Child Deprivation in Gujarat

Abstract

This article analyses the deprivation faced by children below five years of age in Gujarat, one of the most economically developed states of India, on the basis of a child deprivation index that has been constructed for the purpose. The analysis reveals that the deprivation faced by the children of the state is quite pervasive and there are significant within state variation with the situation appearing alarming in some districts. The analysis calls for a district-based approach of addressing the deprivation faced by the children of the state with a focus on building the capacity of women.

Keywords

India Gujarat child deprivation child rights child deprivation index

Introduction

Children, along with women, constitute the most vulnerable group of the population. Children are not full economic and social agents. They cannot secure resources necessary for their survival, growth and development until a certain age. They have no or very limited freedom in making decisions related to their own welfare and benefit (White, Leavy, & Masters, 2002). They depend on elders of the household, including their parents, in meeting their basic needs. They also have to rely up to a significant extent upon the production of public goods and services in education and health (Gordon, Nandy, Pantazis, Pemberton, & Townsend, 2003a, 2003b; Minujin, Delamonica, Gonzalez, & Davidziuk 2005; Notten & de Neubourg, 2011; Waddington, 2004; White et al., 2002). These and many other dependencies of children get manifested in poor social and economic settings. Poverty, at the early stages of life, has enduring consequences on those who survive into adulthood. It condemns them to recurrent poverty spells and a life full of hardship (Grinspun, 2004). As such, there has been an increased interest in recent years on the impact of poverty on survival, growth and development of children (Lister, 2004). This focus is best reflected in the Millennium Development Goals (United Nations, 2000). Two of these goals—universal primary education and reduction in child mortality—are directly related to children while the goal of eradicating extreme poverty and hunger has an indirect impact on the well-being of children.

Another reason behind the increased attention to the well-being of children is the United Nations Convention on the Rights of the Child which lays down principles of non-discrimination in the best interest of the child along with common standards for various rights of children. It takes into account different cultural, social, economic and political realities in which children live (United Nations, 1989). By ratifying the Convention in 1992, India has committed herself to protecting and advancing children’s rights; to developing and undertaking all actions and policies in the light of the best interests of children; and to hold herself accountable for this commitment before the international community.

Rights of children in India have been enshrined in the Fundamental Rights and the Directive Principles of the state policy as inscribed in the Constitution of India. These rights were reaffirmed in the National Policy on Children, announced, for the first time, in 1974 (Government of India, 1974). Since then, there have been a number of attempts of mainstreaming child rights issues in the development discourse of the country. These include the Campaign against Child Labour launched in 1992; the Campaign against Child Trafficking launched in 2001 as part of the International Campaign against Child Trafficking, the Child Rights Group within the World Social Forum and so on. A renewed thinking at the government level is also reflected in the Integrated Child Protection Scheme which aims to promote the best interests of the child, to prevent violations of child rights through appropriate punitive measures against perpetrators of abuse and crimes against children and to ensure rehabilitation for all children in need of care and protection (Government of India, 2007).

Despite all these provisions and commitments, protecting rights of children in India remains a major development challenge. Although child protection is increasingly being recognised as a human rights concept, yet, children, and their conditions, are still considered as their father’s property who is seen as the natural guardian of the child. Traditional structures of patriarchy and other social groupings continue to justify extreme forms of chastisement of children (Kushwah & Prasad, 2009). In order to realise the rights of the child and tackle child poverty, robust measures quantifying the nature and extent of child deprivation are required. These measures should focus specifically on children rather than simply treating children as elements of the household.

This article analyses the deprivation faced by children below five years of age in Gujarat. Although Gujarat ranks amongst one of the most advanced states of India in terms of economic growth, yet, in terms of survival, growth and development of children below five years of age, Gujarat needs serious introspection. Recent estimates indicate that the risk of death during the first five years of life in Gujarat is 45 deaths for every 1,000 live births which is only marginally lower than the national average of 49 deaths per 1,000 live births (Government of India, 2013). Among the leading states of India, Gujarat ranks eighth in terms of the under-five mortality rate according to the sample registration system.

Similarly, the proportion of under-nourished children in the state continues to be exceptionally high despite the economic growth that the state has witnessed in the recent past. According to the rapid survey of children recently carried out by the Government of India, Gujarat ranks 25 in terms of the proportion of children below five years of age who are not low height-for-age (Government of India, 2015). Similarly, the state ranks at 28 in terms of the proportion of children below five years of age who are not low weight-for-height. Under-nutrition is a major contributing factor in the persistence of high to very high risk of death during infancy and early childhood.

The article follows the child rights framework which has been evolved in recognition of the fact that a child is a human being who is dependent upon other human beings till the time she or he grows and develops into a responsible yet productive member of society. This implies that, as human beings, children have a certain moral status that needs to be recognised by the family and the society. This also means that there are things that should not be done to children and things that should be done to children to ensure that they grow and develop as responsible and productive assets of the family and society. The child rights perspective is an attempt to ensure that children are treated in the family and in the society in a morally right way.

The rights that serve the ‘best interests’ of children can be articulated in many ways and can be categorised as positive rights and moral rights. Positive rights are recognised by the law. Moral rights are recognised by some moral theory but accepted by the society as a norm. One argument is that possession of a right is sufficient to outweigh or discount all other moral considerations (Nozick, 1974). The counter argument is that possession of rights may not out-balance every other moral claim because children do not have the necessary capability to exercise the rights bestowed upon them. By just possessing a set of rights, children, obviously, cannot serve their own ‘best interests’. There must be conditions in place to ensure that child rights are actually exercised in an effective yet socially acceptable manner. These conditions are not controlled by children.

The United Nations Convention on the Rights of the Child and the National Policy on Children provide the policy framework to enhance capabilities of children. The realisation of child rights also requires an operational framework to serve as the basis for planning, implementing and evaluating programmes directed towards improving capabilities of children. Such an operational framework can be derived following the capabilities approach (Nussbaum & Sen, 1993; Sen, 1985, 2000), which has widely been applied in welfare economics as the new paradigm of development.

The article is organised as follows. The next section develops the child deprivation index (CDI) to measure child deprivation. The third section describes data sources, while the fourth section presents estimates of the CDI for Gujarat and its districts and discusses inter-district variation in child deprivation on the basis of the CDI and its different domains. The last section summarises findings of the analysis and discusses policy and programme implications of child deprivation in the state. The article suggests that the proportion of children facing one or more than one forms of deprivation in Gujarat is quite substantial and there are significant variations in child deprivation and in different domains of child deprivation across the districts. The deprivation faced by children in the state has implications for their survival, growth and development.

Child Deprivation

Deprivation may be defined as circumstances or situations that are highly likely to have adverse implications to the well-being of an individual. Child deprivation, then, means circumstances or situations or both that are highly likely to have adverse implications to child’s well-being. Mitigating deprivation is critical as addressing survival, growth, development and protection needs of children is essential to achieve their full potential (Minujin, Delamonica, Davidziuk, & Gonzales, 2006). Deprivation measures reflect the degree to which a child needs are actually met (de Neubourg, 2012). Since the basic needs of children are very diverse in nature, it is obvious that child deprivation is a multi-dimensional concept. Condensing different forms of deprivation faced by children into a single index of deprivation, however, is not straightforward and efforts in this direction have been found to be controversial (Ravallian, 2010a, 2010b) or challenging (Atkinson, 2003). Although it makes sense to look into different dimensions of deprivation one-at-a-time, yet resorting to a composite deprivation index is unavoidable when one looks into the breadth of deprivations faced by children (Apablaza & Yalonetzky, 2011).

Deprivation is usually measured through the ‘counting’ approach which counts the number of dimensions or domains in which people suffering deprivation fall below the pre-set threshold (Atkinson, 2003). In case of children, dimensions or domains of deprivation can be identified in different contexts. One context that is most widely used is the child well-being context. Many studies have identified distinct domains of child well-being (Brown, 1997; Child Trends, 2003; Hauser, Brown, & Prosser, 1997; Land, Lamb, & Mustillo-Kahler, 2001, cited in Carroll, 2002; Pollard & Lee, 2002; Reidy & Winje, 2002). Other contexts of child deprivation include child rights (Ben-Arieh, Hevener-Kaufman, Bowers-Andrews, Goerge, Joo-Lee, & Aber, 2001); child needs (Ryan & Deci, 2001); child development (Mickelwright & Stewart, 1999); and child outcomes (Maryland Partnership for Children, Youth and Families, 2002). Dimensions or domains of child deprivation can also be defined following the capabilities approach first propounded by Sen (1985) and later discussed in Nussbaum and Sen (1993) and Nussbaum (2000). Sen’s capability approach has now been universally adopted and applied to characterise social and economic development in the multi-dimensional context. In terms of Sen’s capability approach, domains of child deprivation may be defined in terms of child endowments, child capacities and child opportunities. There exists a congruence between domains identified through the Sen’s capability framework and domains identified through the child well-being framework (Chaurasia, 2010).

The dimensions or domains of child deprivation essentially reflect the theoretical construct of the deprivation faced by children. In order to operationalise this construct, measurable variables or indicators need to be identified. An indicator may be conceived as a measure of a condition or status or behaviour that can be tracked over time, across people or geographical units (Child Trends, 1997). Indicators are defined according to an objective criteria that may be simple as well as complex. Friedman (1997, cited in Carroll, 2002) has suggested a three-point simple criteria. Ben-Arieh et al. (2001) advocate a two-dimensional approach, the first of which is related to validity and relevance while the second is related to policy and programme. Moore (1995, 1997, 1999) has suggested a 13-point criteria, many of which are similar to those suggested by Ben-Arieh et al. (2001). Indicators are usually selected on the basis of data availability but they can also be selected through a policy perspective or on the basis of an underlying theory (Hanafin & Brooks, 2005). It is argued that all the three approaches should be combined to select indicators of child wellbeing (Bauer et al., 2003).

There are two approaches of aggregating different deprivation indicators (Mickelwright, 2001). The first approach combines different elements of deprivation at the individual level, which are then summed over individuals to form an aggregate index at the macro-level. The second approach, on the other hand, sums across individuals to estimate deprivation prevalence in a given dimension and then combines deprivation prevalence in different dimensions to arrive at a composite index of deprivation. The second approach is similar to the human poverty index recommended by Anand and Sen (1997). There are many studies that follow the first approach (Alkire & Roche, 2012; Bradshaw et al., 2009; de Neubourg, Chzhen, Main, Martorano, & Menchini, 2012; Foundation for Child Development, 2013; Nyangara, Nyberg, Murphy, & O’Donell, 2008; Roche, 2013; UNICEF, 2014). On the other hand, Kannamori and Pullum (2013) have followed the second approach to analyse levels and trends in child deprivation in sub-Saharan Africa. Dreze and Khera (2010) have used this approach to construct a summary index of well-being of children in India. The author has used this approach to analyse residence and social class effects of child deprivation in Madhya Pradesh, India (Chaurasia, 2010).

This article follows the second approach to construct a composite index of child deprivation in Gujarat. The CDI constructed in this article is based on 10 indicators grouped into four domains related to the survival, growth and development of children (Table 1). The selection of indicators for the construction of CDI is based primarily on the availability of data, although they have theoretical and policy relevance. The first two indicators are household level indicators that reflect two aspects of household poverty. The first reflects income poverty while the second reflects poverty in terms of household command over resources and assets. Deprivation in the housing domain is captured through the condition of the house and the practice of open defecation which essentially means non-availability of latrine in the house. It is well known that children from households without access to improved sanitation facilities are at increased risk of contracting serious illnesses like dysentery, diarrhoea and typhoid fever. Deprivation in the health domain is captured through three indicators specific to children and are important to survival and growth of children. Finally, deprivation in the opportunity domain is captured through indicators related to women as they play a critical role in the survival and growth of children. Deprivation of women has a strong effect on the deprivation of children.

Table 1

Domains and Indicators Used for the Construction of the Deprivation Index for Children Aged 0–5 Years

Domain and Indicator
Material well-being domain
Proportion of households below poverty line	X ₁
Proportion of households having none of the specified assets¹	X ₂
Housing domain
Proportion of houses not in good condition	X ₃
Proportion of households defecating in the open	X ₄
Health domain
Proportion of children not fully immunised	X ₅
Proportion of children not breastfed within 1 hour of birth	X ₆
Proportion of children not exclusive breastfed for the first 6 months	X ₇
Opportunity domain
Proportion of currently married women (15–49 years) illiterate	X ₈
Proportion of girls marrying before 18 years of age	X ₉
Proportion of women not received full ANC	X ₁₀

Source: Author’s own.

Notes: ¹ Household assets for which information was collected at the 2011 population census include (1) radio/transistor; (2) television; (3) computer with or without internet; (4) phone landline or mobile; (5) bicycle; (6) auto two wheeler motorcycle/scooter/moped, etc. and (7) auto four-wheeler car/jeep/van, etc. Households having none of these seven household assets have been classified as households with no asset at the 2011 population census.

Child Deprivation Index

Using the 10 indicators described in Table 1, the CDI is constructed in the following manner:

All indicators listed in Table 1 range from 0 to 1. Since all indicators range between 0 and 1, there is no need of normalising the indicators.

In the first stage, a domain-specific deprivation index is constructed following the approach suggested by Anand and Sen (1997). More specifically, the deprivation index in material well-being domain is calculated as

I_{m} = {(\frac{x_{1}^{3} + x_{2}^{3}}{2})}^{\frac{1}{3}}

The deprivation index of housing domain is calculated as

I_{u} = {(\frac{x_{3}^{3} + x_{4}^{3}}{2})}^{\frac{1}{3}}

The deprivation index in health domain is calculated as

I_{h} = {(\frac{x_{5}^{3} + x_{6}^{3} + x_{7}^{3}}{3})}^{\frac{1}{3}}

Finally, the deprivation index in opportunity domain is calculated as

I_{o} = {(\frac{x_{8}^{3} + x_{9}^{3} + x_{10}^{3}}{3})}^{\frac{1}{3}}

In calculating the mean of order 3, the relative weight given to an indicator increases as the deviation of the value of the indicator from the mean value of all indicators increases. Therefore, the mean of order 3 addresses the problem of additive compensability associated with simple mean or mean of order 1. There is, however, an escapable arbitrariness in the selection of the order of the mean. When the mean of order 3 is calculated, the relative impact of the dimension with highest deprivation is four times the relative impact of the dimension with lowest deprivation. With the increase in the order of the mean, the relative impact increases very rapidly leading to extreme results.

The CDI is then calculated as the unweighted mean of order 3 of the four domain-specific deprivation indexes:

C D I = {(\frac{I_{m}^{3} + I_{u}^{3} + I_{h}^{3} + I_{o}^{3}}{4})}^{\frac{1}{3}}

It is obvious that the CDI varies between 0 and 1. When CDI = 0, there is no deprivation. When CDI = 1, there is total deprivation which means as the CDI increases, the deprivation faced by the children also increases. The extent of deprivation can be termed as low if CDI is less than 0.2; medium if 0.2 ≤ CDI < 0.3; high if 0.3 ≤ CDI < 0.5; and very high if 0.5 ≤ CDI < 0.7. When CDI $ 0.7, deprivation may be termed as extreme.

Each domain of the CDI reflects a different aspect of deprivation in children below five years of age. This precludes the need of making judgements about the complex inter-linkages across different domains of child deprivation and the contribution that each domain would make to the CDI. It is, however, possible that the same child is captured in more than one domains of deprivation because child deprivation, by its very nature, is multi-dimensional.

Data Source

Two data sources are used in the present analysis. The first data source is the 2011 population census while the second is the District Level Rapid Household Survey (DLRHS) 2010–11 (GSIDBS, 2012). DLRHS covered 11,4675 households; 99,207 currently married women (CMW) aged 15–49 years in Gujarat and covered a population of 514,888 persons. The response rate in DLRHS was 94 per cent for the households and 96 per cent for CMW.

Table 2

Inter-district Variation in Indicators Used in the Analysis

Indicator	Minimum	Median	Maximum	Mean	Standard Deviation	Coefficient of Variation	Skewness
X ₁	10.5	30.3	73.9	32.9	13.686	0.416	1.178
X ₂	7.0	18.1	53.2	22.9	12.461	0.544	0.947
X ₃	22.7	33.4	57.7	34.8	7.684	0.221	1.107
X ₄	11.5	43.9	86.1	48.6	18.801	0.387	0.091
X ₅	5.7	28.0	65.1	30.0	15.430	0.515	0.634
X ₆	3.5	27.0	59.9	29.0	14.925	0.514	0.399
X ₇	16.9	54.3	89.5	51.0	19.630	0.385	–0.115
X ₈	10.2	30.3	53.6	29.9	10.895	0.364	0.251
X ₉	1.2	15.6	56.1	19.8	12.904	0.653	1.138
X ₁₀	47.8	73.6	92.7	73.2	10.635	0.145	–0.568

Source: Author’s calculations.

Estimates of the 10 indicators used for the construction of the CDI are given in the Appendix Table for Gujarat and for its 26 districts as they existed at the time of the 2011 population census. Summary measures of inter-district variation in these indicators are presented in Table 2. All indicators vary widely across districts. The inter-district variability is the highest in the proportion of girls marrying before 18 years of age (X₉) but the lowest in the proportion of women not receiving full antenatal care during their last pregnancy (X₁₀). Inter-district variability has also been found to be quite low in the proportion of houses not in good condition (X₃) but quite high in the proportion of households without specified assets (X₂); proportion of children fully immunised (X₅) and proportion of children breastfed within one hour of birth (X₆). Inter-district distribution is found to be positively skewed in eight indicators but negatively skewed in two indicators. The extent of skewness also varies widely across indicators. This means that inter-district variability in the 10 indicators used in the construction of CDI is essentially different which justifies the use of the indicator set for constructing the CDI.

An important requirement in the construction of any composite index is that indicators used in construction are not collinear. Table 3 shows the simple zero order correlation coefficients among the 10 indicators used for constructing the CDI. The table shows that all correlation coefficients among the variables used in the analysis are well below 0.90. This suggests that the indicators used in the construction of the CDI are not multicollinear.

Table 3

Simple Zero Order Correlation Coefficient among Indicators

	X₁	X₂	X₃	X₄	X₅	X₆	X₇	X₈	X₉
X ₂	0.487
X ₃	0.349	0.797
X ₄	0.353	0.851	0.869
X ₅	–0.134	0.327	0.145	0.299
X ₆	0.462	0.337	0.318	0.234	–0.244
X ₇	0.256	0.053	0.246	0.188	–0.030	0.276
X ₈	–0.068	0.222	0.286	0.245	0.233	–0.053	0.229
X ₉	0.334	0.503	0.298	0.428	0.293	0.473	0.453	0.215
X ₁₀	–0.019	0.223	0.158	0.120	–0.190	0.410	0.030	–0.028	0.350

Source: Author’s calculations.

Child Deprivation in Gujarat

The CDI for Gujarat is estimated to be 0.415 which means that children of the state face a high degree of deprivation in the context of their well-being. Deprivation appears to be relatively the highest in the opportunity domain of child well-being but relatively the lowest in the material well-being domain (Figure 1). The high degree of deprivation faced by the children of the state appears to be one of the reasons behind the abnormally high risk of death faced by children of the state during their first five years of life as is reflected through the under-five mortality rate.

Figure 1

Source: Author’s calculation.

The CDI is estimated to be the highest in Banas Kantha District followed by Dahod District, the only two districts with the CDI more than 0.6 (Figure 2). Other districts where the CDI is estimated to be very high are Narmada (0.592), The Dangs (0.546) and Panchmahals (0.528). On the other hand, the CDI is estimated to be the lowest in Bhavnagar (0.331) followed by Navsari (0.357); Surat (0.359); Junagadh (0.376) and Jamnagar (0.381). There are seven districts where child deprivation may be termed as medium as the CDI in these district, ranges from 0.3 to 0.4. In 13 districts, child deprivation appears to be high as the CDI ranges from 0.4 to 0.5. Finally, there are six districts—Sabar Kantha, Panchmahals, The Dangs, Narmada, Dahod and Banas Kantha—where child deprivation appears to be very high as the CDI is more than 0.5. There is no district where child deprivation may be termed as low or very low.

Figure 2

Source: Author’s calculation.

Table 4

Inter-district Inequality in CDI and Domain-specific Deprivation Indexes

State/District	Index of Deprivation in				CDI
State/District	Material Well-being Domain	Housing Domain	Health Domain	Opportunity Domain	CDI
Ratio of highest to lowest index	5.160	3.975	3.288	1.826	1.712
Coefficient of variation	0.384	0.314	0.273	0.136	0.161
Gini index of inequality	0.209	0.175	0.155	0.075	0.090

Source: Author’s calculations.

Inter-district variation in child deprivation is quite revealing. In Banas Kantha, the district with the highest child deprivation in the state, the CDI is more than 70 per cent higher than the CDI in Bhavnagar District, the district with the lowest child deprivation (Table 4) and is reflected in a coefficient of variation of 0.161 and a Gini index of 0.090. The inter-district disparity in child deprivation suggests that there are significant local-level factors that contribute to the deprivation of children below five years of age in terms of basic needs necessary for their well-being. This observation suggests that addressing child deprivation requires a decentralised district-based approach. It is important that local-level factors that contribute to deprivation among young children are identified and addressed through district-based planning and programming. Strong inter-district inequality also suggests that within the state, disparity in child deprivation appears to have persisted over time.

Figure 3

Source: Author’s calculation.

Figure 4

Source: Author’s calculation.

Inter-district variation in deprivation in different domains of child well-being is different. The inter-district variation in deprivation is found to be the highest in the material well-being domain but the lowest in the opportunity domain of child well-being (Table 4). Deprivation in the material well-being domain is found to be the highest in Narmada District (0.629) but the lowest in Porbandar District (0.122). The CDI in Narmada District is more than five times higher than that in Porbandar District. There are four districts—Porbandar, Junagadh, Gandhinagar and Bhavnagar—where the deprivation in material well-being is very low but in four districts—Tapi, Sabar Kantha, Dahod and The Dangs—it is very high (Figure 3).

In the housing domain, the index of deprivation varies from District Dahod (0.746) to District Surat (0.188). The index of deprivation in District Dahod is almost four times the index of deprivation in District Surat (Figure 4); the coefficient of variation is 0.314 and the Gini index is 0.175. There are only three districts where deprivation in this domain is either low or very low, whereas in eight districts, deprivation is very high. Deprivation appears to be extreme in District Dahod in this domain.

In the health domain, deprivation appears to be the lowest in District Junagadh (0.199) but the highest in District Banas Kantha (0.654) followed by District Sabar Kantha (0. 607) (Figure 5). The index of deprivation in District Banas Kantha is more than three times the index of deprivation in District Junagadh; the coefficient of variation is 0.273 and the Gini index is 0.155. Junagadh is the only district in the state where deprivation in this domain of child well-being is very low. On the other hand, in seven districts, deprivation in this domain appears to be very high.

Figure 5

Source: Author’s calculation.

Finally, in the opportunity domain, deprivation is the lowest in Navsari District (0.375) but the highest in Banas Kantha District (0.688) (Figure 6). The inter-district inequality in this domain is the lowest among the four domains of child well-being. The index of deprivation in Banas Kantha District is more than 1.8 times the index of deprivation in Navsari District; the coefficient of variation is 0.136 and the Gini index is 0.075. There is no district where deprivation in this domain is either low or very low, whereas in 20 districts, deprivation in this domain is very high. Deprivation is medium in two districts—Navsari and Tapi—and high in four districts—Surendranagar, Bhavnagar, Ahmedabad and Kachchh.

Figure 6

Source: Author’s calculation.

Child deprivation appears to have gained alarming proportions in six districts—Banas Kantha, Dahod, Narmada, the Dangs, Panchmahals and Sabar Kantha. The CDI is estimated to be more than 0.5 in these districts. In Banas Kantha, Dahod, Narmada and Panchmahals, the index of deprivation is very high in three of the four domains of child well-being. In The Dangs and Panchmahals, the index of deprivation is very high in two domains. There are only four districts—Bhavnagar, Navsari, Jamnagar and Ahmedabad—where the index of deprivation is not very high in any of the four domains of child well-being. In Surat, Junagadh, Rajkot and Porbandar districts, deprivation appears to be very high in one domain of child well-being but low to very low in the remaining three domains.

It would be interesting to analyse how deprivation in different domains of child well-being contribute to deprivation in children below five years of age. The automatic linear modelling approach (Yang, 2013) is applied for this analysis with CDI as predicted variable and the index of deprivation in material well-being, housing, health and opportunity domains as predictor variables. Automatic linear modelling is a fundamental data mining tool and is an improvement over the traditional regression model. It explores all possible combinations of predictor variables to identify the best predictive model for the predicted variable. SPSS was used to carry out automatic linear modelling.

The results of the analysis are presented in Table 5. The accuracy of the model is found to be very high (97.9 per cent) and coefficients of all predictor variables are found to be statistically significant. The relative importance of the deprivation in the opportunity domain in deciding the level of CDI is the highest among the four domains of child deprivation closely followed by the housing domain. In contrast, deprivation in the material well-being domain is found to be the least important in predicting CDI.

Table 5

Contribution of Deprivation in Different Domains of Child Well-being to Child Deprivation

Predictor	B	SE (B)	B	‘t’	Importance	Accuracy
Deprivation in opportunity domain	0.389	0.031	0.384	12.443	0.358	97.9
Deprivation in housing domain	0.246	0.020	0.460	12.097	0.338
Deprivation in health domain	0.194	0.020	0.310	9.545	0.210
Deprivation in material well-being domain	0.151	0.024	0.239	6.385	0.094

Source: Author’s calculations.

Table 6

Results of the Regression of CDI and Domain-specific Deprivation indexes on Proportion Urban, Proportion Scheduled Castes and Proportion Scheduled Tribes

Dependent Variables	Standardised Regression Coefficient			R²
Dependent Variables	Proportion of Urban Population	Proportion Scheduled Castes	Proportion Scheduled Tribes	R²
Child deprivation index (CDI)	–0.507*	0.474*	0.682*	0.626
Index of deprivation in material well-being domain (I_m)	–0.196	0.196	0.825*	0.630
Index of deprivation in housing domain (I_u)	–0.632*	0.399*	0.583*	0.741
Index of deprivation in health domain (I_h)	–0.437	0.210	.0131	0.107
Index of deprivation in opportunity domain (I_o)	–0.135	0.326	0.294	0.070

Source: Author’s calculations.

Note: * significant at p=0.05.

Districts of the state vary widely in terms of at least three population characteristics that may influence the deprivation faced by children—level of urbanisation, the proportion of Scheduled Castes population and the proportion of Scheduled Tribes population. Table 6 shows how inter-district variation in CDI and index of deprivation in different domains of child well-being are influenced by inter-district variation in the proportion of population urban, the proportion of the Scheduled Castes population and the proportion of the Scheduled Tribes population. The proportion of the Scheduled Castes population and the proportion of Scheduled Tribes population have a positive statistically significant effect on the CDI but the proportion of urban population has a negative statistically significant effect. The same is the case with the index of deprivation in the housing domain, but in case of the index of deprivation in the material well-being domain, the effect of the proportion of the Scheduled Tribes population is statistically significant and positive, whereas the effect of the proportion of urban population and the proportion of Scheduled Castes is not statistically significant albeit in the expected direction. Lastly, in case of the index of deprivation in health and opportunity domains, the effect of all the three variables has not been found to be statistically significant, although the direction of the regression coefficient is in the expected direction. Obviously, the inter-district variation in urbanisation and proportion of Scheduled Castes and Scheduled Tribes population has an impact on the inter-district variation in the CDI and on the index of deprivation in the housing domain but not on the index of deprivation in health and opportunity domains. On the other hand, the inter-district variation in the index of deprivation in the maternal well-being domain is influenced by the inter-district variation in the proportion of Scheduled Tribes population but not by the inter-district variation in the proportion of urban population and the proportion of Scheduled Castes population.

Discussions

The pervasiveness of the deprivation faced by children below five years of age in Gujarat is clearly evident from the present analysis. It appears that the dividends of rapid economic progress that the state has witnessed in the recent past have not been translated into a positive social and economic environment that contributes to improved survival, growth and development of children below five years of age.

Among the four domains of child well-being that are the focus of the present analysis, deprivation in the opportunity domain appears to be alarming as it appears to be unaffected by such factors as degree of urbanisation and the proportion of Scheduled Castes and Scheduled Tribes population. In all but a few districts, deprivation in this domain remains very high irrespective of the degree of urbanisation and the proportion of Scheduled Castes and Scheduled Tribes population. All indicators used to characterise deprivation in this domain are related to women. This means that efforts to address child deprivation must incorporate efforts directed towards addressing the deprivation faced by reproductive age women. Improving the educational status of these women, increasing the age at marriage and ensuring that all pregnant women receive full antenatal care appear to be critical towards addressing child deprivation.

Deprivation in other domains of child well-being is generally district specific. Deprivation in housing domain is generally high in those districts where urbanisation is low and a large proportion of population is either Scheduled Castes or Scheduled Tribes. Similarly, deprivation in the health domain is generally high in those districts where urbanisation is low, although there are exceptions. For example, in the Narmada District, only about 10 per cent of the population lives in urban areas and more than 80 per cent of the population comprises Scheduled Tribes but deprivation in the health domain is quite low. Obviously, there are district-specific factors other than urbanisation and the share of Scheduled Castes and Scheduled Tribes in the population which have a dominating effect on inter-district variation in levels of deprivation in different domains of child well-being. One such factor may be the implementation efficiency of efforts directed towards child well-being. Very little is currently known about the efficiency of child well-being efforts being implemented in the state. In any case, it is obvious that a district-based approach is the most suited to address the challenge of deprivation of young children in the state which remains quite pervasive by all considerations.

Conclusions

This article presents a simple approach to measure and monitor deprivation faced by young children on the basis of the child deprivation index that has been deve-loped following the child well-being framework. The analysis suggests that deprivation faced by children of Gujarat is quite pervasive and has implications for the well-being of children below five years of age in the state. The situation appears to be alarming in six districts that constitute the north-eastern rim of the state. Moreover, in all but four districts, deprivation appears to be alarming in at least one of the four domains of child well-being.

The situation that prevails in Gujarat calls for a district-based approach of addressing child deprivation. This approach must be contextualised to district-specific dimensions of child deprivation so as to make it the most effective. It also appears that improving key indicators related to the women throughout the state is necessary to address the challenge of child deprivation. Building the capacity of women, after all, is the first and perhaps the most important step towards securing and sustaining child well-being. A policy level initiative needs to be taken in this regard.

Footnotes

District-level Database Used in the Analysis

State/District	Material Well-being Domain		Housing Domain			Health Domain			Opportunity Domain
State/District	Households below Poverty Line (Per cent)	Households Having None of the Specified Assets (Per cent)	Houses Not in Good Condition (Per cent)	Open Defecation (Per cent)	Children Not Fully Immunised (Per cent)	Children Not Breastfed Within 1 Hour of Birth (Per cent)	Children Not Exclusively Breastfed for First Six Months (Per cent)	Currently Married Women (15–49) Illiterate (Per cent)	Females Married before Reaching 18 Years of Age (Per cent)	Pregnant Women Not Received Full ANC (Per cent)
Kachchh	20.0	34.0	40.6	61.4	51.3	12.3	25.4	28.7	15.0	71.4
Banas Kantha	33.4	28.1	40.2	73.3	42.1	36.4	89.5	53.6	56.1	86.5
Patan	24.1	24.8	38.0	52.3	20.6	51.0	67.1	36.4	33.8	84
Mahesana	33.6	32.4	41.0	60.6	5.7	35.7	54.8	14.3	14.5	78.5
Sabar Kantha	56.1	18.4	32.2	43.9	37.0	59.9	74.1	12.9	37.1	71.9
Gandhinagar	18.6	16.1	33.1	38.9	49.6	14.9	64.9	26.8	15.5	72.5
Ahmedabad	40.8	8.0	27.3	13.3	17.7	16.7	60.7	42.2	7.8	65.7
Surendranagar	23.3	16.4	34.7	62.4	39.0	15.5	39.0	16.9	1.2	62.3
Rajkot	26.3	7.0	25.8	27.7	23.1	38.7	40.1	23.5	15.6	80.2
Jamnagar	30.6	10.4	28.1	39.7	37.6	11.8	58.2	35.6	4.4	57.8
Porbandar	10.5	13.5	29.4	38.7	23.4	14.0	22.1	30.6	7.3	85.5
Junagadh	18.7	12.1	31.3	38.4	19.2	22.3	17.6	31.9	6.9	76
Amerli	24.6	14.5	29.4	39.6	11.5	38.4	82.5	29.9	11.5	77.3
Bhavnagar	23	13.1	31.2	43.8	16.8	25.5	28.5	34.4	10.9	61.9
Anand	36.9	24.1	38.5	44.7	24.0	28.1	49.4	10.2	24.4	73.6
Kheda	29.9	16.2	25.0	35.0	32.7	3.5	58.8	22.3	31.4	72.9
Panchmahals	27.2	37.6	48.9	71.6	65.1	15.2	59.7	51.4	15.0	66.8
Dahod	47.5	44.8	57.7	86.1	19.1	57.4	70.5	43.0	28.0	74.0
Vadodara	33.8	17.8	33.6	36.7	8.1	24.7	53.8	32.5	16.4	78.9
Narmada	73.9	45.5	44.2	76.4	17.4	42.0	35.2	23.5	18.6	92.7
Bharuch	45.3	21.7	33.1	37.4	19.8	38.7	66.4	32.5	17.1	70.9
Surat	24.8	12.6	22.7	11.5	31.9	36.4	16.9	29.6	13.5	73.5
The Dangs	37.8	53.2	36.4	69.3	61.7	45.7	32.8	34.9	44.7	83.4
Navsari	38.3	16.5	31.2	40.0	33.5	29	45.1	19.6	13.5	52.9
Valsad	24.8	20.6	34.7	48.9	35.1	25.8	60.7	22.0	25.3	83.9
Tapi	51.1	36.2	36.7	70.5	36.5	15.4	51.8	38.5	28.1	47.8
Gujarat	32.0	18.7	32.7	40.4	30.6	28.9	53.7	30.1	19.9	74.1

Source: Government of India (2015).

References

Alkire

, & Roche

J.M.

(2012). Beyond headcount: Measures that reflect the breadth and components of child poverty. In Minujin

& Nandy

(Eds), Global child poverty and well-being: Measurement, concepts, policy and action (pp. 103–134). Bristol: The Policy Press.

Anand

, & Sen

(1997). Concepts of human development and poverty: A multidimensional perspective. Poverty and Human Development. Human Development Papers. New York: United Nations Development Programme.

Apablaza

, & Yalonetzky

(2011). Measuring the dynamics of multiple deprivations among children: The cases of Andhra Pradesh, Ethiopia, Peru and Vietnam. Young Lives Research in Progress. Oxford: University of Oxford.

Atkinson

A.B.

(2003). Multidimensional deprivation: Contrasting social welfare and counting approaches. Journal of Economic Inequality, 1, 51–65.

Bauer

, Davies

J.K.

, Pelikan

, Noack

, Broesskamp

, & Hill

(on behalf of the EUPHID Consortium). (2003). Advancing a theoretical model for public health and health promotion indicator development: Proposal from the EUPHID Consortium. European Journal of Public Health, 13(3), 107–113.

Ben-Arieh

, Hevener-Kaufman

, Bowers-Andrews

, Goerge

R.M.

, Joo-Lee

, & Aber

J.L.

(2001). Measuring and monitoring children’s well-being. Dordrecht, The Netherlands: Kluwer Academic Publishers.

Bradshaw

, Brook

, Huby

, Rhodes

, Sinclair

, Gibbs

, Noble

, McLennan

, & Wilkinson

(2009). Local index of child well-being. Summary Report. London, Communities and Local Government.

Brown

B.V.

(1997). Indicators of children’s well-being: A review of current indicators based on data from the federal statistical system. In Hauser

R.M.

, Brown

B.V.

& Prosser

W.R.

(Eds), Indicators of children’s well-being (pp. 3–15). New York: Russell Sage Foundation.

Carroll

(2002). The well-being of children: Four papers exploring conceptual, ethical and measurement Issues. Dublin: Irish Youth Foundation.

10.

Chaurasia

A.R.

(2010). Spatial and social class variations in child deprivation in Madhya Pradesh, India. Indian Journal of Human Development, 4(2), 369–390.

11.

Child Trends. (1997). Building common definitions and moving towards common constructs. The First National-Level Meeting of the Planning Phase of the Project on State-Level Child Outcomes, Washington, DC, 6–7 November 1996. Washington, DC, Child Trends. Retrieved 20 January 2016, from http://www.childtrends.org

12.

Child Trends. (2003). Child trends databank. Washington, DC, Child Trends. Retrieved 20 January 2016, from http://www.childtrendsdatabank.org/

13.

de Neubourg

, Chzhen

, Main

, Martorano

, & Menchini

(2012). Child deprivation, multidimensional poverty and monetary poverty in Europe. Florence, UNICEF Innocenti Research Centre. Innocenti Working Paper No. 2012-2.

14.

Dreze

, & Khera

(2010). Regional patterns of human and child deprivation in India. Economic & Political Weekly, 47(39), 42–49.

15.

Foundation for Child Development. (2013). National child and youth well-being index. New York: Foundation for Child Development.

16.

Gordon

, Nandy

, Pantazis

, Pemberton

, & Townsend

(2003a). The distribution of child poverty in the developing world. Centre for International Poverty Research, University of Bristol.

17.

Gordon

, Nandy

, Pantazis

, Pemberton

, & Townsend

(2003b). Child poverty in the developing world. Policy Press: Bristol.

18.

Government of India. (1974). National Policy on Children. New Delhi: Department of Social Welfare.

19.

Government of India. (2007). Child Protection in the Eleventh Five-Year Plan (2007–2012). New Delhi: Ministry of Women and Child Development.

20.

Government of India. (2013). Sample Registration System Annual Report 2013. New Delhi: Office of the Registrar General and Census Commissioner.

21.

Government of India. (2015). India: Rapid Survey of Children 2013-14. New Delhi: Ministry of Women and Child Development.

22.

Grinspun

(2004). Editorial. In Grinspun

(Ed.), In focus: Children and poverty. Brasilia: United Nations Development Programme, International Poverty Centre.

23.

GSIDBS. (2012). Evaluation of MCH Services in Gujarat: District Level Rapid Household Survey (DLRHS-1). Gandhinagar: Gujarat Social Infrastructure Development Board Society.

24.

Hanafin

, & Brooks

A.M.

(2005). Report of the development of a national set of child well-being indicators in Ireland. Dublin: The National Children’s Office.

25.

Hauser

R.M.

, Brown

B.V.

, & Prosser

W.R.

(Eds) (1997). Indicators of children’s well-being. New York: Russell Sage Foundation.

26.

Kannamori

M.J.

, & Pullum

(2013). Indicators of child deprivation in sub-Saharan Africa: Levels and trends from demographic and health surveys. Claverton, ICF International. DHS Comparative Report No. 32.

27.

Kushwah

S.S.

, & Prasad

(2009). Child survival and health: Policies and Programmes. In Ranjan

Aalok

(Ed.), Child survival and health: India 2009. Bhopal: Shyam Institute and New Delhi: Indo-American Books.

28.

Lister

(2004). Poverty. Cambridge: Polity Press.

29.

Maryland Partnership for Children, Youth and Families. (2002). Maryland’s results for well-being. Baltimore, Maryland: Maryland Partnership for Children, Youth and Families.

30.

Mickelwright

(2001). Should UK government measure poverty and social exclusion with a composite index? In CASE 2001. Indicators of Progress: A Discussion of Approaches to Monitor Government’s Strategy to Tackle Poverty and Social Exclusion. CASE Report No. 13. London: London School of Economics.

31.

Mickelwright

, & Stewart

(1999). Is the well-being of children converging in the European Union? York: University of York. Social Policy Research Unit.

32.

Moore

K.A.

(1995). New social indicators of child well-being. Vienna: European Centre for Social Welfare Policy and Research.

33.

Moore

K.A.

(1997). Criteria for indicators of child well-being. Cited in Hauser, Brown, & Prosser (1997).

34.

Moore

K.A.

(1999). Indicators of child and family well-being: The good, the bad and the ugly. A presentation to National Institutes of Health Office of Behavioural and Social Sciences Research.

35.

Minujin

, Delamonica

, Gonzalez

, & Davidziuk

(2005). Children living in poverty: A review of child poverty definitions, measurements and policies. Presented at UNICEF Conference Children and Poverty: Global Context, Local Solutions. New York: UNICEF, New School University.

36.

Minujin

, Delamonica

, Davidziuk

, & Gonzales

E.D.

(2006). The definition of child poverty: A discussion of concepts and measurements. Environment and Urbanization, 18(2), 481–500.

37.

Notten

, & de Neubourg

(2007). Poverty in Europe and the USA: Exchanging official measurement methods. Maastricht Graduate School of Governance Working Paper 2007/005.

38.

Notten

, & de Neubourg

(2011). Monitoring absolute and relative poverty; ‘not enough’ is not the same as ‘much less’. Review of Income and Wealth, 57(2), 247–269.

39.

Nozick

(1974). Anarchy, state and utopia. Oxford: Blackwell.

40.

Nussbaum

M.C.

(2000). Women and human development: The capabilities approach. Cambridge: Cambridge University Press.

41.

Nussbaum

M.C.

, & Sen

(Eds) (1993). The quality of life. Oxford: The Clarendon Press.

42.

Nyangara

, Nyberg

, Murphy

, & O’Donell

(2008). Child status index: A tool for assessing the well-being of orphans and vulnerable children – Manual. Chapel Hill: University of North Carolina, Carolina Population Centre. Measure Evaluation.

43.

Pollard

E.L.

, & Lee

(2002). Child well-being: A systematic review of the literature. Social Indicators Research, 61(1), 59–78.

44.

Reidy

, & Winje

(2002). Youth indicators in use in states as of September 2002. Chicago: University of Chicago. Chapin Hall Centre for Children.

45.

Ravallian

(2010a). Mashup indices of development. Washington, DC, The World Bank. Policy Research, Working Paper 5432.

46.

Ravallian

(2010b). On multidimensional indices of poverty. Washington, DC, The World Bank. Policy Research, Working Paper 5580.

47.

Roche

J.M.

(2013). Monitoring progress in child poverty reduction: Methodological insights and illustration to the case study of Bangladesh. Oxford, University of Oxford: Oxford Poverty and Human Development Initiative (OPHI). Working Paper No. 57.

48.

Ryan

R.M.

, & Deci

E.L.

(2001). On happiness and human potentials: A review of research on hedonic and eudaimonic well-being. Annual Review of Psychology, 52, 141–166.

49.

Sen

(1985). Commodities and capabilities. Oxford: Oxford University Press.

50.

Sen

(2000). Social exclusion: Concept, application and scrutiny. Manila: Asian Development Bank.

51.

United Nations Children’s Fund. (1989). United Nations convention for the rights of the child. New York: United Nations.

52.

United Nations Children’s Fund. (2014). Situation analysis of child poverty and deprivation in Uganda. Uganda: UNICEF.

53.

United Nations. (2000). Millennium development goals. New York: United Nations.

54.

Waddington

(2004). Linking economic policy to childhood poverty: A review of the evidence on growth, trade reform and macroeconomic policy. CHIP Report No. 7.

55.

White

, Leavy

, & Masters

(2002). Comparative perspectives on child poverty: A review of poverty measures. Journal of Human Development, 4(3), 379–396.

56.

Yang

(2013). The case of being automatic: Introducing the Automatic Linear Modeling (LINEAR) procedure in SPSS Statistics. Multivariate Linear Viewpoints, 39(2), 27–37.