Abstract
Given that population sex ratio and sex differentials in mortality are the two commonly employed indicators of gender-discrimination/anti-female biases in intra-household distribution of resources, this article performs two major tasks: (a) it reviews existing estimates of sex differentials in mortality among tribal (ST) children (including neonates) up to 6 years of age (in comparison with those of their counterparts of SC and other communities) and (b) it presents and analyses the author’s own estimates of sex-specific infant and child mortality rates separately for ST, SC and other social categories at all-India level by utilising three consecutive National Family Health Survey data. It is shown that the picture of tribal gender relations, while having traditionally been more balanced and equitable, has of late been one of trailing or having caught up with long-standing, albeit unenviable, mainstream anti-female features, resulting in an annihilation of a potentially rewarding opportunity for mainstream to learn from pristine tribal culture ideologically supportive of gender balance and equality.
Background: Issues and Objectives
In our earlier demographic study on India’s tribes as a whole (Maharatna, 2000, 2005), the following two distinct propositions, inter alias, relating to India’s tribal mortality were established—of course as much conclusively as could be possible—on the basis of the then existing (albeit scattered) secondary (official and individual) information:
Historically (at least up and till the early to mid-1980s), Indian tribes on the whole—with perhaps some regional variations of degree—have had not only a relative mortality advantage both in infancy, childhood and adulthood ages most pronouncedly vis-à-vis scheduled caste (SC) peoples (i.e., non-tribal low castes), but they have also evinced a somewhat distinct indication of a female advantage in mortality, in general, and in infant childhood mortality, in particular. But by the early to mid-1990s, there have been somewhat strong indications of tendencies towards reversal of both of the above-noted features of tribal mortality.
In a more recent study (Maharatna, 2015), we have reviewed new evidence made available since the publication of our earlier study, thanks to the lately conducted large-scale sample surveys [National Family Health Survey (NFHS) 2 and 3] and also other related data (e.g., NCER’s Human Development Survey), to check the nature and direction in more recent comparative trends of tribal mortality indicators. Notably, our recent review of these new evidence has reaffirmed our above-mentioned earlier findings and has indeed revealed a further accentuation of the tendency towards a reversal of tribal mortality advantage vis-à-vis both SC and others, although historical advantage in tribal mortality (especially infancy and childhood) vis-à-vis SC group as well as mortality advantage of tribal female infants and children vis-à-vis male counterparts appear to have remained somewhat valid (at the all-India level) even during the late 1990s (see Maity, 2016, Tables 4 and A2; Sinha, 2015).
While all this points to, inter alias, a steady relative worsening of tribal mortality and, by implication, of tribal health, nutrition, health-care facilities and their accessibility, this finding is in complete consonance with a few other recent researches depicting a distinctly greater (relative) tribal deprivation in terms of poverty, undernutrition, healthcare and child mortality as compared to SC and, of course, non-SC/ST social groups (Bhattacharya & Chikwama, 2012; Das, Hall, Kapoor, & Nikitin, 2012; Maity, 2016; The World Bank, 2011). Such a clear adverse trend of tribal mortality is well corroborated by findings of district-level regression analyses (based on various census data) that essentially seek to identify statistically significant correlates of child mortality, sex ratio (female–male ratio for total population or/and among children aged 1–6 years) or ratio of female to male child mortality—the latter being widely taken as a proxy for gender relations/bias and/or of relative female advantage/disadvantage/discrimination in infancy and childhood ages (e.g., Arokiasamy, 2004; Chaudhuri & Jha, 2011).
For example, in a careful econometric exercise with district-level census data for 1981, Murthi, Guio, and Drèze (1995) reported a highly significant district-level inverse (causal) relationship of proportionate share of ST population not only with under-five child mortality (i.e., probability that a child would die before it reaches its fifth birthday) (−0.6 at 5% significance level) but also with the extent of female disadvantage in childhood mortality (−0.01 at 5% significance level). [Gender disadvantage in child mortality is measured as 100 (Q5F – Q5M)/Q5F, where Q5F is mortality among female children and Q5M is mortality among male children.] Interestingly, a subsequent econometric regression analysis, based inter alias on district-level census information both for 1981 and 1991, by Drèze and Murthi (2001, Table 6), showed a positive—albeit very weak or statistically insignificant (causal)—relationship between relative concentration of tribal population in a district and under-five child mortality, pointing to a fairly dramatic disappearance of historic mortality advantage of tribal children especially vis-à-vis their SC counterparts. Indeed, this phenomenon of rapid reversal of historic tribal advantage in child mortality has been tellingly reaffirmed in a more recent district-level regression analysis involving census information from 1981 to 2001 (see Table 4 in Bhattacharya & Chikwama, 2012).
While these studies have mostly relied on census-based ‘indirect’ mortality estimates based on census information (rather than on vital registration or birth history data), the econometric studies using NFHS-based birth history data separately for ST and SC are extremely rare, and there is hardly any study on comparative evaluation of trends in gender differentials in infant and child mortality among ST, SC and others. However, a very recent study, which estimates hazard ratio (HR) of infant and childhood mortality for the ST group in rural India with NFHS 1, 2, 3 information, reports that although the risk of male infant mortality is higher than that of female infant mortality in tribal communities from 1992 to 2006 taken together, they do evince a male mortality advantage (by approximately 30 per cent point) vis-à-vis females in childhood years during this period (Sahu, Nair, Singh, Gulati, & Pandey, 2015). This finding, of course, provides (at least partial) clue to what was aptly posited by a section of media as a burning national issue, namely a drastic (indeed ‘the largest’ quantum) decline in child sex ratio (female/male children aged 1–6 years; see Table 1) among aggregate ST population in the latest census of 2011. 1 This is because sex differential in mortality or child mortality in particular—which is one of the major determinants of sex ratio—is widely known to be a reflection of societal gender discrimination and hence of unequal gender relations. However, there is pretty strong evidence that still shows a greater female autonomy, insubstantial son preference and/or lesser incidence of pre-natal sex determination/abortion among STs vis-à-vis SCs as late as 2006 (e.g., Arokiasamy & Goli, 2012). Although there has been a common general pattern of movements in sex ratios (for both children and total population) among all major social groups during entire post-independence period, this can hardly distract us from the fact that ST group happens to have still much higher (indeed the highest) female–male ratios among both children and total populations than those in other major social categories. Indeed, as is clearly discernible from Table 1,while child sex ratio (F/M) has been declining since long (at least since 1951) for all-India population and for SC and ST groups as well since 1981 (separate SC and ST child sex ratios are unavailable before 1981), an increasing magnitude/pace of its declines among ST children with passage of time over recent period appears quite striking, especially in combination with diminishing absolute quantum of decline among other social groups, particularly the SC. This could be highly plausible reflection of a lagged percolation of mainstream anti-female patriarchal culture, norms and practices into tribal world. This, if true, is indicative of far more serious sociocultural implications and ramifications of ongoing assimilation of tribal culture into mainstream than are commonly thought (Maharatna, 2011a).
Trends in Sex Ratio (females per 1,000 males), ST, SC, All Populations, India 1951–2011
With this backdrop, the primary objective of the present article, based on newly available evidence from various sources, is twofold: (a) to examine or perhaps re-examine recent trends in India’s tribal mortality vis-a-vis SC and others and (b) to investigate whether or not a tendency towards reversal of the historically observed female advantage in tribal mortality (which is aptly taken as a reflection of a more balanced/equal gender relations in a society; see Maharatna, 2005, 2011a and references cited therein) is of late getting further hastened and even completed. If the answer is found to be ‘yes’, which, as we will discuss, is the case, this should constitute one extremely sad commentary on an unenviably negative aspect—especially from a broad civilisational standpoint—of India’s ongoing (and indeed long-lasting) sociocultural assimilation of, or sometimes even domination over, tribes and tribal sociocultural traditions into mainstream (or Hindu) sociocultural patterns, apart from a stark inequity and discrimination against tribal people in terms of their access to modern healthcare and related facilities and provisions.
Levels and Trends in Infant and Childhood Mortality by Social Group: Growing Tribal Disadvantage
We better begin by emphasising that up and till the late 1980s, there has been little empirical/statistical evidence showing any higher mortality level among Indian tribes (both at individual and aggregative levels particularly in infancy and childhood ages) than not only those of SC counterparts but also even those of total population (see Maharatna, 2005, especially chapter 4). Ironically enough, over about last three decades, there has been a plethora of studies/evidence that posit an increasingly higher tribal mortality levels vis-à-vis those for SC and, of course, ‘other’ social groups (ibid.). Table 2 presents temporal trends in relative mortality levels of ST, SC and non-ST/SC groups (measured as ratios to each other) between 1980s and the mid-2000s. These ratios are based on respective estimates obtained from three rounds of NFHS birth history data. The estimates are available in National Institute of Medical Statistics (NIMS), Indian Council for Medical Research (ICMR), and UNICEF (2012). As can be seen, from 1981 to 1986, all indicators of mortality starting right from neonates up to childhood ages posit tribal social category (ST) as having had a mortality advantage vis-à-vis SC and even others in the particular case of neonatal mortality. (Estimated rates or levels of respective mortality are not shown here because they are all given in Table 4.3[a] in NIMS et al. [2012, p. 45.]) In fact, this tribal advantage in infant and child mortality over SC group, while already waning by the 1980s, continued to be manifest even during the latter half of the 1980s (except for child mortality between ages 1 and 5 years). This (plausibly) implies that the intrinsic superiority of traditional infant and child care practices/cultural norms of tribes could continue to show up in respective mortality differentials between ST and SC so long as very limited overall reach and standards of modern public health facilities and very low overall economic footing kept both the groups somewhat equally vulnerable and deprived vis-à-vis others.
Differential Improvement in Mortality Among ST, SC and Other, India 1981–2006
It is true that social differences in mortality movements—especially in infancy and childhood ages—should normally correspond inter alias to differential fluctuations in their economic circumstances, vulnerability to death and accessibility to modern healthcare facilities. But in an overriding context of very rudimentary modern public health and child care facilities and mass poverty, the neonatal, post-neonatal and perhaps even infant mortality levels are relatively more substantially shaped, than survival chances of children above the age of infancy, by postnatal-care-related sociocultural practices and norms including timing/duration and other features pertaining to breastfeeding (Maharatna, 2000, 2011a). This appears to have been reflected rather starkly in the evidence of mortality advantage among tribal children within infancy vis-à-vis non-tribal counterparts up and till the early 1990s. But when relative economic adversities and healthcare deprivations among tribal people (vis-à-vis SC) have continued unabated (or even increasing) over a fairly long period of time, as it has, indeed, been the case, it is little wonder that this would eventually neutralise the socioculturally (and historically) shaped mortality advantage of tribal children even within infancy as compared to their non-tribal counterparts.
Indeed, it was during the second NFHS survey representing the period 1993–1998 that IMR of ST group relative to ‘others’ surpassed that of SC community (see Table 2). Herein arguably lies the clue as to why tribal infants and neonates could continue showing an edge—albeit of declining degree with time—over their non-tribal counterparts up to the early 1990s, while tribal children aged above 1 year appear to have lost their mortality advantage vis-à-vis non-tribal counterparts a little earlier, that is, by the 1980s (since child mortality estimates from NFHS-1 data refer to the period 1982–1992). Indeed, tribal child mortality relative to both SC and others has been rising steadily since the latter half of the 1980s (Table 2). A recent World Bank-sponsored study, which has estimated separately odds ratios of dying among children starting from birth up to 4 years of age on the basis of NFHS-3 data (that corresponds to 1999–2005 period), has reported the highest (about 2.4 times, see Table 2) and statistically highly significant, odds ratio for tribal children aged 1–4 years vis-à-vis non-ST/SC peers. Since this was found to be so even after controlling for other factors including household welfare (i.e., some indicators of wealth/income), the authors conclude that there exists a specific (extra-economic) tribal effect on mortality of tribal children beyond infancy vis-à-vis their non-ST peers, although they stop short of providing or indicating its exact notion, clue or factors behind this so-called ‘tribal effect’ (Das, Hall, Kapoor, & Nikitin, 2010, pp. 4–10). As, according to the authors’ interpretation of the results, ‘the problem of high child mortality is explained by concentration in tribal groups’, they point to the importance of the need for going beyond ‘addressing poverty alone’ or more particularly, for specific government interventions designed to reach out specifically to tribal pockets. Another recent district-level decomposition analysis based on census information from 1981 to 2001 has reported that while members of scheduled tribe communities had no independent or intrinsic effect or contribution to the district-level inequality in under-five mortality in 1981, tribes are generally found of late (i.e., from 1991 to 2001) to have been increasingly concentrated typically in less-developed districts (Bhattacharya & Ckikwama, 2012). In any case, all this, overall, is supportive of, and consistent with, our emphasis on a contemporary phenomenon of transition from a regime of relative tribal mortality advantage in infancy and childhood ages to its complete reversal particularly in comparison with their SC and other counterparts.
As can also be seen from Table 2, magnitudes of improvements of mortality in infancy and childhood ages under 5 years during the 1980s, 1990s and 2000s have been the largest for other (non-ST/SC) population group and the least for ST category, with the position of SC falling in-between. Since the 1980s, there has been a steady worsening of relative mortality indicators of ST children vis-à-vis those belonging to SC and other groups. This has been the case not only with infants and children below 5 years but for adolescents too. For example, in terms of odds ratios of mortality of children aged 2–5 years and 6–18 years among three social groups, the SC children have had higher mortality than that of ‘other’, the ‘reference’ category, by about 15 and 35 per cent, respectively, while for ST counterparts, the respective figures have been as much as 71 and 94 per cent during the 1990s (see Table 2 and Subramanian et al., 2006). All this reaffirms the already-noted irony, namely while tribal infants and children historically continued to appear least vulnerable to death particularly in comparison with their SC counterparts even up to the early 1980s, just over the following couple of decades, the former have turned out to be most vulnerable to death vis-à-vis SC and, of course, ‘other’ social groups. This, in turn, clearly reflects tribal people’s growing relative disadvantage/inaccessibility in reaping mortality and health benefits that have emanated from the country’s more recent (post-liberalisation) development programmes, in general, and reproductive and child health initiatives, in particular. This key demographic finding corroborates well with the overall worsening of the relative plight of India’s tribal population in myriad ways and means that are often being eloquently portrayed in the sociological, anthropological or political literature (Das et al., 2010, 2012; Guha, 2007; The World Bank, 2011, among others). What, thus, emerges from Table 2 is not only (by-now perhaps a litany) of tribal people lagging behind SC communities (and of course the rest as well) in receiving benefits of modern medicines and child healthcare improvements—however limited—over the recent decades but it also powerfully drives home a distinct and indeed a disturbing irony—often unnoticed and/or overlooked by official and other influential (and a section of academic) circles—that it was the tribes who have historically been quite ahead of (at least) SC social group on all these child mortality counts even up to a few decades after India’s independence (Maharatna, 2005, 2011a). While all this signifies an acute material/physical deprivation/discrimination being meted out to Indian tribal societies on the whole, we now turn to the phenomenon of tribal sociocultural subjugation (or assimilation or rather degradation) by the majoritarian mainstream patterns and norms with special reference to recent trends in gender differentials in infant and childhood mortality among ST as compared to SC and others.
Recent Trends in Sex Differential in Mortality by Social Group: How Complete Is the Reversal of Gender Equality Among Tribes?
In Table 3, we present fairly comprehensive information pertaining to recent trends in sex differentials in children’s mortality beginning right from birth up to the fifth birthday separately for three social groups considered and compared here. The source of information is overwhelmingly the NFHS which has been conducted under the larger rubric of DHS. We have used data from the first three consecutive NFHS surveys starting from 1992 to 2006 [the latest NFHS-4 was conducted in 2015–2016 of which detailed data are yet to be made available/accessible]. Although there have been quite a few empirical studies on the major demographic aspects of ST and SC as separate groups on the basis of NFHS data, the specific issue of the trends in gender differential in mortality by social group over recent decades has remained almost entirely unexplored. 2 While such an analysis at the state level is likely to be impaired because of thinness of sample size on demographic events such as births and deaths by sex and by social group at least in many states, this reason can hardly deter an all-India level analysis. In fact, using NFHS data from first three consecutive surveys, we have estimated mortality levels among children starting with neonates by gender separately for ST, SC and non-ST/SC categories at all-India level (Table 3).
First, while neonatal mortality among girls is less than that of boys in almost all social groups (except SC in 1998–1999), a suggestion of temporal weakening of this female advantage among ST and SC is quite discernible. Similarly, a growing female disadvantage in the post-neonatal mortality over the recent decades is particularly noticeable for both ST and general population. Given very limited (or perhaps even absent) available evidence pertaining to social, behavioural and familial determinants of neonatal and post-neonatal mortality, it is not readily clear whether this relative worsening of survival chances among female neonates and post-neonates reflects a growing anti-female attitude and associated neglect/discrimination within households. In case of IMR, however, there has been a female advantage—albeit of varying degrees across social groups—around the early 1990s, with its absolute magnitude being the largest among the ST group but only until the mid-2000s when female advantage got reversed among nearly all groups including ST. Indeed, a steady decline in the HR of male IMR (vis-à-vis females) among tribes as a whole has been found across first three NFHS surveys (see Table 3). However, in sum, during the entire period of infancy, which is the prime time for breastfeeding as a major source of child nourishment and survival chances, the relative female advantage in mortality, though waning over time, has continued to distinguish overall tribal community from the rest of the Indian society only until very recently. That this long-standing female advantage in tribal IMR has been reversed of late could well be suggestive of a growing pace of infusion of mainstream anti-female sociocultural attributes and practices into the former.
Trends in Infant and Childhood Mortality and their Sex Differentials, Scheduled Tribes (ST), Scheduled Castes (SC) and Other, India 1981–2006
All figures/estimates, if not otherwise stated, are based on NFHS data.
Figures within ( ) brackets are respective female to male ratios estimated on the basis of Census 2001 data (Government of India, 2010, New Delhi). Figures with [ ] brackets are ratios of female to male mortality based on vital registration data: (a) for the 1970s, (b) for the 1980s, (c) for the 1990s and (d) for the 2000s (United Nations, 2011). Figures with { } brackets are respective female to male ratios of mortality indicators estimated by using 1991 census data and indirect techniques (Mohanachandran & Rajan, 2001). The category ‘other’ is exclusive of both SC, ST and Other Backward Communities (OBC) in case of NFHS II and III, while it is inclusive of OBC in NFHS-I. All mortality rates, except for those indicated otherwise, are calculated from the respective NFHS unit-level data files.
Neonatal mortality rate is the number of neonates dying before reaching 28 days of age per 1,000 live births in a given year.
Post-neonatal mortality rate is the number of resident newborns dying between 28 and 364 days of age in a specified geographic area (country, state, county, etc.) divided by the number of resident live births for the same geographic area (for a specified time period, usually a calendar year) and multiplied by 1,000.
In contrast with lower mortality rates observed particularly among tribal female neonates and infants, especially until the 2000s, one can see a distinct turnaround among children aged 1–4 years, with female child mortality exceeding that of males among all social groups including ST and for all NFHS surveys since the early 1990s. Notably, however, the magnitude of excess mortality of female children vis-à-vis male counterparts has been the least among tribes throughout except for NFHS-3 (2005–2006) when tribal female–male ratio of child mortality has exceeded—albeit slightly—that of non-ST/SC group (see Table 3). It thus seems quite clear that there still exist considerable remnants of age-old tribal cultural traditions characterised by much higher degree of gender equality and hence much lesser familial/social discrimination/neglect against tribal girls/daughters than that among mainstream and other non-tribal populations (see Maharatna, 2000 for relevant evidence). But the fact of female–male ratio of child mortality for non-SC/ST people turning lower than even that of ST in the NFHS-3 (2005–2006) reflects the effects of former’s lessened degree of familial discrimination against those daughters/girls who were decidedly welcome to be born on some consideration or other (e.g., lower birth order female foetus) in the face of widespread practice of pre-natal sex-selective abortion targeting often higher birth order female foetuses (see Anukriti, Bhalotra, & Tam, 2015 for detailed evidence of this phenomenon). This aforesaid argument, of course, rests on a highly plausible assumption that the incidence of pre-natal sex-selective abortion is negligible or far less among tribal communities.
The overall conclusion of a lower degree of anti-female discrimination in tribal societies appears to be reaffirmed even more strongly by estimated tribal female–male ratios of under-five mortality being far less than unity at least for the first two NFHS surveys respectively, while it has exceeded unity rather noticeably in NFHS-3 conducted in 2005–2006, reflecting a complete reversal of tribal female advantage in under-five mortality by the mid-2000s. Notably for the SC group, the degree of female disadvantage in under-five mortality has increased over the entire period covered by three consecutive NFHS surveys. It is also of interest that the extent of anti-female gender differential in estimated child mortality (1–4 years) rates is found noticeably larger than that for under-five (0–4 years) mortality estimates (i.e., inclusive of infant deaths) for non-tribal social groups, while tribal female advantage in under-five mortality has remained prominent up and till the NFHS-3 survey in 2005–2006. All this seems consistent with our previous observation that adverse health/mortality effects of conscious familial gender bias, neglect and discrimination are likely to be relatively less pronounced among infants (as they survive predominantly on breast milk, which help to keep infants less exposed to contaminated water) than among children aged between 1 and 4 years (when food and medical attention take on relatively greater importance in survival chances). For instance, a recent careful study, which utilises NFHS-3 data, has not only shown that the estimated gender gap in breastfeeding duration accounts for only 14 per cent of excess female child mortality (deaths between the ages of 1 and 5 years) in India, but it has also revealed that instead of the oft-alleged reason of ‘valuing girls’ health less than boys’, ‘excess female mortality arises because subsequent fertility decisions, by being intertwined with breastfeeding decisions, have unintended health consequences’ (Jayachandran & Kuziemko, 2009, p. 30). (As a mother is biologically protected against conception during almost the entire duration of breastfeeding, the shortening of breastfeeding duration in case of ‘unwanted’ female child paves the possibility of hastening of the next child birth.)
In any case, nutrition, health and mortality of children aged between 1 and 5 years are not as crucially linked to breastfeeding practices, its intensity or duration as they normally are during infancy. This clearly leaves a greater scope and manoeuvrability for conscious gender-specific discrimination (e.g., in terms of food distribution and medical attention) in childhood years than in a child’s infancy. Thus, comparative figures of child mortality and under-five mortality from 1992 to 2006 (as presented in Table 3) appear amply suggestive (albeit indirectly) that the mainstream Hindu patriarchal features, values and sociocultural norms began making a perceptible dent on historically/traditionally far egalitarian tribal sociocultural milieu not really recently, but much earlier (e.g., during the 1970s and the 1980s) than the early 2000s as is apparently implied in the trends in neonatal and post-neonatal mortality differentials between sexes across social groups. This, of course, lends even stronger reaffirmation of what had been indicated in our earlier studies, namely pervading anti-female influences of mainstream sociocultural milieu on tribal communities on the whole (Maharatna, 2000, 2005). Interestingly, this revelation in our earlier studies happens to be also consistent with one of the key findings of a recent rigorous econometric study with the all-India household level data from the first three consecutive NFHSs starting from 1991 to 1992, which shows that increased prosperity and education over time have accentuated gender bias (proxied by number of girls per thousand of boys aged 0–6 years) across the board, that is, among general, SC, ST and Muslim population groups (Chaudhuri & Jha, 2011).
Gender Differentials in Undernutrition, Vaccination and Medical Attention: Is Gender Bias Fast Penetrating Tribal Society Too?
We now examine whether the picture sketched above of anti-female movement of sex ratios backed by somewhat concomitant increases in anti-female differential in infant and child mortality among Indian tribes over past several decades is matched or supported by available indicators/evidence of commensurate changes in gender biases in other proximate spheres like nutrition/undernutrition and medical attention/vaccination received.
Somewhat strangely, studies utilising information of NFHS surveys for estimating sex differentials in undernutrition among three broad social groups considered here are extremely rare or virtually non-existent. But there exist a few studies that posit very high (indeed the highest) child undernourishment among ST people. For example, one recent study, based on NFHS-2 (1998–1999), shows an extremely high degree of child (≤ 3 years) undernourishment, especially among ST and SC groups (55.9% and 53.5%, respectively) vis-à-vis ‘other’ (41.1%) (Das, 2008). Also, this study reports quite a high and statistically significant odds ratio of child undernutrition with respect to ST and SC categories even after controlling for various social and biological factors such as religion, place of residence, standard of living of household, work status of mother, education of mother, exposure of mother to television, mother’s body mass index (BMI) and anaemia status, ante-natal check-up visit by mother, whether iron folic tablets are given, tetanus injection, previous birth interval of the child, birth order of the child, age and sex of the child and morbidity. All this reflects what the author calls an ‘ethnicity effect’ on the nutrition of ST and SC children.
There are, however, a fairly large number of micro-level (anthropometric) studies on specific tribes, mostly in their adulthood ages, in diverse tribal pockets across India. A recent paper (Das & Bose, 2015) has provided a review/summary of findings of a variety of nutritional studies among adult tribes very often separately for males and females over last couple of decades. By drawing on this review article, we have constructed Tables 4(a) and (b), which present respectively the percentage of undernourished persons as well as the mean BMI level for both sexes separately among several individual tribal groups across diverse Indian locations/states. There is, of course, a variation in methodology and sample size adopted in these micro-level studies. However, since the focus of our present exercise is squarely on pattern of gender differentials among tribal communities, we do not need to be particularly sensitive to methodological or sampling variations that might have been entailed in these individual micro-level anthropometric studies, as long as the same method is applied by an individual study for both the sexes.
Percentage (%) of Undernourished Persons by Sex, Tribal Populations and Selected Indian State/Locations
As can be seen from Table 4(a), there is considerable variation in the incidence of tribal undernutrition both across locations and individual tribes. For instance, while Kora Mudi tribe in West Bengal shows about half of the adult males having undernutrition, the corresponding figure for Birhor, another tribe in the same state, is a little over 20 per cent. But since we are not sure about methodology and evaluative criterion adopted for estimating levels of undernutrition by these respective discrete micro-studies, it seems appropriate not to draw a firm conclusion pertaining to comparative evaluation between levels of undernutrition across tribes and locations. Apropos patterns of gender differential in adult undernutrition, the overall picture that emerges is the one which is starkly against females in case of most of the tribes. But a very limited number of empirical studies on age-sex composition of nutritional status and intra-family food distribution among select tribes and locations, conducted in the 1980s or even earlier, were, on the whole, ‘not suggestive of a definitive female discrimination among tribes, albeit with a few exceptions’ (Maharatna, 2005, pp. 170–171 and references cited). Of course, it is difficult to be completely doubtless about the extent to which glaring anti-female gender biases in adult tribal nutrition in majority cases over recent past considered here do reflect a fallout of deliberate/conscious female discrimination and neglect in terms of intra-household distribution of food and medical attention. However, this finding per se is not inconsistent (at least apparently) with the above-noted trends towards a complete reversal of (traditional/historical) gender equality in tribal childhood mortality. Interestingly, Table 4(b), which contains summary results of various micro-level studies, evinces rather little gender differential in BMI (an important indicator of nutritional status) among select tribal adult population groups in specific locations. This is not necessarily contradictory with the above-noted evidence of growing anti-female bias in childhood mortality among tribal communities over the recent past. Moreover, as the statistical biases, particularly selection bias, are likely to be present in these micro-level studies, an interpretation of these results calls for considerable caution.
Mean Body Mass Index [BMI = Weight in Kg/(Height in Meters)2], Tribal Populations, Select Tribes and States, India
A recent (draft) paper of Anindita Sinha (2017) presents percent distribution of children aged 0–4 years who have been covered under four immunisations (e.g., BCG, polio, DPT and measles) and also of those children who have received medical attention/treatment during infliction of two illnesses, namely diarrhoea and acute respiratory infection (ARI)/fever, by sex separately for each of the three social groups (ST, SC and other) over first three consecutive NFHS surveys. Apropos gender differential in child immunisation, although immunisation coverage for both sexes has been the lowest among the ST group, there is no or negligible evidence of discrimination/differential against tribal girls—a fact that contrasts rather sharply with the finding of some stark anti-female differentials in child immunisations among SC community and non-ST/SC social group (Table 5[a]). Sex differential in child immunisation rate, however, is not a very effective indicator of familial conscious gender discrimination, as child immunisation across India is—especially of late—carried out in (particularly) rural areas by public health personnel almost on a home delivery basis (and of course free of cost) or through a closer institutional surveillance/follow-up mechanism, thanks to Intensive Child Development Scheme and National Rural Health Mission.
Percentage Distribution of Children Aged 0–4 years Covered by Immunisations by Sex and Social Group, India, NFHS-1, 2, 3
Regarding the question of gender differential, if any, in medical attention during illness, Table 5(b) shows, first, that the proportion of children of both the sexes who have received medical treatment against the two illnesses has been distinctly lower overall among ST than in cases of SC and others, respectively during the entire period, confirming a scenario of tribes having much lesser access to healthcare facilities. Also, the proportion of girls treated in the ST community has somewhat declined from the early 1990s, whereas for boys the corresponding figures show a somewhat rising tendency. This could reflect a growing anti-female bias in medical attention reportedly received in case of two common health ailments among tribes vis-a-vis SC and others.
Percentage Distribution of Children Aged 0–4 years, Who Have Received Medical Treatment, by Illness, Sex and Social Group, India, NFHS-1, 2, 3
Concluding Discussion
India’s tribes, as per 2011 Census count, constitute little less than 10 per cent of total population and could hence be seen just as an ‘ethnic minority’ of the country. But at the same time, tribes are widely held to be the original inhabitants of a land which is known for boasting as one of the oldest civilisations of the world. From this latter standpoint, India’s tribal society, culture and demography carry a lot of historical significance and contemporary relevance from the perspective of shaping India’s future.
Tribes are viewed as distinct groups of people who remain outside mainstream civilisation. But in the Indian context, the task of defining tribe and its identity is particularly difficult. Unlike most other parts of the world, the rising civilisation in the Indian subcontinent neither eliminated nor quite absorbed these primitive inhabitants of the land, thereby leaving room for their continuity side by side the ‘mainstream’. Apart from the question of its origin, the nature of evolution through long-standing processes of transformation and assimilation within mainstream caste society constitutes a major facet of Indian tribal identity.
How far and/or whether India’s tribes are indigenous and autochthonous people of the land has been a subject of much academic debate (e.g., Béteille, 1998; Guha, 1999, pp. 1–9; Xaxa, 1999; Xaxa, 2003, pp. 377–379 among others). As the mainstream narrative goes, tribes had long been settled on the plains and river valleys before the Aryan invasion (see Ray, 1972: especially 11–15). Indo-Aryan speaking peoples with a superior social organisation and techno-economy are believed to have forced these indigenous peoples to move bit by bit to more and more inaccessible region of forests, hills and mountain slopes. Then followed the prolonged period of rise and spread of Hindu civilisation, in which caste-based social organisation eventually became its nucleus, while ‘tribes’ continued living in their ‘natural setting’ even very close to such renowned ancient and medieval centres of civilisation as Gaya, Ujjain and Maduari well into present days (Béteille, 1986, p. 309; Mandelbaum, 1970, p. 585).
Indeed, the relationship between Hindu and tribal societies is somewhat paradoxical. For instance, although a traditional tribal society is far more homogenous and far less stratified than a typical Hindu society, both these societies share certain homology in their strikingly similar ways of perpetuating respective collective identities. In Andre Béteille’s (1986, p. 311) words, ‘[i]t is no accident that observers down the ages have so persistently mistaken castes for tribes, and tribes for castes’. This signifies interpenetration, rather than just unidirectional adaptation of one by the other. As D. D. Kosambi (quoted in Béteille, 1986, p. 312) writes, ‘[t]he entire course of Indian History shows tribal elements being fused into a general society’. Similarly, tribal endogamy reflects, arguably, the influence of Hindu civilisation on tribal societies.
On a longstanding (and slow) process of assimilation or absorption of diverse tribes within mainstream society, one received explanation—so-called ‘Hindu method’—harps on a symbiotic, albeit unequal, relationship between tribes and the larger caste society (Bose, 1941, 1975). As the argument runs, when technologically less advanced tribal economic base became precarious due to population expansion or for other reasons, they or a section sought economic security through closer attachment to the wider Hindu society. In this view (based on fieldwork as well as historical and classical texts), the newly attached tribe was given the lowest position in Hindu caste hierarchy. There are, however, a few instances of some richer and more powerful sections of tribal groups laying claim to being Kshatriyas [warrior caste], though they originally had started their career as tribes (Srinivas, 1977, p. 227). All this said, the prolonged fusion of tribal elements could hardly prevent the collective identities from ‘outliving the conversion of tribe into caste’ (Béteille, 1986). For instance, what inter alias distinguishes India’s tribes as a whole—socioculturally —from the mainstream is the former’s long tradition of far greater degree of gender equality and female autonomy than in the latter (for a fuller discussion on this issue, see Maharatna, 2000, 2005, 2011a and references cited therein).
In fact, this tribal (traditional) characteristic of greater gender equality could offer an important clue to India’s marked regional differences in gender relations, women’s status and autonomy particularly between more starkly patriarchal ‘north’ and comparatively more gender-equal ‘south’ (see the pioneering paper by Dyson & Moore, 1983 and subsequent literature produced on this issue). In our earlier studies on India’s tribal demography, we have found reasons to hypothesise that the greater is the extent of historic tribal infusion and conversion into mainstream society and culture (as was historically the case in large parts of southern India), the more balanced and equal is the latter’s gender relations and the higher is its female status and autonomy (Maharatna, 2005). But over recent past, the Indian ‘south’ has been evidently and increasingly taking to the anti-female features of the Indian ‘north’—a trend which can hardly be called ‘healthy’ or ‘welcome’ from gender relation perspective (Maharatna, 2011b).
The above brief historical backdrop on India’s tribes should help us better appreciate the wider significance of key demographic findings of our present exercise. Findings of present statistical analysis reaffirm, in the light of newer available statistical evidence, that, first, the historic tribal (relative) advantage in infant and child mortality vis-à-vis SC counterparts has been completely reversed by now. This is testified also by overall tribes’ growing relative deprivations in modern medical child care benefits and their accessibility. Second, while a greater gender equality/balance in sociocultural spheres—one well-known marker of modern Western world—has also historically characterised India’s tribal world up and till recently, this globally admired feature, found historically among India’s aggregate tribes, has been almost entirely replaced by a majoritarian dominant—yet (admittedly) contemptible—anti-female sociocultural patterns and mores. The moot question, then, is whether we should let gender-equal sociocultural traditions of tribes get further supplanted by unenviable mainstream anti-female patriarchal patterns or we should learn and imbibe an admirable culture of gender equality that has traditionally characterised tribal India.
Footnotes
Acknowledgements
Author acknowledges valuable statistical assistance provided by P. Ramesh, Sumoni Mukherjee and Anindita Sinha without whose contribution and help in digging up NFHS data the paper could not have reached its present version. Thanks are also due to Professor B. R. Bhagat for his support and cooperation. Useful comments from anonymous referees and the Managing Editor are also thankfully acknowledged.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Funding
The author received no financial support for the research, authorship and/or publication of this article.