I S S U E B R I E F P O V E R T Y A N D S O C I A L E X C L U S I O N I N I N D I A Adivasis Maitreyi Bordia Das and Soumya Kapoor Mehta1 development status of the SCs and Adivasis is that while the former lived among but were segregated Tribal groups or Adivasis are considered to be the socially and ritually from the mainstream and from earliest inhabitants of India. The term Adivasi is upper caste groups, the latter were isolated physically, commonly translated as ‘indigenous people’ or ‘original and hence socially (Béteille, 1991). inhabitants’, and literally means ‘Adi or earliest time’, and ‘vasi or resident of’. The state and discourse in While India is widely considered a success story India, however, reject the term “indigenous peoples”, in terms of growth and poverty reduction, Adivasis as it is considered “divisive, undermining the unity in 2004–05 were 20 years behind the average. The of the Indian nation” (Ghurye in Chopra 1988). The poverty headcount index for Scheduled Tribes fell government instead recognizes most Adivasis under by 31 percent between 1983 and 2004–05. In com- the Constitutional term “Scheduled Tribes” derived parison, poverty fell by about 35 percent among the from a schedule in the Constitution Order of 1950. Scheduled Castes and by 40  percent at an all-India While the Order declared 212 tribes located in 14 states level (table 1). The relatively slower decline in pov- as Adivasis, the Government of India today identifies erty among Adivasis means that they are increasingly 533 tribes. To many therefore, any aggregate analysis concentrated in the poorest deciles. Comprising about of Adivasis is meaningless because it cannot capture 8 percent of India’s population, Adivasis accounted for the uniqueness that defines tribal groups. a fourth of the population living in the poorest wealth decile in 2005. More worryingly, in states with high Scheduled Tribes are often conflated with Scheduled tribal populations (more than 10 percent of the state’s Castes in the development literature, although they total population), Adivasi households exhibited pov- are completely different social categories. The for- erty rates that were higher than the rates across the mer do not strictly fall within the caste hierarchy, and nation as a whole in 2004–05. In the state of Orissa, have distinct (often considered non-Hindu) cultural for instance, almost 75 percent of Adivasi households and religious practices and social mores. Also, they do fell below the official poverty line. not face ritual exclusion, say in the form of untouch- ability, as do the Scheduled Castes or ‘Dalits’2. But These findings need to be nuanced considering the when exclusion is defined more broadly in terms of highly unequal results across states and by urban/ being “prevent(ed) . . . from entering or participating” rural residence. In states where the Adivasis are a or “being considered or accepted” (Encarta online majority (such as in the north-east) and where they edition), Adivasis fit squarely within the concep- have gained from education, their status is different tion of excluded people. The major difference in the from the status of Adivasis living as marginalized minorities in the interior reaches of the central or west- ern states. We also find intra-group variation in poverty for Adivasis over time. Growth Incidence Curves that 1 This brief is based on a chapter in the forthcoming volume, Poverty and Social Exclusion in India. It is not a formal publication we calculated in rural and urban areas indicate that of the World Bank. It is circulated to encourage thought and discus- expenditures among Adivasis grew more rapidly at the sion, and its use and citation should take this into account. Maitreyi higher end of the expenditure distribution than in the Bordia Das is Lead Specialist in the Social Development Department lower end. This was particularly true in urban areas, of the Sustainable Development Network in Washington DC. Soumya Kapoor Mehta is an independent consultant in New Delhi. and may in part be explained by particularly large 2 For the purposes of the graphs and the tables in this note, we income gains among those with access to and benefits use the term SC/ST as these are standard administrative/survey from reserved jobs. This result may also explain why categories. In the text though, we use the self-preferred terms i.e. Dalits and Adivasis (or tribals) interchangeably with SCs and STs poverty rates among STs in urban areas have fallen respectively. relatively quickly. 2 Issue Brief: Poverty and Social Exclusion in India  |  Adivasis Table 1: Poverty Rates: Adivasis are 20 years behind the average population National Family Heath Survey Population below the poverty line, % (NFHS) 2005/6. Location, social The analysis for the India group A. 1983 1993–94 B. 2004–05 % change (A, B) Poverty and Social Exclusion Rural Report (Das et al, 2010) has ST 63.9 50.2 44.7 –30 three clear findings. First, a SC 59.0 48.2 37.1 –37 disproportionately high number Others 40.8 31.2 22.7 –44 of child deaths are concentrated All 46.5 36.8 28.1 –40 among Adivasis, especially in Urban the 1–5 age group and in those ST 55.3 43.0 34.3 –38 states and districts where there is SC 55.8 50.9 40.9 –27 a high concentration of Adivasis. Others 39.9 29.4 22.7 –43 Any effort to reduce child moral- All 42.3 32.8 25.8 –39 ity in the aggregate will have to focus more squarely on lowering Total mortality among the Adivasis. ST 63.3 49.6 43.8 –31 Second, the gap in mortality SC 58.4 48.7 37.9 –35 between Adivasi children and Others 40.5 30.7 22.7 –44 the rest appears after the age of All 45.6 35.8 27.5 –40 one. In fact, before the age of Source: Staff estimates based on Schedule 1.0 of the respective NSS rounds and official poverty line; table 1.1 one, tribal children face more in report or less similar odds of dying as other children (figure  1). While Adivasis saw significant gains in health indi- However, these odds significantly reverse later. This cators, some of which improved at rates that were calls for a shift in attention from infant mortality or more rapid than the average, a stark marker of tribal in general under-five mortality to factors that cause a deprivation is excess child mortality. Under-five wedge between tribal children and the rest between mortality rates among tribal children remain startlingly the ages of one and five. Third, the analysis goes con- high (at about 96 deaths per 1,000 live births in 2005 trary to the conventional narrative of poverty being the compared with 74  among all children). An average primary factor driving differences between mortality Indian child has a 25 percent lower likelihood of dying outcomes, because even after controlling for wealth under the age of five compared to an Adivasi child. quintiles, the effect of being Adivasi stays robust. This In rural areas, where the majority of Adivasi children indicates that belonging to the Adivasi community live, they made up about 11 percent of all births but disadvantages children in other ways as well. 23 percent of all deaths in the five years preceding the Malnutrition among Indian children has remained stubborn to poverty reduction and programmatic Figure 1: Rural Adivasi Children: Lower Risk of Dying at Birth, intervention. But Adivasi children show even worse but Greater Risk by Age 5 levels of malnutrition. The gap between the Adivasi and non-Adivasi children in severe stunting and in 120 wasting3 appears within the first 10 months of birth 100 and persists—with some variation throughout early 80 childhood. The rise in severe wasting among Adivasi per 1000 60 children during the first 10 months of life is particu- 40 larly alarming. Microstudies on food insecurity among ■ Rural SC 20 ■ Rural ST Adivasi households provide a contextual picture of the ■ Rural OBC ■ Rural Other causes of chronic malnutrition. 0 al l nt ild ve ta at fa Ch rfi na on In eo de Ne -n Un 3 Stunting reflects long-term effects of malnutrition; while wast- st Po ing measures the current nutritional status of the child, i.e. his/her Source: Staff calculations based on 2005 NFHS data; figure 1.1 in report food intake immediately prior to the survey. Issue Brief: Poverty and Social Exclusion in India  |  Adivasis 3 Child deaths among Adivasis have led to frequent the positions of nurses or female health workers. As a public outcry. Public interest law suits have been result the positions either remain vacant or are filled filed on behalf of families that lost their children4, and by non-tribal, non-resident providers. state governments have been repeatedly directed by the courts to take remedial action. Governments have There have been a number of successful initiatives undoubtedly become more vigilant on this issue than in both the nongovernmental and the public sec- they were before, but solutions are still ad hoc and in tors to improve access to services in many tribal crisis-mode. For instance, during an emergency large areas.5 Studies from around the world suggest that numbers of medical personnel are deployed to the the adaptation of health services to the culture of vulnerable areas, but in normal times absenteeism of the beneficiaries (for instance, by recognizing the doctors is endemic in rural and especially tribal areas. importance of traditional medicine) leads to better Additionally, poor registration of births and deaths outcomes. Yet, the most well known efforts have often has meant frequent haggling over the real numbers been small, resource-intensive ones, making scaling- of deaths/children affected. up a challenge. Are tribal children more likely to become sick as Mortality outcomes for tribal children in India they grow older, and, so, more of them die? The need to be looked at in the light of larger changes answer is no. Contrary to expectations, tribal children experienced by Scheduled Tribes, especially in the do not get diarrhea or respiratory disease more often past two or three decades. Land and forests are the than other children, although they are more malnour- mainstay of tribal livelihoods but the relationship of ished. But when they do get sick, tribal children are tribals to land is not restricted merely to subsistence much less likely than other children to receive medical cultivation. It extends to their dependence on natural assistance. For instance, only 56 percent of Scheduled resources for livelihoods and for food security. Over Tribe children were taken to a health facility for treat- time, the average landholding has declined more ment of fever and cough in 2005 compared with 67 rapidly among Adivasis than among other groups. percent of non-Scheduled Tribe children. Mothers of This reflects the ‘alienation’ of Adivasis from their tribal children are also less likely to obtain antenatal traditional lands largely through displacement (by or prenatal care from doctors or have an institutional infrastructure projects) or fraudulent private transac- delivery. tions. The government’s 10th Five-Year Plan noted that between 1951 and 1990, 21.3 million people were dis- Poor access to health facilities for tribals is not just placed; 40 percent of them—or 8.5 million—were tribal an issue of low demand for services as medical people. This alienation explains, to a large extent, the practitioners frequently say. There are serious issues poor outcomes among tribals. The loss of control over of supply as well. In most states in India, Scheduled their jal, jungle, and zameen (water, forest, and land) Tribes live in physically isolated hamlets, in remote has alienated Adivasis from public schemes, affected regions and districts and in hilly and forested areas their traditional food practices and forced them to with poorly staffed health centers. Limited coverage migrate to cities to work under harsh conditions. of all-weather roads makes transportation in emergen- cies virtually impossible, even if health centers were Physical remoteness and smaller numbers have attended by medical personnel. Migration of tribals gone together with political isolation and low voice during the lean season to cities and towns makes the in decision making for the Scheduled Tribes. There task of health surveillance for antenatal care or immu- have been measures to assure defacto autonomy nization or growth monitoring of children even more and self-rule to Adivasis, but implementation has difficult. There is also a deep-rooted cultural chasm been patchy. The Indian government’s response to and mistrust between the largely nontribal health vulnerability among Scheduled Tribes has been pro- providers and tribal residents. While administrators active and has included a mix of constitutional and realize the value of recruiting local residents as field budgetary instruments. Both the Panchayat Extension level medical personnel, it is often impossible to find even secondary educated tribal women who can fill 5 In the remote tribal areas of Gadchiroli (Maharashtra), the Society for Education Action and Research in Community Health, for example, is an international success story in maternal and neonatal 4 See for instance, Sheela Barse v/s State of Maharashtra 1993 health. 4 Issue Brief: Poverty and Social Exclusion in India  |  Adivasis to Scheduled Areas Act (PESA) and the Tribal Rights to attaining India’s goal of shared growth. More dis- Act fundamentally question the power relations cussion of tribal aspirations and problems from their between Adivasi and non-Adivasi areas and purport point of view is needed, rather than an examination to transfer greater power to the former. In particular, of such issues through the lens of policy makers, the PESA attempts to give special powers to tribal gram bureaucracy, or the civil society. sabhas (village assemblies) to enhance tribal voice on issues related to mining leases and infrastructure development in tribal areas. It is also unique in that References it is in consonance with customary laws and, rather than on revenue villages, it focuses on tribal hamlets Béteille, A. 1991. Society and Politics in India: Essays in a on the basis of culture. However, it is widely believed Comparative Perspective. London School of Economics that these wide-reaching legislations have not been Monographs on Social Anthropology. New Delhi: Oxford implemented in spirit. University Press. Chopra, Suneet. 1988 “Revolt and Religion: Petty Bourgeois Protest movements among tribal groups have a Romanticism”. Social Scientist, Vol. 16, No.  2, Four long history. Spontaneous uprisings against the state Decades of Economic Development (Feb). pp. 60–67. occurred before the British period, and each such Das, M. B., S. Kapoor, and D. Nikitin. 2010. “A Closer Look at conflagration was an assertion of the cultural and Child Mortality among Adivasis in India.” Policy Research political identity of tribals, as well as a claim on natural Working Paper 5321, World Bank, Washington, DC. resources. A Planning Commission report links move- Government of India. 2008. “Development Challenges in ments since independence and the recent militancy Extremist Affected Areas: Report of an Expert Group squarely with the underdevelopment of Scheduled to Planning Commission.” Report, April, Planning Tribes (Government of India 2008). There is growing Commission, New Delhi. http://planningcommission recognition among policymakers that increasing mili- .gov.in/reports/publications/rep_dce.pdf tancy in tribal areas is not merely a “law and order” Singh, Manmohan. 2009. PM’s address at the Chief problem (Singh, 2009). Instead the causes lie in the Ministers’ Conference on Implementation of the marginalization—spatial, economic and political—that Forest Rights Act 2006, November 4, 2009. Accessed tribal groups have experienced over years. Addressing from http://pmindia.nic.in/speech/content4print the development needs of tribal groups will be central .asp?id=842 1117368