population Sciences. (2021). National Family Health Survey (NFHS-5), 2019–21.

14.

Kathuria

, & Raj

S. N. R.

(2016). Do informal sector wages explain rural poverty? Evidence from India. Journal of Poverty, 20(1), 73–101.

15.

Khan

A. U.

, Saboor

, Hussain

, Sadiq

, & Mohsin

A. Q.

(2014). Investigating multidimensional poverty across the regions in the Sindh province of Pakistan. Social Indicators Research, 119(2), 515–532.

16.

Meher

A. K.

, & Rout

(2020). Rural economy in Odisha can be revived by the handloom industry. Orissa Economic Journal, 52(4), 94–108.

17.

Meher

(1995). The handloom industry and the socio-economic conditions of weavers in Orissa. Journal of Rural Development, 14(3), 301–322.

18.

Ministry of Textile, Government of India. (2019a). Annual reports 2018–19. http://texmin.nic.in/sites/default/files/Textiles-AnnualReport2018-2019%28English%29.pdf

19.

Ministry of Textiles, Government of India. (2019b). Fourth all India handloom census 2019–20. http://handlooms.nic.in/writereaddata/3736.pdf

20.

Mishra

P. K.

, Kumar

, Arif

, Niyazi

A. U.

, & Kaur

(2020). Poverty in multidimensional perspective: Policy insights from selected North Indian districts. Millennial Asia, 13(2), 289–316.

21.

Mohanty

S. K.

, Rasul

, Mahapatra

, Choudhury

, Tuladhar

, & Valdemar Holmgren

(2018). Multidimensional poverty in mountainous regions: Shan and Chin in Myanmar. Social Indicators Research, 138(1), 23–44.

22.

Mohanty

S. K.

, & Vasishtha

(2021). Contextualizing multidimensional poverty in urban India. Poverty & Public Policy, 13(3), 234–253.

23.

NITI Aayog. (2021). National multidimensional poverty index: Baseline report. NITI Aayog.

24.

Sen

(1999). Development as freedom. Oxford University Press.

25.

Shariff

(2015). District development and diversity index: Report for India and major states (USIPI Occasional Paper No. 6). US-India Policy Institute.

26.

Tripathi

, & Yenneti

(2020). Measurement of multidimensional poverty in India: A state-level analysis. Indian Journal of Human Development, 14(2), 257–274.

27.

Tsui

(2002). Multidimensional poverty indices. Social Choice and Welfare, 19(1), 69–93.

28.

Unjum

, & Mishra

(2017). Multidimensional poverty in Uttar Pradesh: Trends and patterns. Indian Journal of Economics and Development, 5(11), 1–6.

29.

World Bank. (2000). World development report 2000/2001: Attacking poverty. World Bank.

30.

Yamane

(1967). Statistics, An introductory analysis (2nd ed.). Harper and Row.