EPW    Discussion December 18, 1999 

Selectivity and Bias

Recent Reporting on Sardar Sarovar Project

D C Sah

DESPITE supposed objectivity within which social scientists function, it is a fact that the knowledge generated and disseminated by them can be better understood vis-a-vis their personal social location and ideology. Given this social reality, some social scientists decontextualise and deconstruct the text and selectively quote others to establish themselves as more committed scholars. Whitehead’s article (EPW, July 10, 1999) is a case in point. She has twisted, and (mis)quoted our ‘findings’. She relies on newspaper reporting and headlines to criticise the Centre for Social Studies (CSS) reports on monitoring and evaluation (M and E) of relief and rehabilitation (R and R) of Sardar Sarovar oustees.

At the outset let me mention five points: one, the CSS has been studying the ‘oustees’ of 19 villages of Gujarat since 1981. All our 24 M and E reports are interlinked. We have stayed in the villages without interpreter, not for flying visit of a week or two but for months together. We have been presenting our understandings in these reports. Information presented in one report need not be repeated in every other. Therefore, if one is honest and desires to have a realistic grasp of issues and analyses, one ought to read the reports together. Second, we do not believe that quantitative methods are better than anthropological. But at the same time we do not discard statistics. Both the methods have their plus and minus points. We more often than not combine both. This is what we have done in our M and E reports. Anthropological method is our strength. We agree with Whitehead that statistical inferences outside their social context tend to become “the types of social scientific obstructions, practices and the philosophies supporting them which erase ‘subjectivity, transforming local and particular experiences into general facts, terms and categories by freeing them of their experimental grounding in lived social relations’ ” [Whitehead 1999:1940]. Third, we have not painted a rosy picture of rehabilitation as Whitehead makes it out. For us the picture is neither black nor white. Therefore, both supporters and critics of the Narmada project have selectively used our reports to suit their different purposes. Fourth, the objective of M and E is to examine what government has and has not done in ‘rehabilitating’ the PAPs. A number of issues – such as first generation problems of relocation, amenities, land related problems, relations with host villages, second generation problems of rehabilitation – have been raised in our different reports. Last, from where has Whitehead learnt that the CSS assumes that ‘tribal’ societies are constituted by a lack of ‘modernity’ and ‘development’? She is putting her words in our studies. Any one who has read the CSS studies on ‘adivasi society’, which are by now many, would agree that we do not have such assumption. The tribal of submerging villages of Gujarat, as we understand, are not a homogeneous group. Therefore, what a group perceives need not be considered the voice of all. However, all social scientists do have ‘a priori biases’. We do not claim to be an exception. But we do contest someone like Whitehead who claims superiority in representing the adivasis of Narmada region.

While being critical of the method used in comparing the non-comparable, Whitehead complains of “little or no ‘before’ statistics, as there has been no attempt made to assign quantitative values to non-marketed inputs and outputs in the submerging villages” [Whitehead 1999:1945]. Had we not been concerned about the situation in the submerging villages, we would not have devoted a whole volume (Report No 15) to socio-economic condition of submerging villages. This report, apart from economic production and consumption, analyses meticulously collected data on gathered forest produce (Report No 15, pp 64-65) and their use in consumption (Report No 15, p 70). Similarly, the data on traditional agricultural implements and on ownership of homestead land and trees (Report No 15, pp 45-47) are documented. Why Whitehead has failed to notice this, she should know. Nonetheless, the pertinent question why the method is quantitative and not qualitative in our reports would better be answered by policy-makers and their consultants who designed these studies. We on our part had made efforts to incorporate as much qualitative insights in these reports as possible.

Whitehead says that “the CSS Reports confidently asserted that the living standards of oustees were better in the resettlement sites” [Whitehead 1999: 1940]. At first, I thought Whitehead is suffering from selective amnesia for, the report concludes that “The finding of M and E Report reveals that though the relocation of over 2981 PAPs had been achieved quite successfully, their rehabilitation has just started and its success calls for a fresh policy intervention” (emphasis added, Report No 24, p i). Even earlier we have said:

Access failures to food is one major problem in the new sites, which erodes well being of relocated PAPs year after year. One may argue that the proportion of PAPs below absolute poverty has reduced compared to what it was in the submerging villages. Academically true. But this is not an academic debate: it is question of well being of a large portion of PAPs, and planners cannot disregard this (Report No 21, p i).

These conclusions have not come out of thin air. We arrived at these by carefully analysing variations, and not averages of consumption intake and factors affecting them. Whitehead touches on the ‘coping with crop failure’ reported by the CSS on 1991-92 drought and concludes that “many of these mechanism seems to have depleted”. What she forgets to present is our insights contained in the reports:

A large proportion of resettlement sites do fall under Narmada command but still there are about 43 new sites – out of total 120 new sites – where Narmada water will not reach (Report No 17, pp 19-22 and 64-68). Even the rest of the sites are vulnerable to fluctuating weather. Although the average farm production of households after relocation has increased, there is wide variations in yield across new site (Report No 19, pp 29-30). In Bhanadra, Chametha, Dhefa, Dhamsia, Golagamdi, Karmaliapura, Kundi, Kasundar, Nasri, Rustampura, Sanoli, Tarsava, Vadadli, Vadesia and Vasla, there was a total paddy crop failure. The same situation prevailed for other crops as well. In fact, about 40 to 50 per cent cultivators have lost their total crop in the case of paddy, jowar, bajri, pulses, cotton, and oilseeds. 1991-92 had been one of the worst agricultural years in the recent history of the region. Rainfall was about 45 per cent of normal. Irrigation in the area is nominal and two out of every five years during last 15 years ending 1993 had been abnormal in the region.

  Crop failure has a number of simultaneous and sequential repercussions which disrupt the economy of affected area (Report No 19, pp 44-48). In submerging villages, access to forest and river on the one hand, and social institutions and mutual exchange relations of the closely knit society on the other, insulate the low level economy to sustain; even during the worst of the draughts, the households’ consumption intake was marginally affected. The fall in crop production in new sites, however, directly reduces households’ income and availability of food for consumption. It also reduces avenues of farm employment. While the current income is reduced, the need for cash to purchase food increases. The households in such situations make various adjustments to minimise the impact of the crop failure. These include decision to opt for casual labouring, migration of some family members, reduction in the use of purchased inputs, taking up to sale of milch animals, reducing, expenditure, and changing the consumption pattern to less expensive items. The findings of the analysis clearly establish that (i) those households which are engaged in casual labour are worst affected by crop failure and their calorie intake is significantly low; (ii) the cattle sale were made in distress and the households who sold cattle were households with significantly lower calorie intake; (iii) households whose agricultural income is higher are households with significantly higher calorie intake; (iv) households with less number of family members to feed have significantly larger calorie intake; and (v) households with larger out-migration have significantly higher calorie intake. Another way to interpret these findings is as follows: (1) in 1991-92 when calorie intake per adult reduced by roughly 480 calorie per day per adult, some members of the household took decision to work as casual labour; (2) if households’ size increases by one member, it reduces the calorie intake by about 98 calorie per adult per day; (3) if households agricultural income per head increases from low to medium level, its calorie intake increases by about 65 calorie per adult per day; and (4) if out-migration increases by about 1 unit, the calorie intake of the members remaining in the village increases by about 200 calorie per adult per day.

  But the PAPs coping mechanism is operating without any external support. With traditional sustaining mechanism distorted due to randomness of resettlement, all these options in new site may create further social disruptions.

Using a case study presented in Report No 19, we have tried to bring out the contrast between PAPs and the host villagers with respect to technology adoption, yield and income: gross income is significantly higher in farms located in host villages compared to the income of PAPs. The difference in income grows larger as one moves from dry maize and paddy zones towards irrigated cotton zone. This difference is mainly because of a mix of superior cropping pattern, higher crop yield and better prices received by the farmers located in the host villages, specifically in irrigated cotton zone. Evidently, the new settlers have much less negotiating capability when it comes to new technology and institutions; this is true whether it is relating to choosing a cropping pattern, yield of new crops or bargaining for remunerative prices for their crops. The impact of forced commercialisation, on product and credit markets was also examined:

Although, the institutional and private informal credit sources coexist in this drought prone zone, what is unique in this area is that proportion of PAPs availing of any working credit from formal or informal sector is significantly low. Despite having large land assets, only a minuscule of oustees were members of co-operative society, and no one had access to co-operative short-term credit. Only a few oustees had access to bank finance, and fewer have shown any interest on private production credit sources. This low reliance on formal and informal production credit in the new sites is not very surprising when one introduces behavioural factors in the analysis. The dominant phenomenon of exploitative nature of interlocked credit, land, and output markets is widely feared among tribal of south Gujarat; the small borrowers resulting into generations of bondage to debtor is well recorded. This interlocking of credit and labour markets is also well understood by tribal resettlers in new sites. Thus, the cautious approach of ‘self-financing’ the current production and avoiding risk of losing land to debtor is not surprising. It should also be noted that oustee farmers are late comers in credit nexus, where besides assets, social links also play important role. Credit co-operatives are faction ridden and supporter of the dominant faction gets easy access to credit and other benefits than those who belong to dissident faction and or the new comers. On the other hand, output market links are well developed. But it is observed that the proportion of total cereal marketed is significantly higher among those who have relatively low intake. This is the group who should have retained its produce more for home consumption (Report No 21, pp 24-27).

Another limitation of the CSS Reports according to Whitehead is the findings on health care. She says that the CSS reports claim that health facilities are better in or near resettlement sites as compared with submerging villages. She observes that “for most ailments, people tended to prefer a bhuwa, who used ayurvedic remedies and who lived at a distance of three kms from the village” (Whitehead 1999:1944). The CSS findings on health services are developed in Report Nos 19 and 22. We have said that “In general, the oustees who are residing in locations which have health centre in the village, feel an increased access to as well as qualitative change in health services after relocation. The need is to develop this confidence among all the oustees”. Such locations are not many. In order to understand changes in health perceptions and practices of the oustees, the findings of Report No 22 are worth noting:

Relocation is responsible for rapid diffusion of alternative health care system amongst the oustees, irrespective of their tribal identity, their past experiences with modern amenities, or their present locational advantage; tribal whether they are tadvi, rathwa or bhil, whether they originally belonged to Zone I or Zone III, or whether they are presently located near a town or in relatively remote new site, their dependence on allopathic system is noticeably high. At the cognitive level, illness arose perceptions not in congruence with their health practices. Although noticeable increase was observed among oustees for showing their faith for allopathic system, their perceptions about ill health are still guided by their tribal environment and cultural beliefs formed in their submerging villages. And thus, dual treatment still exists; for common diseases they still get satisfaction from their indigenous practices which save their time and money. On the other hand, for uncured sickness, they prefer to visit nearest PHC, hospital or a private practitioner. If the diseases still remains uncured, the search for Bhuwa again starts (Report No 22, pp 116-18).

While discussing basic amenities in the new sites we have reported that “mere existence of service in the new site does not mean that these services are easily accessible to all, and oustees residing in these new sites have improved their well-being” (Report No 21, p 36). Relating to drinking water facilities are CSS report says:

Hand pumps are the main source of drinking water in new sites. Apart from this, well, pond, rivulet and in some locations tap connection also supplement the drinking water supply. The new sites, both relatively old as well as of recent origin, have not yet organised themselves in maintaining their drinking water sources, especially hand pumps. Consequently, trouble free drinking water services are often disturbed. In Sandod, Gopalpura, Farakua, Jetpur and Devaliya hand pumps have become non-functional as a result oustees have to carry water for all their use – drinking, washing and for cattle – from host villages. The new sites Haripura, Samkwa, Bonya, Nasri, Vasla, Kasumbia are totally dependent on host villages for drinking water. Supply of water through tanks is not uncommon: oustees in Dormar, Kundi, Laver, Vardi and Paniya-1 reported having frequently transported water from Bhadroli new site. Oustees from many other new sites also have reported such transportation of water from host villages or other new sites. Oustees have come from a geographical area which had perennial source of drinking water. As a result, they become nostalgic when talking about drinking water facilities. This is not surprising but their nostalgia coupled with frequent disruption in supply of drinking water has given rise to dissatisfaction (Report No 21, pp 43-47).

Another distorted presentation of the CSS findings is relating the oustees’ perceptions about their economic condition in new sites in comaprison with that in submerging villages. For us it is disconcerting if a third of the relocated households are dissatisfied. We are also concerned to find factors responsible for this situation. To put the issue in perspective, I present some of the findings (Report No 23, pp 9-10 and Report No 22, pp 101-02):

The oustees were asked to report their perceptions regarding their economic condition in new sites in comparison to what it was in their submerging villages. In 1993-94 about a third of the households have shown dissatisfaction with their condition. It would be worthwhile to analyse the reasons behind the increased dissatisfaction amongst the oustees. It is hypothesised that such negative feelings are closely related to falling avenues of cultivation, lower availability of irrigation, falling levels of farm production, and stress that force the household member to migrate in order to do some labouring. Findings reveal that net cultivated area, gross cropped area, percentage area irrigated and farm production (at constant prices) were significantly higher in the case of households who had perceived that their economic condition has improved or remained same. What it means that those households who had low cropped area, less irrigation and low agricultural output were relatively dissatisfied with their economic condition. When all the households were provided with equal amount of land, a lower net or gross cultivated area is on account of a proportion of area remaining uncultivated. Analysis of Universal Household responses reveals that in every new site there were one or two PAPs who feel that they were given less land; a large proportion of PAPs in Malu 1 and 2, Krushnapura, Simlia, Pansoli, Lunadra and Ferkuva believe that they have been provided with unproductive land which requires treatment; Some PAPs located in Kanteshwer, Kali-Talavadi, Pansoli, Malu-1, Pania-1 and 2 and Sanoli have reported that their land remains water-logged during kharif; a large proportion of PAPs located in Krushnapura, Malu-1, Pachhisgram, Sanoli and Zamb have reported that part of their land is coming under canal and road. To some extent, access to irrigation helps in improving oustees’ perceptions for it allows technology transfer and higher productivity. This is further vindicated by the fact that households with low farm production had reported that their economic condition in comparison to submerging villages has deteriorated. Negative and significant correlation reveals that lower the farm production, the higher the probability that the household shall be reporting deterioration in its economic condition. On the other hand, those households who have reported deterioration in their economic condition also were those who have taken up to agricultural labouring and casual labouring more. It also means that when production do not fully sustain current consumption, some members of the household move out in search of employment.

In this connection the following insight on developing dissatisfaction relating to unproductive land is worth mentioning:

Low gross cropped area is mainly because of large proportion of net area remains uncultivated. Evidences have revealed that some households have serious problems relating to the quality of agricultural land. For example, households in Ambavadi and Piparvati have such a severe dissatisfaction relating to land quality that they now refuse entry to outsiders in the village until they are relocated to some other new sites. Situation in Malu is slowly deteriorating. There is widespread dissatisfaction among the resettled oustees in Malu. We have contacted Malu oustees for detailed discussions. Some of the salient features of this discussion are as follows: In Malu 50 PAPs have perceptions that they have been given unproductive kochar-dabwali agricultural land which they think warrants either relocation or some treatment. Amongst these, there are some cases of short changing in land, some have reported that the land is unirrigated, and a few have lands where waterlogging during kharif is quit problematic. In all about 32 families were not staying in Malu, bulk of them are in their submerging village. In order to ascertain the actual situation, we went to submerging village Gadher on November 8, 1995. We found that 22 oustees, some of whom with their families, were staying in one of the hillocks under the leadership of Dalsukh Sonji Tadvi. All these PAPs believe that they had been provided with unirrigated, unproductive land. They also believe that sustenance on land is not possible, for yield from their land is extremely low. Some of them moved out after leasing their lands. These 22 families are staying in Gadher since May 8, 1995. Dalsukh Sonji says, “Year after year since 1992, I am facing a risk. Although it is not unusual for a farmer to take risk but for me crop failures have become a recurring phenomenon. When I will earn? How long I should rely on my savings? Every year my crop washes out due to flooding from adjoining river and land remains waterlogged during both kharif and rabi seasons. Moreover, when it was decided that the oustees would be resettled in irrigated land, why we are not provided with irrigation? Narmada water shall be coming in the area in next five to seven years, that is all right but we have to live today and need our sustenance today. We can no more wait for tomorrow” (Report No 21, pp 13-14).

Within the given frame of monitoring against certain indicators, the CSS reports are able to bring out (a) level of living of the oustees in terms of access to health care and other amenities, agricultural income and other avenues of employment, consumption and expenditure; and (b) problems faced by PAPs in their relocation and rehabilitation. Our findings clearly show that pressure on land to augment resource has led to diversification and double cropping even in the dry lands. But in absence of viable technology, yield risks have not been reduced. Consequently, some of the members from the adversely affected families have taken to labouring or seasonally migrated to support the consumption of the remaining members of the family. In the process, hitherto unknown institutional arrangements like sharecropping, fixed tenancy and land mortgage have also developed. Despite this, about a third of PAPs during normal years and about three-fourths of PAPs during abnormal years live in abject poverty.

By choosing not to mention these above findings, Whitehead perhaps wants the reader to believe that her own research findings are not based on the discussions in M and E reports. Her kind of research relying on 15 days field study could only be used to test the ideas and conclusions reached elsewhere. In order to appear original, she not only has had to be selective in quoting the findings of her sources (the CSS reports) but also to discredit them.


Sah, D C (1993): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 15, Centre for Social Studies, Surat.

– (1993a): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 16, Centre for Social Studies, Surat.

– (1994a): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 17, Centre for Social Studies, Surat.

– (1994b): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 19, Centre for Social Studies, Surat.

– (1995): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 20, Centre for Social Studies, Surat.

– (1995a): ‘Development and Displacement: National Rehabilitation Policy’, Economic and Political Weekly.

– (1996a): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 21, Centre for Social Studies, Surat.

– (1996b): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 22, Centre for Social Studies, Surat.

– (1996c): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Report No 23, Centre for Social Studies, Surat.

– (1997a): Monitoring and Evaluation of Resettlement and Rehabilitation Programme for Sardar Sarovar (Narmada) Project, Draft Report No 24, Centre for Social Studies, Surat.

– (1997b): ‘Second Generation Problems of Involuntary Migration’, Indian Journal of Agricultural Economics, 52(4).

– (1998): ‘Avoiding Impoverishment Risks’, Review of Development and Change, 3(1).

– (1999): ‘Pressures on Land, Employment and Migration: Evidences from Sardar Sarovar Project’, Review of Development and Change, (forthcoming).

Current Issue    Previous Issue    Archives    Subscription
About the EPW    Sameeksha Trust    EPW Research Foundation
Advertising    Announcements    epw@vsnl.com    Home    The Site