Socio-economic Determinants of Multidimensional Poverty in Rural West Bengal: A Household Level Analysis

Rowntree (1901) defined poverty primarily as those families whose earnings are insufficient to meet the minimum necessities for the maintenance of merely physical efficiency. In the life-cycle of an individual poverty can be found by the sudden fluctuations in the income and employment of his father beside other family members. He defined the poverty line using work efficiency of individuals that, in turn, is related to biological requirements.

A family with income less than one half the median family income can be defined as relatively poor (Fuchs 1967). Townsend (1974) has argued that relative poverty should incorporate objective conditions with ‘feelings of deprivation'.

The biological approach was under attack due to difficulty in determining minimal nutritional requirements, which were varied internationally. Sukhatme (1981, Sukhatme and Paresh 1983) has argued that the energy requirements of an individual are not fixed and it varies between individuals (due to inter-individual genetic differences) and minimum calorie norms depends on age, sex, body weight, type of work, occupation, etc. The income or expenditure method identifies the poor as those who lack the purchasing power to meet a subsistence level of living. This technique entails the estimation of the poverty line, i.e., the minimum income needed to meet the basic needs and persons with money income or expenditure below this threshold line are regarded as poor (Rao 1977; Dandekar and Rath 1971; Bardhan et al. 1974; Deaton 1997). Since both spending and income are measured in money, the choice of the cut-off or poverty line is always somewhat arbitrary (HDR 1996).

Capability refers to a person's or groups' freedom to promote valuable functioning. Or, in other words, the capability is a set of values of functioning, reflecting the person's freedom to lead the decent life (Sen 1992). By the term ‘Freedom', Sen (1999) has mentioned five kinds of ‘freedom'. These are political, economic facilities, social opportunities, transparency guarantees and reproductive securities. Basic capabilities refer to (a) leading a life free of avoidable morbidity (b) being informed and educated and (c) being well-nourished (Sen 1997a, b). A basic capability is a capability to enjoy a functioning that is defined as a general level and refers to a basic need (i.e., a capability to meet a basic need) or, in other words, a capability to avoid malnourishment, a capability to be educated and so on (Alkire 2002).

The three variables in the capability poverty measure covers substantial ground—indication of nutrition and health for the population as a whole (measured by underweight children), access to reproductive health services and a concrete test of access to health services in general (measured by unattended trained health personnel) and basic educational attainment plus information on gender inequality (measured by female adult illiteracy). The index relating to capability poverty measure emphasizes critical areas where progress is needed must remain especially in poor countries (HDR 1996). Therefore, capability poverty measure (CPM) is defined as: \( \textit{CPM} = \frac{{P_{1} + P_{2} + P_{3} }}{3} \), where P₁ = percentage of children under age six who are under-weight, P₂ = percentage of births unattended by a trained health professional and P₃ = percentage of adult women who are illiterate.

Human Poverty Index, \( P(\alpha ) = \left[ {\frac{{w_{1} P_{1}^{\alpha } + w_{2} P_{2}^{\alpha } + w_{3} P_{3}^{\alpha } }}{{w_{1} + w_{2} + w_{3} }}} \right]^{{\frac{1}{\alpha }}} \), where \( P_{1} \) is the percentage of people who do not survive after age 40, \( P_{2} \) = percentage of adults who are illiterate and \( P_{3} = \frac{{P_{31} + P_{32} + P_{33} }}{3} \), P₃ is a composite variables in which \( P_{31} \) = people without access to safe water (in percentage), \( P_{32} \) = people without access to health services (in percentage) and \( P_{33} \) = moderately and severely underweight children under five (in percentage); here, α is the order of the average (?).

In the context of development and deprivation perspective, Kamdar and Basak (2005) has made a distinction between the output (and attainment) and input (or process) indicator. Output or attainment indicators reflect the status of the population in that particular dimension of development; for example, life expectancy is an outcome indicator of the health status of the population indicating the number of years a person is expected to survive at age one or birth. Access to health services is an input indicator or a process indicator as it contributes to higher life expectancy.

The human poverty index for OECD countries (HDR 1998) measures deprivations in the same dimensions as the HPI for the developing countries and also captures social exclusion. Thus it reflects deprivations in four dimensions: (a) A long and healthy life—vulnerability to death at a relatively early age. The probability at birth of not staying alive to age 60 measures this. (b) Knowledge—exclusion from the world of reading and communications, as measured by the percentages of adults (aged 16–65) lacking functional literacy skills. (c) A decent standard of living—as measured by the percentage of people living below the income poverty line (50% of the median disposable household income). (d) Social exclusion—as measured by the rate of long-term unemployment (12 months or more).

The health indicator is a percentage of people not expected to survive beyond age 40 (P₁). The second index (P₂) is comprised of two variables: illiteracy rate for the population in the age group 7 years and above and proportion of children in the age group 6–18 years not enrolled in the schools. The economic deprivation indicator (P₃) is a composite of four parameters: percentage of people below poverty line (P₃₁); proportion of children in the age group 12–23 months not fully vaccinated (P₃₂); proportion of people living in kutchha houses (P₃₃); and proportion of people have not been able to get access to safe drinking water, electricity and health facility (P₃₄).

Human Poverty Index, P(α) does satisfy some essential properties like homogeneity, substitutability, convexity but not decomposability; generally, P(α) cannot be considered as a head-count ratio (HCR). Following the endnote 6, if we assume P₁ = P₂ = P₃ = P and α = 1, then and only then P(α) can be considered as HCR, but in reality, it is rarely happened (UNDP 1997).

The Global Hunger Index (GHI) developed by International Food Policy Research Institute (IFRI) is a multidimensional statistical tool used to describe the starvation condition of different countries. It is released annually by IFPRI since 2006. The GHI is calculated by taking into account four indicators; these are (1) undernourished population (1/3rd weight), (2) child wasting (1/6th weight), (3) child stunting (1/6th weight) and (4) infant mortality rate (1/3rd weight). Stunting is defined as the deficiency in height about age reflecting chronic undernutrition; wasting means low weight about child's height reflecting acute malnutrition.

A broad discussion of poverty in India among different socio-ethnic and religious groups can be found in Meenakshi et al. (2000), Panagariya (2008), Panagariya and Mukim (2013), Thorat (2006) and Thorat and Deubey (2012).

The generalized entropy (GE) class of measure of inequality has some advantages over gini-coefficient; GE measure is sub-group decomposable, and it is distribution sensitive, and one can quickly reconcile between GE measure of inequality with Atkinson measure. Theil Index belongs to GE(α) measures of inequality for if α = 1. Therefore, following Cowell and Jenkins (1995) and Litchfield (1999), we can write the Theil Index [GE(1)] in m-group decomposable form as:

The within-group inequality and between-group inequality are captured by the first and second term of the above formula. Here m stands for a number of groups, n_k be the population size of the kth group, n be the total population, \( \bar{y}_{k} \) means mean income (in our case, it is deprivation score) of the kth group, \( \bar{y} \) be the mean of the overall population. Theil Index [viz. GE(1)] of the kth group is represented by T_k.

FGT(α) can be expressed as:

here n* be the number of poor people, x* be the threshold level of income used to determine poverty \( g_{i} = x^{*} - x_{i} \), and x_i is the income of the ith poor. Equation (A1) can be expressed as:

here, \( K_{i} = \frac{{g_{i} }}{{x^{*} }} \)

If incomes of the poor are arranged as \( x_{1} < x_{2} < \ldots x_{n* - 1} < x_{n*} \), therefore, FGT(α)₁is defined as:

Expanding (14) and taking the limit α → ∞ \( \frac{{FGT(\alpha )_{1} }}{FGT(\alpha )} = 1 \), this proves that FGT(α)₁ = FGT(α). Here, FGT(α)₁ stands for the poorest of the poor.