Drought Proofing Rajasthan Imperatives Experience Prospects

8/10/2019 Drought Proofing Rajasthan Imperatives Experience Prospects

1/50

Discussion Paper Series - 5

Drought Proofing in Rajasthan:Imperatives, Experience and Prospects

by

Reetika Khera

The analysis and policy recommendat ions of this Paper do not necessarily reflect the views of theUnited Nations Development Programme, its Executive Board or its Member States.


2/50

ACKNOWLEDGEMENTSACKNOWLEDGEMENTS

This paper is based on fieldwork for a Ph. D. at the Delhi School of Economics. The author would like to thank the UNDP for a grant that made this possible. During the fieldwork the author was helped by Kavita Srivastava,

various NGOs and also by many people in the villages. Without their logistical help and support, this paper would not have been possible. The author would also like to thank Piers Blaikie, Arudra Burra, Nikhil Dey,K. Seeta Prabhu, Suraj Kumar, Francois Leclercq, Ritu Mathur, Elena Borsatti, Ben Rogaly, Janet Seeley andKunal Sen for helpful suggestions in writing this paper. Finally, special thanks are due to Jean Dreze forguidance during fieldwork and otherwise.


3/50

PROLOGUEPROLOGUE

Droughts have affected many peoples lives in Rajasthan. The adverse impact is due to widespread cropfailure, which leads to shortages of food, fodder and drinking water, and migration.

This paper calls for a comprehensive drought policy, as droughts are a recurring phenomenon in the State. Butit also argues that successful policies need a change in the understanding of droughts. The impact on peopleslives should be taken into account and diversification of livelihoods in Rajasthan should be promoted. Theanalysis is also based on fieldwork carried out in four districts of the State.

This paper is one of the studies commissioned by the HDRC to have a deeper understanding of the livelihood

situation in a few States.

Reetika Khera is currently pursuing a Ph.D. (Economics) at the Delhi School of Economics. She is studying government interventions and their effectiveness during the recent drought in Rajasthan.


4/50

v

CONTENTSCONTENTS

EXECUTIVE SUMMARY 1

I DROUGHT: ITS PERIODICITY AND MEASURES 2

Types of Drought 2

Alternative Indicators 3

II THE DROUGHT OF 1999-2001: METHODOLOGY USED 6

Selection of the Districts 6

Selection of the Villages 6

Selection of the Households 7

The Villages 8

III THE SEVERITY OF THE DROUGHT OF 1999-2001 13

Agriculture 13

Livestock 15

Coping Strategies 16

IV GOVERNMENT POLICY 21

Drought Mitigation 21

Drought Proofing 25

V CONCLUSION 30


5/50

v i

STATISTICAL ANNEXURES 31

Table 1 Some Indicators of Drought in Rajasthan, 1981-2 to 2002-3 31

Table 2 Rainfall and Famine-affected Villages, 2001 32

Table 3 Basic Features of the Sample Villages 33

Table 4 Land Owned and Area Cultivated ( Bighas ) by Various Caste Groups, 2001 35

Table 5 Changes in Area Sown and Total Output of Main Foodgrains 36

Table 6 Change in Total Livestock Wealth of the Sample Households 37

Table 7 Coping Strategies of Households in Order of Importance 38

Table 8 Vulnerable Groups in the Sample Population 39

REFERENCES 40


6/50

vi i

ABBREVIATIONSABBREVIATIONS

CR F Calamity Relief Fund

DDP Desert Development Programme

DPAP Drought Prone Areas Programme

GOI Government of India

GOR Government of Rajasthan

IRDP Integrated Rural Development Programme

OBC Other Backward Castes

NCCF National Calamity Contingency Fund

PDS Public Distribution System

SC Scheduled Castes

ST Scheduled Tribes

SWRC Social Work and Research Centre

URMUL Uttar Rajasthan Milk Union Limited


7/50

vii i

GLOSSARYGLOSSARY

Anna akaal Food drought

Bajra Pearl millets

Bania Trader caste

Bigha Measure of land

Chana Chick peas

Chowktis Casual labour market

Dhanis Hamlet

Gauna Beginning of effective married life

Girdawari Land-use report

Gram sewak Gram Sabha Secretary Gwar Cluster beans

Jal akaal Water drought

Jeera Cumin

Jow Barley

Jowar Sorghum

Kasidakari Embroidery

Kharif Monsoon crop

Luhar Caste of blacksmiths

Makka Corn

Mochi Caste of cobblers

Moong Green gram, a kind of pulse

Moth Haricot, or dew bean

Nadi Pond

Nai Caste of barbers

Panchayat An administrative unit comprising of a few revenue villages

Patwar Land Revenue Circle

Patwari Land Revenue Officer

Rabi Winter crop


8/50

i x

Sunars Caste of goldsmiths

Suthars Caste of carpenters

Tanka Underground water storage structure

Til Sesame seeds

Trinakaal Food, fodder and water droughtTuar Arhar lentils

Urad Black gram


9/50

1

Executive SummaryExecutive Summary

Droughts are a recurring phenomenonin Rajasthan. This paper discusses vari-ous aspects of this phenomenon: how people deal with drastic crop and live-stock losses, especially when coupled

with limited employment opportunitieselsewhere (Coping Strategies); gov-ernment policies to provide relief in

such times (Drought Mitigation); andgovernment policies to prevent the oc-currence of droughts (Drought Proof-ing). Currently, droughts are perceivedas sporadic events so policies are not de-signed in the best possible manner. Thispaper argues that successful policies fordrought mitigation and proofing requirea change in our understanding of droughts. Thus a review of drought-re-

lated policies must be accompanied by a review of the definitions adopted todeclare a drought.

This paper is partly based on fieldwork carried out in four districts of Rajasthanfrom May 2002 to February 2003, col-lecting information on the drought of 1999-2001. The first section looks athow droughts are defined, and what evi-dence is used to judge their frequency.

The second section outlines the meth-

odology used and attempts to judge theseverity of the drought on the basis of

various indicators. Section III discussesthe drought of 1999-2001, based on datacollected during fieldwork. It also con-tains a discussion of the coping strate-gies that emerged in the sample areas.

The concluding section provides a shortoverview of the governments droughtpolicies, and discusses the policy impli-

cations and recommendations thatemerge from the discussions in theearlier sections.

Successful policies for drought mitigation and proofing require a change in our understanding of droughts


10/50

2 DROUGHT PROOFING IN RAJASTHAN: IMPERATIVES, EXPERIENCE AND PROSPECTS

Drought: Its Periodicity and Measures

SECTION I


Currently, droughts are not treated as aregular feature of Rajasthan, but ratheras sporadic events. In this section, welook at various indicators of droughtand their shortcomings. The period un-der study spans approximately two de-cades from 1981-2 to 2002-3. The dis-cussion will bring out clearly the need

to change our perceptions of drought,because these perceptions influencepolicy design.

1.1 Types of Drought

What counts as a drought? Put simply, adrought is a failure of rain, leading tomoisture stress, that in turn leads to ag-ricultural losses and other forms of so-cial and economic hardship. There aremany definitions and classifications of drought, including that of the NationalCommission on Agriculture (quoted inBokil 2000) which has defined threetypes of drought:

Agricultural : When crops are affecteddue to moisture stress and lack of rainfall.

Meteorological : When there is more than

25 per cent decrease (from normal)in rainfall in any area.

Hydrological : When recurring meteo-rological droughts result in decreasein surface water and groundwaterlevels.

Droughts are a matter of concern as they affect the lives of people. We will focuson agricultural droughts. In keeping withthis focus, we may classify droughts ac-cording to the shortage of three com-modities: water ( jal akaal), food ( anna akaal) and fodder ( trinakaal).

1.1.1 Rainfall

Rajasthan can be divided into four broadagro-climatic regions and indeedphysiographical regions: the arid regionin the west, the eastern plains (whichare fertile and receive a good deal of rainfall), the hilly region of the Aravallisextending from the north to the southin the eastern part of Rajasthan, and theplateau in the south-eastern part of the

State. Rajasthan receives most of itsrainfall over a period of three months roughly from July to September. Theaverage annual rainfall is 531 mm,compared to the all-India average of 1100 mm. Rainfall is not only low, butalso uncertain. Further, there are wide

variations in rainfall across the State from as little as 193 mm in Bikaner inthe west, to as much as 607 mm in theeastern plains in normal years.

The Indian Meteorological Departmentdefines a failure of the monsoon as ayear in which the actual rainfall has been20 per cent less than the normal rain-fall.1 By this definition, in Rajasthan the

A drought is a failure of rain,leading to moisture stress, that in turn leads to agricultural losses and other forms of social and economic hardship

1 Normal rainfall is calculated as the long period average.


11/50

3

monsoon failed in just two years 1987and 2002 (Table 1, p. 31).

Rainfall and agricultural operations inRajasthan, and indeed in many parts of India, are very closely related. Therefore,

rainfall and land area sown tend to bepositively correlated.2 In Rajasthan, only 25 per cent of agricultural land is irri-gated. However, the relationship be-tween rainfall and area sown is quitecomplex and depends on the regionaldispersion, the timing and duration of each spell, as well as the gap betweensuccessive spells. These factors are likely to be as important for agricultural out-

put as total rainfall. Meteorologicaldroughts do not take these factors intoconsideration and therefore cannot beused to make judgements about agricul-tural droughts.

The area sown is a slightly better indi-cator of a drought because it reflects oneof the ways in which the lack of rain-fall affects human life. Since most agri-culture in Rajasthan is rain-fed, andpeoples livelihoods are still quiteheavily dependent on agriculture, fall inarea sown provides us with a good start-ing point to judge the severity of adrought. Table 2 (p. 32) demonstratesthat rainfall data can be misleading be-cause there is only a loose relation be-tween amount of rainfall and area sown.It provides a cross-tabulation between

the amount of rainfall and the propor-tion of villages that were declared fam-ine-affected in the year 2001. By theState governments classification, anarea is declared scarcity-affected

when it reports crop losses of 50-75 percent, and famine-affected when thelosses exceed 75 per cent. If one wereto judge the occurrence of a droughtbased only on whether rainfall was

normal or not, only 6 districts wouldhave been classified as drought-hit(column 3). However, when we look atcrop losses, we find that in as many as16 out of the 25 districts that receivednormal to excess rain, more than fifty per cent villages had been declared fam-ine-hit. Thus we see that lack of rain-fall (a meteorological drought) doesnot adequately capture what actually af-

fects peoples lives and we need to usealternative criteria to determine whetheror not a drought has occurred.

1.2 Alternative Indicators

While it is true that meteorological rea-sons have been the cause of the worstdroughts in Rajasthan (1987 and 2002),it is important to separate the triggerevent from its consequences. Though

droughts may be triggered by lack of rainfall, they affect people in many ways(e.g. loss of home-produced foodgrain,loss of employment, difficulty in getting

water for daily domestic needs and foragriculture, lack of fodder for livestock,etc.). Change in level of agricultural out-put, availability of fodder, employmentand water are therefore good indicatorsof drought.

When the State government declares adrought, it uses indicators that, in prin-ciple, include those mentioned above. Inthis section, we examine changes in

2 In his analysis of rainfall data for the period from 1877-1986, Sivasami finds that when rainfall hasbeen less than 20 per cent of normal, the area sown in the country can fall by as much as 50 per cent.

When the shortfall is between 10-20 per cent, the affected area ranges between 20-40 per cent (Sivasami,2000: 1991).


Lack of rainfall does not adequately capture what actually affects peoples lives


12/50


agricultural output and livestock popu-lation, government expenditure on relief

works, and person-days of employmentgenerated. The analysis relies on second-ary data published by the Government

of India or the Government of Rajasthan(Table 1, p. 31).

The Scarcity Manual (formerly knownas the Famine Code) for Rajasthan laysout the rules and procedures to be fol-lowed in declaring a drought. In this pro-cess, the girdawari report plays a key role.

The girdawari report is a land-use reportand is prepared by the patwari (landrecords official) of each panchayat . Theother criteria in the Scarcity Manual in-clude distress migrations, increase inthefts, news of starvation deaths, etc.

While the Scarcity Manual includes many criteria, in practice, the State govern-ment has come to rely almost exclusively on the girdawari report and the losses insowing and production reported therein.

To calculate the losses, the current years

figures are compared with area sown andproduction in normal years (definedas the average production for the pastfew years). On the basis of this, calcu-lations of affected population are made.

Affected population

Probably the most important indicatorof the severity of a drought is the num-ber of people it affects. In 14 of the 22

years between 1980 and 2002-03, agri-cultural operations in more than 10,000 villages were declared either scarcity -or famine-hit. Only in three years dur-ing this 22-year period, was the numberof affected villages less than 500.

Labour employed

It has been mentioned that during droughts the State government providesrelief for the able-bodied by opening employment sites in the affected villages.

Only in two years over the entire period1981-2003 did the State governmentdeem it unnecessary to provide such re-lief work. In each of the remaining 20years, the government generated, on anaverage, 3.35 lakh person days of employment. During the drought of 1987-8, nearly 220 lakh person days of employment were generated.

Government expenditure

Government commitment to providerelief measures can also be gauged fromits drought relief expenditure (Table 1,p. 31). These funds are used largely toinitiate food-for-work programmes, toensure water availability during adrought, the supply of subsidized fod-der for the conservation of livestock and the provision of gratuitous relief

for the weak, old and disabled. Thegovernments commitment is quite clearfrom the financial resources it sets asidefor these obligations. In the period from1981-2 to 1989-90, the State govern-ment expenditure on relief operations

was Rs. 1515 crores and in the periodfrom 1991-2 to 1999-2000, it wasRs. 1057 crores.3

Agricultural production

Given the overall importance of agri-culture in the State, in terms of its con-tribution to the Gross State DomesticProduct (27.4 per cent in 1999-2000),as well as in terms of proportion of

The most important indicator of the severity of a drought is the number of people it affects

3 This expenditure is at current prices. If we look at expenditure at real prices, the fall in expenditure in the1990s is likely to be larger still.


13/50

5

people who rely on agriculture as theirmain means of livelihood, it makes senseto look at the growth of this sector.

Trends in the production figures of foodgrain for the State show that, since

the 1980s, in four years, the fall in agri-cultural production was much higher(more than 20 per cent) than that of theprevious peak year.

Livestock population

The Government of India carries out alivestock census every 5 years. The lastcolumn in Table 1 (p. 31) reports thetotal cattle population at various points

of time. This brings out clearly the im-pact of the drought of 1987-8, whichcaused the total cattle population todecline to such an extent that it had notrecovered to its pre-drought levels evenat the time of the last census in 1997.

This probably also has to do with achange in the composition of livestock:

there is a move away from cattle towardssheep and goats, which are known tobe sturdier.

Thus, going by many of the above indi-cators, while there has been a meteoro-

logical drought in just two years of the22 years under study, by all the alterna-tive indicators, only three or four years(depending on which criterion one uses)have remained drought free.

From the available data it is not possibleto say whether the intensity or severity of droughts has been increasing over theyears, as has been suggested by previ-

ous research.4

For doing so, one wouldhave to look not only at the above indi-cators, but also at the duration of eachdrought cycle and at the regional spreadof each cycle. Localised droughts orsingle-year droughts are known to be lesssevere on the human populations af-fected by them.

4 Jodha (1991) and Vidya Sagar (1995) both have suggested that the intensity of droughts has increasedover time.


Localised droughts or single-year droughts are known to be less severe on the human populations affected by them


14/50


The Drought of 1999-2001:Methodology Used

The Drought of 1999-2001:Methodology Used

This section discusses the methodol-ogy used to study the impact of thedrought of 1999-2001. Data was col-lected from nearly 400 households ineight villages by means of a question-naire. In addition to collecting dataon the background characteristics of the households and individuals, the

questionnaire covered the severity of the drought, coping strategies, thepublic distribution system and food-for-work programmes. This wassupplemented by group discussionsand informal discussions with variouspeople in the villages, including thesarpanch , ward panches , pa twar i , gram sewak and others.

2.1 Selection of the DistrictsSince the survey could not cover very large parts of the State, or provide asample in which all regions got ad-equate representation, it was carriedout in four districts, chosen to reflectregional variations. The sample districts

we re Barmer, Bika ner, Ja ipur andUdaipur. Barmer and Bikaner are in thesouth-west and north-west of Rajasthanrespectively (along the border with Pa-kistan). Jaipur lies in the north-easternpart of Rajasthan and Udaipur in thesouth. These districts not only give afair geographic spread, but also reflectagro-climatic variations and differing levels of development in different parts

of Rajasthan. Bikaner and Barmer areboth semi-arid regions with sandy soil.

Jaipur is mostly flat, with fertile land, while Udaipur lies in the Aravallis andhas fertile but not level land, and richforest reserves.

2.2 Selection of the Villages

Once the districts were selected, a two-stage random sampling procedure wasadopted. The first stage was a selec-tion of clusters , which were taken tobe revenue villages according to the1991 census.

The household population of the rev-enue villages tends to vary quite a lot-thus smaller villages would standproxy for a small number of house-holds and larger villages (with a popu-lation of more than 1500 persons)

would be proxy for a larger number of households. But the bulk of the popu-lation lies in villages with a popula-tion of 800-1500. A simple randomsample of villages would give rise tobias in the estimates of populationcharacteristics because it would give

equal probability of selection tohouses in small and large villages as it would to houses in medium-sized vil-lages. To improve the precision of thesample, such (i.e. medium-sized) vil-lages should have a greater probabil-ity of selection. This suggests that toget unbiased sample estimates for the

SECTION II

The survey was carried out in four districts,chosen to reflect regional variations


15/50

7

population, households in mediumsized villages should have a higherprobability of selection than those insmall or large villages. This can beachieved by employing probability pro-

portional to size sampling.5

Another consideration in the selectionof villages was whether it would be pos-sible to stay in the village for the dura-tion of the fieldwork. This meant thatonly those villages in which some con-tact was available - mainly throughNGOs working in that district - wereincluded in the final frame from whichthe sample was chosen.6 This gave a listof villages from which the sample vil-lages were chosen.

Replacement villages were also chosenat this stage. In cases where there hadnot been any relief work in the sample

village during the summer of 2000, a re-placement village where there had beenrelief work during this period was sur-

veyed. Three of the villages in this

sample are replacement villages. In onecase, the sample village was inaccessiblesince it lay across a river which couldnot be crossed because of the monsoon.In this case, the last accessible villageon the road to the sample village satis-fying the population criterion was cho-sen for the survey. Thus the final sampleincludes 4 sample villages (Biramsar,Dulmera Station, Baasri Jogiyan and

Sukaliya), three replacement villages

(Morda, Kharad and Badli) and one other village (Birothi).

2.3 Selection of the Households

At the second stage, a simple random

sample of 400 households was chosenfrom the voter list of the village pre-pared by the district authorities. The

voter lists are updated each year andan attempt was made to use the elec-toral roll for 2001 (the latest available).But this was not possible in all cases,in which case an older electoral roll hadto be used. The oldest electoral rollused for selection of sample households

was from 1998. While sampling, up to 10 replacementhouseholds were also selected to makeup for missing or unresponsive house-holds. However, all the sample house-holds proved willing to respond; in fact,in most villages, it was those who wereleft out of the sample who were un-happy. In most cases, this was due to themisconception that the households that

were being surveyed would receive somebenefits, which non-sample households

would be deprived of. Replacementhouseholds were surveyed when

Members of sample households wereunavailable even after the second visitto the house, or during the investiga-tors stay in the village. (This was themost common reason for replacing a

sample household).

5 See Deaton (1997: 15).6 The NGOs were URMUL in Bikaner, SURE in Barmer, Sewa Mandir in Udaipur, SWRC and Vishaka in

Jaipur. As it turned out, in three cases the NGO did not really work in the sample village but did in aneighbouring one. As a result, they were able to introduce the author to an acquaintance with whom she thenstayed for fieldwork. In the other four villages, Biramsar, Dulmera Station, Birothi and Morda, people knew about the NGO through whom the author had made contact, and quite often had benefited from it in some

way or another. In one village, the local contact did not work for an NGO but for a youth group with whomthe author was acquainted.

The Drought of 1999-2001: Methodology Used

A consideration in the selection of villages was whether it would be possible to stay in the village for the duration of the fieldwork


16/50


The sample households had migratedfrom the village.7

Due to pressure of time, the samplehousehold seemed too far away to

visit.

There was no surviving member fromthe sample household.

Since the voter lists were slightly out-dated in some cases, it often turnedout that sample households had splitinto more than one household sincethe list was compiled. In such cases,the oldest member of the family, orthe most easily available member, was

surveyed. Finally, the sample alsocontains some households that wereneither in the list of sample house-holds nor in the list of replacementhouseholds. This includes extremely destitute people who were encoun-tered during the survey.

The interviews were car ried out withadult members of each family. Bal-ance was maintained between thenumber of male and female respon-dents. Sometimes an interview wasrepeated with two members of thesame household at different times tocorroborate answers given by the firstrespondent. At other times, the inter-

view was carried out in the presenceof more than one adult member of the family, with different persons an-swering different questions.

2.4 The Villages

2.4.1 Demographic and SocialCharacteristics

Th is secti on pr ovides som e back-ground information on the villages that

were surveyed. Table 3 (p. 33) givesa detailed picture for each village inthe sample.

Caste Structure

The largest caste group in the sample(42 per cent) is that of other back-

ward castes (OBCs), a category thatincludes a large number of castes.

This group is far from homogenous.It includes politically articulate andeconomically powerful castes suchas the Jats , as well as castes lowerdown on the social scale such as

Nais , Luhars , Suthars , Sunars smallin number as well as less articulate.In fact, the Jats have managed to havethemselves classified as OBCs inRajasthan precisely because of their

political clout.8 The other large groupin the category of OBCs is that of the Muslim castes (nearly 11 per centof the OBC population of 42 per centare Muslim)9.

The Scheduled Castes and Tribes form17 and 19 per cent respectively of the population, with the other castes( Brahmins and Rajputs ) co mpr ising 22 per cent of the population.

7 The greatest number of such cases was in Badli where the migration was not drought-related, but wasrelated to a local feud which had caused all the members of one family to flee from the village. In theother villages, when there were such households, their numbers were quite small and never more thanthree households.

8 As a result of a recent political mobilization, the Jats were able to attain the status of OBCs so that they couldbenefit from the policies of affirmative action, mainly reservations in government jobs. Jats were added to thecategory of OBCs as recently as 1999 (Datta, 1999).

9 Many of the Muslim households said that they too had a caste (e.g. Alisar , Langa , etc.) and attested to ahierarchy among these castes.

It often turned out that sample households

had split into more than one household since the list was compiled. In such cases,the oldest member of the family, or the most easily available member, was surveyed


17/50

9

Education

More than half of the individuals in thesample households were illiterate.

About a third had completed school upto primary level, and just 11 per cent

had finished secondary school.

Sex Ratios

The sample villages had a female-maleratio of 913, i.e. 913 women per 1000men. According to the 2001 census, thefemale-male ratio in Rajasthan is 922,

which is slightly lower than the All-In-dia figure.

Age distribution One-fifth of the population was chil-dren aged less than six years and abouta third of the population was in theschool-going age group (7 years to 18years). Another third of the populationconstituted the working age group(19-40 years) and 5 per cent of the popu-lation was aged 60 years and above.

Marital Status Forty three per cent of the individuals inthe sample households were currently married; another 2 per cent were mar-ried but had not performed the gauna 10.Four per cent of the individuals were wid-owed and the remainder (approximately 50 per cent) had never been married.

2.4.2 Physical and Social

InfrastructureLocation and Status

The geographic location and administra-tive status of a village is significant interms of providing (or denying) accessto panchayat officials and also to govern-

ment services such as healthcare centres,schools, and especially fair price shops(PDS ration shops). Each panchayat hasup to 10 villages, one of which is theadministrative headquarters for that

panchayat , and after which the panchayat is usually named. Not belonging to the pa nc ha ya t headquarters (the gr am panchayat village) can mean not know-ing when the ration shop will be open or

wasting many days to get a signaturefrom the patwari or gram sewak. Only twoof the eight sample villages were also

gram panchayat headquarters: these wereBadli and Birothi, both in Udaipur. One

village, Biramsar, had its own ration shopand high school, though it was not a gram panchayat village; these privileges werecourtesy the local MP, who hailed fromit. The other five villages did not haveration shops, so people faced tremendousdifficulty just in getting their monthly ra-tions. Since ration shops were not openon fixed days, they had to make repeatedtrips outside their village, incurring the

cost of transport as well as losing wages. Access to primary schools is no longer aproblem in any of the villages. In fact,except for two, the remaining villageseven had a middle school within 3 km,if not within the village itself.

On the other hand, access to health fa-cilities is still a huge problem. Some vil-lages even lack access to an Auxiliary

Nurse Midwife (ANM). For people liv-ing in Sukaliya, the closest Primary Health Centre is more than 10 km away.

Miscellaneous infrastructure

All the villages but one have access to aroad. The exception is Sukaliya, which

10 Gauna is the ceremony after which the bride leaves her natal home to live with her husband.


Access to primary schools is no longer a problem in any of the villages. On the other hand, access to health facilities is still a huge problem


18/50


is 6 km from the nearest road. Four vil-lages have a direct bus service. Two vil-lages are not electrified and the others,though electrified, have erratic supply.In half of the villages, some homes had

telephones.Drinking water

The drinking water problem is mostacute in Bikaner and Barmer, where it isa perennial problem. Here people haveto rely on rain-water harvesting and ongovernment supplies (generally througha pipeline) for their needs. The mainform of rain-water harvesting in the

sandy regions is the construction of tankas (underground storage tanks). Incontrast, people in Jaipur and Udaipur,rely mainly on handpumps for water,

which is available most of the time, ex-cept occasionally, when handpumps dry up or stop working. Four years of con-secutive drought have lowered the wa-ter table, and some handpumps havegone dry as a result. The quality of wa-ter is also a matter of concern. In west-ern Rajasthan, water salinity is a prob-lem, whereas in Jaipur, high fluoride con-tent in drinking water was reported asthe cause of many health problems.

The worst-off villages were those whichgovernment supply of water did notreach. People had to walk for up to4-5 hours to get their supply of drinking

water from the nearest governmentsource. In Sukaliya (Barmer), it was com-mon to hear stories of men who left afterthe morning meal and returned with wa-ter when it was time for the evening meal.11

Where water is supplied by the govern-ment, it is generally piped water suppliedto a water tank in the middle of the vil-lage. However, this does not solve theproblem entirely. The water supply re-mains uncertain: when water is not sup-plied to the water tank, villagers are leftstranded. In western Rajasthan, thedominant pattern of settlement is indhanis where only a few families of therevenue village reside near the watertank and the majority live a few kilometres away. Thus even when wa-ter does come to the village tank, those

who live in the dhanis (hamlet) still have

to spend some hours each day to get tothe village and fill water. Those who areable to afford it, pay for the water to beput into their private tankas . The costis, however, quite high in Biramsar,people were paying up to Rs. 250 permonth for their water supplies.

2.4.3 Occupational Structure

A large proportion of the population of

Rajasthan relies on agriculture and re-lated activities such as cattle rearing andcollection of forest produce. This is in-dicated by the secondary data on theoccupational structure in Rajasthan andborne out by the present field study. Thedata indicates that more than 60 per centof the population still considers self-em-ployment in agriculture as their primary occupation, though this does not mean

that agriculture is their sole activity. There isnt much variation in this pro-portion either by caste or region.

The share of the population earning aliving from other forms of livelihood is

11 In the same village one womans husband was suffering from TB and diarrhoea, and she did not have any animal on which to fetch the water. She walked 3 hours each day in the sand dunes to the nearest water sourceto fetch one matka of water on her head for the days needs.

Four years of consecutive drought have lowered the water table, and some handpumps have gone dry as a result


19/50

11

quite small. For example, only 5 per centreported being self-employed in non-ag-ricultural occupations. This often meantthat they were employed in their casteoccupation e.g. working as Suthars ,

Luhars , Mochis , Sunars, Naisetc. Theother main category was that of casuallabour, where again only 5 per cent re-ported it as their primary occupation.

This is indicative of the lack of devel-opment of occupational choices otherthan agriculture and traditional caste oc-cupations for the rural population. How-ever, it should not be taken as evidenceof lack of diversification in the occu-pational structure. Of the 397 house-holds visited, as many as 379 house-holds (95 per cent) reported a second-ary occupation as well.

The diversification of livelihood strat-egies is best studied at the individuallevel. Only when all members of ahousehold report the same occupationdoes it indicate a lack of diversifica-

tion of the occupational structure.Members of a particular householdcould diversify in two ways: a) eachperson does a different job, or b) thesame person may do different jobs atdifferent points of time.

Thus, there may be houses where onemember has a regular job whereas theother members concentrate on agricul-ture. Or, those who report agricultureas their primary occupation may migratein search of work during the lean agri-cultural season. Such examples are notdifficult to find girls who help in agri-cultural operations, or in grazing cattle,

may supplement the household incomeby doing kasidakari (embroidery).

In households where traditional casteoccupations were reported as the primary occupation, close to 60 per cent reported

being engaged in agriculture as well.Even for those with regular jobs, wherethe flexibility to move between jobs ispresumably the least, approximately 40per cent reported agriculture as their sec-ondary occupation.

2.4.4 Ownership of Land

It is important to note that the data onlandholding is not entirely reliable. Many

people seemed to be afraid to state thetotal area of their land holdings. Thismust be due to the fear that if they wereseen to be rich then they would be de-prived of any potential benefits thatcould come as a result of this survey. Inone village we were able to get the landrecords from the patwari . In other cases,two members of the same household

were interviewed separately on different

days, or at least at different times to crosscheck the answers, or the same person

was asked some of the questions a sec-ond time, on another day to verify ear-lier claims. Quite often, these cross-checks revealed that there had been un-der-reporting of assets and sometimeseven output of cereals.12

A majority of the sample households(38 per cent) are marginal and smallfarmers, with land holdings of less thanone hectare. The second largest group,comprising nearly a quarter of thehouseholds, have land holdings of morethan 4 hectares. One must bear in mind


In households where traditional caste occupations were reported as the primary occupation, close to 60 per cent reported being engaged in agriculture as well

12 These crosschecks were not done in a systematic manner therefore, one cannot say with certainty the extentof measurement errors. Similar crosschecks were also applied wherever possible for the other data collected.

The most glaring differences were in the case of land owned.


20/50


that in the western part of the State,population density is low and in gen-eral, the size of land holdings tendsto be higher. However, these largerland holdings are accompanied by much lower productivity of land in theregion. Thus a high degree of regionalinequality in the distribution of landholdings is offset by the differences inproductivity of land. Table 3 (p. 33)clearly shows that the land holdingsin the four villages of westernRajasthan are the largest, whereasholdings in the villages of Udaipur arethe smallest.

To look at inequality in the distributionof land as a measure of inequality in

wealth, we would have to delve furtherand study the distribution of land acrosshouseholds. However, since the quality

of land data is slightly suspect, this hasnot been attempted here. To give an in-dication of such intra-village inequali-ties, the average size of landholdingsfor various caste groups has been shown

in Table 4 (p. 35). This table providesa clear indication of the favourable po-sition of the other backward castesgroup on the one hand, and of the smallland holdings for the scheduled tribesgroup on the other. However, one mustbear in mind that this category includesa large number of castes, ranging fromquite prosperous farmers belonging tothe Jat and Bishnoicastes, to extremely

poor ones such as the Kumbhars ,Luhars , Nais , Sunars , etc. For example,the average holdings vary from 45 bighas for the Bishnois, to only 2 bighas forthe Nais .

A high degree of regional inequality in the distribution of land holdings is offset by the differences in productivity of land


21/50

13

The Severity of the Drought of 1999-2001The Severity of the Drought of 1999-2001

There are several ways by which the se- veri ty of the drought of 1999-2001could be evaluated. Since Rajasthanseconomy is primarily agrarian, it makessense to look at losses in the agriculturalsector by studying various parameters.

The ones that were included the ques-tionnaire related to the area sown and

the fall in agricultural output. The west-ern part of the State, as mentioned ear-lier, is arid. As a result, people here rely on livestock rearing as much as they doon agriculture.13 Thus it is important tolook at changes in the aggregate live-stock wealth of the sample households,as well as changes in the compositionof their livestock assets. Since migrationoutside the village, and in some cases

even outside the State, is not only a cop-ing mechanism to seasonality anddrought in the region, but an importantsource of livelihood, it will also be dis-cussed briefly.

3.1 Agriculture

Land conditions in the sample villages vary considerably in the four districts.Some of the agro-climatic differences

have already been discussed. Here wedescribe the main crops and their pro-duction levels, as well as losses in eachof these regions.

Barmer and Bikaner both have sandy soilthat is suitable mainly for bajra . Agricul-

tural operations in this region rely on themonsoon rains. Consequently the kharif (monsoon) crop of bajra , the staple, isthe main crop of the year. The only ex-ception here is Dulmera Station wherebajra cannot be sown because the rainscome very late. Thus no cereals aregrown here in the kharif season. How-

ever, there are variations within thesample villages as well. Irrigation facili-ties are available in two of these villages,Kharad (Barmer) and Dulmera Station(Bikaner). In Kharad, some fields are ir-rigated with groundwater, while inDulmera Station a few fields are in thecommand area of the Indira Gandhi

water canal.

Udaipur has fertile soil, but fields are not

level. Here again,kharif is the main crop,though a small proportion of the popu-lation do have irrigation (river water andgroundwater) and are able to sow a rabi crop as well. The staple in this region ismaize, supplemented with rice. Those

who do grow a rabi crop sow wheat.

Finally, Jaipur probably has the most pro-ductive soil of the four sample districts,in addition to having level fields. Herethe norm is to sow both the kharif andthe rabi crops. Jaipur does receive somerain in the winter as well, so the secondcrop is not entirely dependent on irriga-tion facilities. The main cereals grownin the monsoon are bajra and jowar ,

13 Rajasthan has the largest livestock population in the country, and a substantial part of the States economy consists of earnings from its cattle wealth (Rajesh, 2000).

SECTION III

It is important to look at changes in the aggregate livestock wealth of the sample households, as well as changes in the composition of their livestock assets


22/50


whereas wheat and jow are the main ce-reals grown in the rabi season. The fieldsthat are irrigated are irrigated by tubewells and dug wells.

Other Kharif crops: The other impor-

tant kharif crop is gwar . It is used mainly as fodder and sold on the market. Thekharif pulses include moong and moth in

wes ter n Ra jas tha n and Jaipu r andmoong , tuar and urad in Udaipur. Theseare also generally grown for self-con-sumption, though it is not uncommonfor them to be sold in good agriculturalyears. In Dulmera Station, people areable to sow groundnut in the monsoon,

which is harvested in early December.Some people in Biramsar (Bikaner) alsogrow groundnut. Sesame ( til ) is anothercrop that is widely grown in the regionfor its oil.

Other Rabi Crops: Apart from the rabi cereals wheat and jow chana , jeera andmustard seed are the other rabi crops. Jeera (cumin) is grown extensively in Kharad

and is a highly remunerative crop. Chana is mainly grown for self-consumption.

3.1.1 Crop losses

As discussed above, the rainfall in 2001 was quite close to normal and that yearhas been considered as the benchmark in evaluating the area sown and outputlevels in 2000, which was a droughtyear. Table 5 (p. 36) shows the short-

fall in area cultivated for the main ce-real crops. A small shortfall in the areaunder cultivation is not necessarily agood sign. This could just mean thatfollowing a good rain at the beginning of the monsoon season, farmers under-took their sowing operations expecting

that the good rain would continue. If,after an initial good spell, the rains fail,it could mean a double blow for thefarmers not only do they lose theircrops, they would also have to bear the

costs of the sowing operations whichinclude purchasing the seed, expendi-ture on ploughing the fields, and, insome cases, the cost of fertilizers andpesticides. Over and above this, is theimputed labour cost of their own timespent in the sowing operations. There-fore, the shortfall in area sown must beconsidered in conjunction with the out-put data.

An example of a small fall in area sownaccompanied by a large fall in total out-put is clearly demonstrated in the caseof bajra . Though area under cultivationof bajra fell by just 4 per cent, the fall inoutput was to the tune of 70 per cent!

This is in spite of the fact that bajra isamong the most sturdy cereals in termsof being able to tolerate a high level of moisture stress. Table 5 (p. 36) suggeststhat the largest decline in output was forbajra , along with rice, jowar and pulses.

Makka seems to have been the only cropthat managed to survive in spite of alarge decline (nearly 60 per cent) in thearea under makka .

Overall, one can see that the failure of the rain in the year 2000 caused a largedrop in the production of bajr a and

wheat, both cereals that are widely con-sumed by the population. On the otherhand, makka , which is grown only inUdaipur, did not see much fall in pro-duction. Pulses, which are an importantpart of peoples diet, also saw a drasticfall in output.

If, after an initial good

spell, the rains fail, it could mean a double blow for the farmers not only do they lose their crops, they would also have to bear the costs of the sowing operations


23/50

15

Most households reported using approxi-mately 100 kg (one bori ) of cereals eachmonth. 14 Using this as a benchmark, wecan look at the average production of cereals per family to see how many

months their home-produced cereals would last them. This also gives us ameasure of their vulnerability. If they areable to produce enough to last them untilthe next harvest (approximately 12 boris )then they are unlikely to be badly affectedby fluctuations in the market price of cereals that may result from a generalshortfall in total output of cereals. Indeed,many families that were purchasing from

the market complained about the priceof bajra having increased from Rs. 3 perkg to more than Rs. 7 per kg in a few months; or from Rs. 5 per kg more thanRs. 10 per kg in the case of wheat.

Using this criterion, for the good year(i.e. 2001) we find that the worst hit vil-lages are Badli and Birothi (Udaipur) andDulmera Station (Bikaner) in whichpeople were able to get only about 3-4months consumption requirementsfrom their own fields and had to rely onthe market for the remaining months. Atthe other end of the spectrum we havethe villages of Jaipur, where farmers

were not only able to produce enoughto last them the entire year, but in thecase of Morda, had enough for 2 yearsconsumption needs. The remaining three

villages produced enough to last themfor 9-10 months.

When we look at the same figures for2000, the greater vulnerability is imme-diately obvious. Even a village like

Morda, which in the normal year pro-duced enough for two years consump-tion, could only produce enough forabout 10 months in 2000. The worst af-fected were the villages of western

Rajasthan, which were able to produceenough for nearly 10 months in a goodyear but produced only between one tofour months consumption in 2000. Theproduction of makka did not vary muchbetween the two years as mentionedabove (Table 5, p. 36). There is not muchdifference in the vulnerability of thetribal villages of Udaipur betweendrought and normal years. For them,

vulnerability remains high under bothcircumstances. Finally, for Dulmera Sta-tion, the source of vulnerability is not thedrought, but the cropping pattern in the

village. As we noted above, the bajra crop cannot be sown there because therains come late to the region; while somefields have access to water from theIndira Gandhi Canal, the supply from theCanal is irregular and unpredictable. In

any case, only a small proportion of fieldsdo actually benefit from the Canal.

3.2 Livestock

The main livestock wealth in the sample villages was in the form of cows, buffa-loes, goats, sheep and camels. Whilecows and buffaloes are important mainly for their milk and to a lesser degree asdraught animals, sheep and goats are

reared for their wool and meat as well asfor their milk. Camels are found only inthe sandy western regions, where they are quite useful as draught animals as

well as for transport.

14 The median consumption of cereals is 100 kg whereas the average household consumption of all cereals is114 kg. This has been calculated by summing up the total monthly consumption of wheat, rice, bajra , makka and jowar .

The Severity of the Drought of 1999-2001

If they are able to produce enough to last them until the next harvest then they are unlikely to be badly affected by fluctuations in the market price of cereals that may result from a general shortfall in total output of cereals


24/50


If one looks at the total composition of livestock wealth in 1999, one finds thatthe smaller animals, i.e. goats and sheep,predominate. Goats and sheep constitute49 per cent and 22 per cent respectively of the total holdings. The only changeover the three drought years is that theshare of cows in total holdings increasesto 24 per cent, pushing that of sheepdown to 18 per cent.

Though goats are believed to be thesturdiest of all livestock, Table 6 (p. 37)shows that the population of goatsnearly halved in the three years of drought (between 1999 and 2002). The

largest losses were in the holdings of sheep (53 per cent).

Death has been the main cause of de-cline in livestock population. This is truefor cows and buffaloes, goats and sheepas well as camels. Between half to two-thirds of the losses can be attributed todeaths, most of which were due to ill-ness or lack of fodder.15 In many cases,the illness was due to the poor quality of fodder upon which the livestock wasforced to subsist.

The second largest cause of loss hasbeen due to sales. Here too, as in the caseof deaths, sales were mainly due to thelack of fodder, or because fodder wastoo expensive. The need to meet daily consumption expenses also figured as amajor cause for livestock sales. Between

5-10 per cent of the losses were classi-fied as being due to other causes, in-cluding thefts, partitioning of house-holds, gifts, etc. Some of these losses

were offset by an increase due to pur-chases or because of new births.

3.3 Coping Strategies

Apart from diversifying their sources of

income, households resort to variousother coping strategies to survive thedrought. Some of these are best viewedboth as coping strategies as well as strat-egies to diversify livelihoods. The pri-mary coping strategies discussed hereare migration, credit, distress sale of as-sets, and change in pattern of food con-sumption and level of consumption.

3.3.1 Migration

Nearly half of all households surveyedreported that at least one member of their family had to migrate in the sum-mer of 2001 (before the normal mon-soon) in search of work. In nearly half of these households, the reasons formigration were related to the droughtand were not instances of usual sea-sonal migration.16

When one looks at individual data, oneneeds to compare migration in the sum-mer of 2001 with migration in the sum-mer of 2000 to ascertain whether morepeople were driven out of the local ar-eas in the second year of drought. If thenumber of migrants increased in 2001compared to 2000 across the samplehouseholds and within each household,this would indicate that migration wasbeing used as a coping mechanism.

A study of migrants seeking work in thechowktis (urban daily labour market) of

15 It is quite likely that that the lack of water was as important as the lack of fodder. However, none of therespondents mentioned this when asked about the causes of deaths.

16 To distinguish usual seasonal migration from drought -related migration, the respondents were asked whether they migrated in that particular season every year or whether migration in the summer of 2001 was something new.

Death has been the main cause of decline in livestock population.Most of these deaths were due to illness or lack of fodder


25/50

17

Jaipur, did in fact find that the numbershad swelled quite considerably eventhough 2001 was closer to being a nor-mal agricultural year (Kerbart andSivakumar, 2002).

The incidence of entire families migrat-ing was quite rare in all the villages vis-ited. When this did happen, it was uni-formly a sign of great distress and im-plied hardship for those concerned. Thedata collected indicates that most often,these families seemed to get work only as casual labourers in the cities and vil-lages to which they migrated.

Migration to save cattle was also un-common. When people migrated tosave cattle, they would take the cattleof others in the village with them, sothat fewer people would need to mi-grate. In Morda (Jaipur), in some Gujjar families, one son took the family cattlealong with that of others in the village,and set off towards the Madhya Pradeshborder in the hope of saving them. Quiteoften, those who had sent their cattleaway said that they were as good as lostsince they would not be returned tothem. This seemed to imply that theperson who had taken the cattle wouldsell them but tell the owners that they had died.

The most common form of migration was when one male member of the fam-ily migrated or commuted to the nearby

city, where he would undertake one of arange of activities. Migrants would beemployed in shops and restaurants, or

would work as casual labourers. In Baasri Jogiyan (Jaipur), most young men go to Jaipur city each day, where some havejobs at shops and similar establishments,others work in Sanganer at dyeing and

printing factories, and yet others to gochowktis to work as masons, porters ascasual labourers. Migration from Morda(Jaipur) was not a very common feature.

Those who did migrate went to the stone

quarries in Ajmer district (Morda is inthe southern most part of Jaipur andquite close to the Ajmer border).

Seasonal migration is most common in western Rajasthan. In the villages of Barmer, seasonal migration has be-come a regular feature of peoples lives.Most people (probably even all) mi-grate to Gujarat to earn a living. Theirearnings from migration are not insub-

stantial, and generally employmentseems guaranteed for the period of mi-gration. For example, in Sukaliya peoplego to Valsad to work as caretakers onmango orchards in the summer monthsand are able to earn enough for theentire year. But there are other seasonalmigrants who are not so well off. They are faced with uncertainty regarding employment and other adverse condi-tions as well. In Bikaner, men wouldmigrate north towards Punjab to work as agricultural labourers in the wheatharvesting season.

The duration of migration seems directly correlated to the distance to the nearestcentre for employment. Thus, in places

where the city is easily accessible (Baasri Jogiyan), or there is a big market centre

nearby (Kharad) the incidence of sea-sonal migration was lower than in vil-lages like Sukaliya, where there is nonearby employment option available andconsequently people go for the entiresummer season. In Dulmera Station, theauthor met one family that had just re-turned from north Rajasthan where they


The most common form of migration was when one male member of the family migrated or commuted to the nearby city,where he would undertake one of a range of activities


26/50


had been working as agricultural labour-ers in conditions that seem to have beenquite miserable camping in the openair, in the middle of nowhere, with noshelter, without social support in terms

of friends and family, or any guaranteeof employment.

3.3.2 Credit

Resorting to credit is known to be oneof the most common ways of tiding over shortages that occur in times of drought. In this sample of 397 house-holds, 247 (62 per cent) had to bor-row money in the drought year. But we

must bear in mind that householdsborrow for a variety of reasons, not allof which are related to drought. Forexample, borrowing for medical ex-penses or to meet marriage costsshould not be attributed to drought, assuch borrowing can occur in normalyears as well. Besides this, householdsmay borrow to accumulate assets.

In our sample, a majority of the house-holds (50 per cent) that borrowed did soin order to meet their food consumptionneeds (see Box). The other major rea-son why people borrowed was for agri-cultural expenses, which included thepurchase of seeds, payment for tilling theland with tractors, digging wells, etc.

Some of these could be drought-related e.g. some households may haveneeded to use tractors to till their fieldsbecause their own draught animals died,or were sold, in the wake of the drought.

The reasons given by nearly three-quar-ters of the borrowing households couldbe linked directly to the drought. Theseinclude borrowing for food consump-tion needs, for other daily consumptionneeds (a very small category, with only 18 households), to buy fodder and to mi-grate. The remaining 25 per cent bor-rowed for reasons that cannot be linkedso easily to the drought these include

medical and marriage expenses, and ex-penses for agricultural operations.

More OBCs borrow as a proportion of their social group than do members of any other group; there is not much dif-ference in the proportions of SC andother households who borrow (Table 8,p. 39). The smallest group is that of STs.

When we look at the reasons for bor-

rowing across the caste groups, the pat-terns do not vary much across house-holds borrowing for daily needs is fol-lowed by borrowing for other reasons.

Table 8 (p. 39) provides some evidencethat borrowing among the better-offhouseholds is higher than among poorerhouseholds. This could indicate thathouseholds are borrowing to accumulateassets. However, we have already seen

that daily consumption expenses formthe main reason for borrowing. So itis more plausible to assume that thelarger share of better-off householdsamong debtors reflects their greater cred-itworthiness. For example, for ST house-holds, borrowing from institutionalsources is limited by economic and

Main Reasons for Borrowing

Of the 247 families that did borrow, *

123 borrowed to meet their food consumption needs81 mainly for agricultural operations46 to buy fodder42 to meet medical expenses30 to meet marriage expenses18 to meet other consumption needs1 to migrate* The total for all the households does not add up to 247 because the categories are not mutually exclusive, i.e., some households borrow for more than one reason.

Resorting to credit is

known to be one of the most common ways of tiding over shortages that occur in times of drought


27/50

19

non-economic factors (these includesmall land holdings, greater social dis-tance from institutional mechanisms,etc.). Borrowing from friends, relativesand family members is equally diffi-

cult as they are likely to be in a simi-lar situation. Other informal sourcesof credit (such as the local bania ) areknown to dry up in times of distress. 17

As a result, even though the share of STs among borrowing households issmall, it is due more to a restriction inthe supply of credit rather than a lack of demand. This also explains to someextent the small proportion of borrow-

ing households in Udaipur, which is where most of the tribal populationof the sample is concentrated.

The main sources of credit were the lo-cal bania (39 per cent), followed by rela-tives and friends (27 per cent). Only 8per cent of households borrowed frominstitutional sources. The larger shareof informal sources of credit is prob-ably linked to the fact that most house-holds are borrowing for their food con-sumption needs, for which credit is notavailable from formal sources. Themain sources for borrowing for agricul-tural purposes were from institutionalsources (23 per cent) as well as infor-mal sources. Relatedly, one finds thatpeople borrowing from informalsources are being charged usuriousrates of interest. The average monthly interest rate is 2.4 per cent, and in somecases, as much as 10 per cent! Thehighest rate of interest is charged by banias . The advantage of borrowing from relatives and friends is that they

sometimes provide interest-free loansfor a flexible time period, which are notavailable elsewhere.

3.3.3 Hunger

A quarter of the houses (101 out of 397)reported that they had been skipping meals around March 2001. Nearly half (47 per cent) of these were in Udaipurdistrict.

When food was short in the house:

105 homes reported eating less thanusual.

85 reported borrowing money, buying food on credit.

18 sought the help of their neighboursand relatives.

13 resorted to changing their foodbasket.

There has been a debate on the se-quencing of coping strategies and theirrelation to the level of distress.18 Jodhahas argued that people will try to pre-serve their holdings of key productiveassets rather than protect a certain stan-dard of consumption. This is in directopposition to Corbetts (1988) position.Chens (1991) research in Gujarat vali-dated Jodhas proposition, and indeedsome evidence for this can be found inthe sample villages as well. The propor-tion of houses that had to resort to dis-tress sales of productive assets (mainly land and cattle) is much smaller than,for instance, the proportion of house-holds that report skipping meals. Chenlists the following main coping strate-gies in their rough order of sequencing:adapting or diversifying activities,

17 See Chen (1991) for more on this. When asked if they had borrowed during the past year, many householdscomplained vehemently that the bania no longer gave them credit.

18 See Chen (1991), Jodha (1975, 1991), Corbett (1988) and Chambers (1981) for more on this debate.


People try to preserve their holdings of key productive assets rather than protect a certain standard of consumption


28/50


reducing or modifying consumption, re-ducing commitments, participating inrelief programmes, borrowing, migrat-ing, mortgaging or selling assets andother drastic measures such as distress

migration of the entire household andbreakdown of families.

This study does not look at sequencing of coping strategies over a period of time. Therefore it is not possible to say much about sequencing. However, wecan get some clues about sequencing by looking at the proportion of householdsresorting to the various strategies at aparticular point in time. If the propor-tion of households using a strategy de-clines as we move from strategies ear-lier in the sequence to strategies later on,this may confirm the sequencing of thosestrategies.19 In Table 7 (p. 38), as we movedown the column, we can see that theproportion of households resorting tothat particular strategy falls.

3.3.4 Vulnerable Groups

Based on the discussion above, we try to highlight some of the particularly vul-nerable groups. In Table 8 (p. 39), someindicators of vulnerability have beentabulated for three socio-economicgroups. These three groups are based onland ownership, primary occupation andcaste group. The most vulnerable group

when we use land ownership as the ba-sis of classification, is predictably thesmall or marginal farmers. Whether we

look at incidence of skipping meals orseeking work at relief sites, we find agradual decrease in proportion of house-holds as we move up the land-owning scale. For credit, we see further evidencesupporting the hypothesis that thegroups that do borrow are the wealthierones because they are perceived as cred-itworthy debtors, and that the low pro-portion of borrowing households among the small and marginal farmers may berelated to them being perceived as un-reliable borrowers.

When we look at the same indicators of vulnerabili ty by occupation, there is

some evidence to show that agriculturallabourers are the worst affected. This isnot only in terms of having to sleep hun-gry or because the proportion of house-holds seeking work on relief sites is very high, but also because the proportion of agricultural labourer households thathad to sell or mortgage land is the high-est among the five occupation catego-ries. The other category that looks quite

vulnerable by these indicators is that of casual labourer households.

When one looks at vulnerability among the various caste groups, the Scheduled

Tribes seem the worst off. The Sched-uled Castes are better off than them, butnot by much. This is further substanti-ated by the fact that they are also over-represented in the class of marginal andsmall farmers, a category that is also very

vulnerable.

19 This assumes that the most frequent strategies are also those that come first, which may not always be true. These results should therefore be interpreted with caution and as something to be explored further.

The most vulnerable group when we use land ownership as the basis of classification, is predictably the small or marginal farmers


29/50

21

Government PolicyGovernment Policy

This section presents a brief but criticalreview of the governments droughtpolicy. It highlights some areas wherechange is possible, without giving con-crete recommendations.

Given that large parts of the country aredrought prone, the government has ahistory of policies and programmes todeal with its consequences. This history can be traced back to the Famine Com-mission Reports of the colonial era.

These identified the lack of work andaccess to food as the main cause of suf-fering, but stressed the development of irrigation facilities as the most impor-tant weapon against drought-related suf-fering.20 Based on this evaluation, the co-lonial governments solution to droughts

was two-pronged:

Generation of work for the able-bodied with gratuitous relief forthe others.

The development of agriculture inde-pendently from the monsoon.

The governments response can be stud-ied under two heads drought proofing

and drought mitigation. The two are dif-ferentiated by the duration for whichthey are deemed to have an effect onpeoples lives. Thus, drought mitigationmeasures are generally treated as short-term measures that evolve to deal witha crisis that might develop if these

measures are not undertaken. The food-for-work programmes fall under this cat-egory of drought policy. Drought proof-ing refers to policies and programmes thatare put in place with the hope of evolv-ing a long-term solution to the problemof drought. Such measures might includedevelopment of irrigation facilities or the

development of alternative sources of livelihood so that dependence on agri-culture itself is reduced. The two typesof measures cannot always be differen-tiated in a clear manner for example,measures for mitigation include the cre-ation of durable assets. Until recently,drought mitigation measures formed thelarger (in terms of expenditure) and moredominant part of the governments strat-

egy to deal with drought. However, therehas been a gradual shift away from miti-gation measures to the more perma-nent proofing measures.

4.1 Drought Mitigation

Government policies for drought miti-gation can be viewed as its short-termresponse to drought. The main respon-sibility of dealing with drought relief

lies with the State Government as it istreated as a State subject. The Scarcity Manuals play a pivotal role in the fram-ing of drought mitigation policy. Theirrole is important not only in defining

what counts as a drought, but also inprescribing when a drought will be

20 See Drze (1990: 15-35) for a description of the evolution of famine policy.

SECTION IV

There has been a gradual shift away from mitigation measures to the more permanent proofing measures


30/50


declared and how relief will be allo-cated. For example, in the allocation of relief work quotas, villages that are de-clared famine-affected are given pri-ority over villages that are declaredscarcity-affected.

The Manuals are very thorough, but werelast revised in 1968. Since then, therehave been many changes in the economy,in politics (e.g. the introduction of

panchayati raj institutions) and in society. The Manuals, as they exist, do not re-flect this changed socio-political reality.

They were designed to provide objec-tive criteria on which to base the dec-

laration of droughts. In practice, how-ever, the declaration of a drought in-

volves a lot of politics as well and thereare various ways in which these criteriacan be subverted. It is possible that acareful revision may still be the best way to deal with drought-related policy is-sues. The discussion that follows furtherhighlights the need for an immediate anddrastic revision and design of the Scar-

city Manual of Rajasthan.Reliance on the Scarcity Manual isfraught with many problems. First , theManuals have been designed with theunderlying assumption that droughts arenot a regular feature of the economy. Itfocuses on timely identification of droughts so that they do not develop intoa crisis. In the first section, we saw thatjudging by various indicators, droughtsare actually a recurring feature of theState, affecting some part or anothereach year, and manifesting themselvesin the form of shortages of drinking

water, crop losses, etc. Thus droughts

should not be treated as separate and/or occasional phenomena to be dealt

with as and when they occur, but ratheras another constraint in the design of policies. They are a chronic problem that

should be addressed as such rather thanin an ad hoc manner as is the case atpresent. All development policy mea-sures should be designed with this inmind. The ad hoc treatment of droughtsmeans that the chronic problems that area direct consequence of droughts, suchas unemployment, underemployment,poor health conditions (especially undernutrition amongst vulnerable

groups), etc. are ignored by governmentinterventions. 21 One clear indicator of the permanent nature of drought con-ditions in Rajasthan is that it is difficultto separate peoples coping strategiesagainst drought from their coping strate-gies in general. Migration is a prime ex-ample of this people migrate not only because of droughts, but also on a sea-sonal basis. Decisions to migrate depend

on whether work is available in the vil-lage. People adjust their migration pat-terns to the nearby city or town to fit in

with their chances of getting employ-ment on the food-for-work site in theirown village.

Second , there are problems with the way the Scarcity Manual is used. Thoughthe Manual lays out other indicators thatshould be included in ascertaining theexistence of a drought, in practice mostgovernments have come to rely almostexclusively on the girdawari report, ig-noring other indicators such as cattledeaths, migration, etc. Cattle deaths are

21 One finds the government giving out contracts for various jobs that are as easily done by manual labour tomechanised firms, while people in the same area clamour for work and sometimes in extreme conditions, evendie of starvation.

Droughts should not be treated as separate and/or occasional phenomena to be dealt with as and when they occur, but rather as another constraint in the design of policies


31/50

23

especially pertinent in the case of Rajasthan, which has the largest live-stock population in the country. For alarge proportion of the rural population,livestock is a more important source of

income than agriculture. Third , undue reliance on the Manualcauses unnecessary delay in the com-mencement of relief work. Since the as-sessment of land use requires that cropsbe close to harvest and since, in principleat least, each field has to be visited by the patwari , it becomes a long drawn outand time-consuming process, which canbe started, at the earliest, in August. Be-

cause of this, the declaration of droughtis often grievously delayed. For example,in 1999-2000, the second consecutiveyear of drought, the government took until January 2000 to declare it. To makematters worse, relief work does not al-

ways begin immediately after the decla-ration of a drought. Quite often there isa further lag of upto two months beforerelief work is initiated. In fact, even in2000-1 relief work began only in Febru-ary 2001 though the drought was declaredin November 2000.

Also, relief work is stopped as soon asthe first monsoon showers come. Thegovernments claim in this case is thatpeople no longer need work becausethey prefer to work on their fields any-

way. There is some truth to this; but the

period from the beginning of the mon-soon until the crops are harvested is alsothe period of maximum distress, sinceby then people have exhausted whateverlittle stocks they might have had from

previous years.22 Since, as we noted ear-lier, people do diversify their sources of livelihood within a household, the con-tinuation of relief work would allow some members to continue using it as a

source of income to tide over this diffi-cult period, while others would be freeto work on their fields.

Fourth , the sincerity with which a patwari may carry out his duty in themeasurement varies tremendously.Quite often the patwaris do not live inthe region they are responsible for. Many are known to remain absent from theiroffices for long periods of time, are cor-

rupt, etc. The girdawari report involvesthe patwari visiting all the fields in his

panchayat to record the area sown, andto provide an estimate of the expecteddamage to the crops. Even if they wantto carry out these measurements prop-erly, the sheer magnitude of the task bears them down, especially given thefact that the patwaris have numerousother responsibilities as well. Besidesthis, in some villages at least, it has beennoticed that when the villagers areaware of the significance of over-stat-ing their agricultural losses in order toget additional relief, the patwaris areoften under a lot of pressure to exag-gerate the losses.23

The Scarci ty Manual also details thetypes of work that should be started in

order to provide relief. This list includestasks that would create durable assets.However, government policy has devi-ated from the Manuals guidelines. Es-pecially since the implementation of

22 This has been referred to as the hungry season by Longhurst (1986). See also Chen (1991: 106). In this study as well, many households did say that the worst season in terms of food availability was the monsoon.

23 This was noticed in the gram sabha meeting of one of the villages that the author did fieldwork in. Many people were asking the patwari to be meherbaan (generous) in reporting losses in their village.

Government Policy

Since people diversify their sources of livelihood within a household, the continuation of relief work would allow some members to continue using it as a source of income, while others would be free to work on their fields


32/50


Fifth Pay Commissions recommenda-tions, State Governments have been fac-ing a severe financial crisis. This resourcecrunch, combined with the objective of providing employment to as many people

as possible, has meant that the Relief Department sanctions only thoseprojects where the materials componentin the total cost is close to zero. As aresult, it is common to find that roadsare being laid in sand dunes where theprevious days work could be wiped outin a single night. Nadis (ponds) dug inany particular year to harvest the follow-ing years rainwater often do not survive

that long.On the other hand, since Panchayati Raj institutions have been put in place, a

gram sabha takes place four times a yearin each panchayat . At these meetings, theelected representatives of the people areasked to submit proposals for theprojects to be undertaken in their

panchayat . These proposals are sent reli-giously to the block administration aftereach gram sabha meeting. However, thelist of sanctioned works that comes back from the district or block administrationrarely includes the proposed work. Thisis partially due to the above mentionedresource crunch.

There are several problems with this ap-proach. First, the top-down approachmeans that projects that do not always

reflect peoples needs and priorities getprecedence, leading to wasteful expen-diture. Second, peoples participation inand ownership of the relief work suf-fers. It is widely recognized that thesetwo factors are vital in ensuring thesuccess of interventions. Third, the

top-down approach is especially detri-mental when it is accompanied by therhetoric of peoples participation.Peoples faith in the panchayati raj insti-tutions suffers when their proposals and

demands are not considered. This hasthe effect of reducing their participationin these democratic local institutions.

Though drought is a State subject, theCentral government does play a very important role. Specifically, the financesfor dealing with a drought are providedmainly by the Central government. TheCentre has two schemes for dealing

with calamities such as droughts,floods, earthquakes etc. These are theCalamity Relief Fund (CRF) and theNational Calamity Contingency Fund(NCCF). Both were started in 2000-1and have a duration of five years. TheCentre contributes as much as 75 percent of the funds to the Calamity Re-lief Fund (CRF). In the case of NCCF,the entire contribution is made by theGovernment of India, and allocationsto the States are made when the Na-tional Centre for Calamity Management(NCCM) finds that the State govern-ment needs to spend in excess of theCRF to meet the calamity. 24

This pattern of funding relief leads todelays of its own. The State governmenthas to prepare a report of the losses andprovide the Centre with estimates of

funds required. Once this is received,the Centre sends its team to the Stateto determine the validity of the claimsand only when the Centre is satisfied,are the funds released. Smooth imple-mentation of relief activities requiresthe existence of good Centre-State

Peoples faith in the panchayati raj institutions suffers when their proposals and demands are not considered

24 See http://ndmindia.nic.in/documents/document.html


33/50

25

relations. However, the case of the re-cent drought in Rajasthan illustrateshow relief operations can suffer on ac-count of political rivalries.25

4.2 Drought Proofing

Drought proofing policies can be of twokinds: those aimed at drought prevention ,i.e. countering the trigger event, andthose aimed at reducing the impact of drought on people. Examples of theformer include attempts to predict theextent of rainfall or devise early warn-ing systems, and extend to reforestationprogrammes (based on the assumption

that the failure of rain is related to en- vironmental degradation).26

Within the second category of respon-ses, i.e. those aimed at reducing theimpact of drought, policies tended tofocus too much on agriculture and

within that, on the development of ir-rigation. This has led to the imple-mentation of large-scale irrigationprojects,27 reforestation programmes,etc. From independence until the1970s the reliance was almost solely on such large-scale irrigation projects.Over the years, there has been a gradualshift away from such policies. However,the focus on the agricultural sector hascontinued. Later programmes such asIRDP and TRYSEM were included indrought proofing efforts.

There have been other policies aimedat reducing the impact of drought onpeople. These policies include measures

that would strengthen peoples position within their existing livelihood patterns such as agricultural extension services,

veterinary services, access to credit andmarkets, etc. Their failure can be traced

to their flawed design as well as to prob-lems in implementation. Drought-resis-tant varieties of seeds, varieties withshort stalks, short ripening period, etc.have also been tried. What seems to havebeen neglected is the already existentdiversity in peoples livelihood strategiesand the need to focus on them.

4.2.1 Government Programmes

In its attempt to drought proof thedrought prone areas of the country theCentral government has undertaken

various programmes such as theDrought Prone Areas Programme(DPAP) and the Desert DevelopmentProgramme (DDP).

The DPAP was launched in 1973-4 andcovers 180 districts spread over 16States. Similarly, the DDP covers 40 dis-

tricts in 7 States and was launched dur-ing the Fifth Plan in 1977-8. Theprogrammes undertaken under theDPAP and DDP include pasture devel-opment, afforestation, and water re-sources development. The unsatisfac-tory progress of the programmes led toa review in 1993 by the HanumanthaRao Commission. As a consequence,there was a concerted move to bring de-

cision-making closer to the beneficia-ries of the programmes as well as carry out decentralisation. This change came

25 For more on Centre-State relations and the fiscal arrangements related to droughts see Mathur and Jayal(1993: 53-62).

26 Wijkman and Timberlake (1984) note that policy responses tend to focus on trigger events even though wehave little control over them.

27 This emphasis can be traced back to British India (Mathur and Jayal, 1993: 31). Baviskar (1995), also commentson this tendency of government policy (see pp. 26-22). See Jodha (1990) for the changing priorities in policy.

Government Policy

What seems to have been neglected is the already existent diversity in peoples livelihood strategies and the need to focus on them


34/50


together with an exclusive focus on watershed program-mes. The recom-mendations came into effect in 1995.

They also changed the financial ar-rangements from a 50:50 partnership

between the State and the Centre to a25:75 sharing arrangement.

The long term objectives of the Centrethat have been laid out in the Contin-gency Plan for Drought 2000, GOI, 28 in-clude the development of drought-resis-tant varieties of crops, a greater empha-sis on small and medium scale irrigationprojects, watershed development andrainwater harvesting, implementation of relief works that are sustainable, affor-estation, etc.

4.2.2 Government Policies

Water Policy

In discussing water policy, it is usefulto separate issues relating to the supply of water from issues relating to its de-mand. The problem of supply has to do

with the availability of water for vari-ous purposes, such as agriculture, drink-ing and sanitation. The problem of de-mand has to do with how the water isused: for instance, whether it is used

wastefully or unsustainably, who hasaccess to the water, etc. Governmentpolicy has tended to focus on the prob-lem of supply over that of demand, andhas concentrated on irrigation over

other needs. Recent interventions suchas water harvesting and watershed man-agement, though commendable, havealso continued to ignore the demandaspect of the water problem.

We have already noted that Govern-ment water policy has concentrated al-most exclusively on irrigation. The em-phasis here has been on large-scale irri-gation projects: during 1951-85, the In-dian government invested 64 per centof its total irrigation expenditure onmajor and medium projects, eventhough they are more costly than minorprojects.29 The cost of irrigating onehectare of land by a large project isRs. 19,310 compared to Rs. 4,520 whenirrigated by a minor irrigation project(Baviskar, 1995: 27). Even in the dry State of Rajasthan there have been at-

tempts to sell the idea of solving the water problem by linking the State withthe Indira Gandhi Canal in the north andthe Sardar Sarovar Project in Gujaratin the south. There is a need to moveaway from large scale irrigation to smalland medium projects which haveproved to be more efficient.

Since rainfall in India is concentratedduring three months, small-scale ef

Documents

Drought Proofing Rajasthan Imperatives Experience Prospects