Jump to ContentJump to Main Navigation
Social Security in Developing Countries$

Ehtisham Ahmad, Jean Drèze, John Hills, and Amartya Sen

Print publication date: 1991

Print ISBN-13: 9780198233008

Published to Oxford Scholarship Online: September 2011

DOI: 10.1093/acprof:oso/9780198233008.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 27 October 2020

Social Security and the Family: Coping with Seasonality and Calamity in Rural India

Social Security and the Family: Coping with Seasonality and Calamity in Rural India

Chapter:
(p.171) 5 Social Security and the Family: Coping with Seasonality and Calamity in Rural India
Source:
Social Security in Developing Countries
Author(s):

Bina Agarwal

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198233008.003.0005

Abstract and Keywords

This chapter draws upon existing studies focusing on the issue of social security essentially in terms of food security. It discusses a conceptual framework for understanding intrafamily relations and puts forward certain analytical propositions on the ways in which seasonality and calamity might impinge on these relations. The chapter examines the nature and effectiveness of the mechanisms adopted by poor agricultural families to provide food security for their members, against seasonality and calamity; the ways in which family coping mechanisms in the two contexts might differ; and intrahousehold inequalities in the burden of coping. It analyses how the family itself may begin to disintegrate in an extreme calamity such as famine, using the Bengal famine of 1943 as a case study, and highlights some aspects of appropriate external interventions.

Keywords:   social security, food security, intrafamily relations, agricultural families, seasonality, intrahousehold inequalities, Bengal famine

Introduction

How do poor agricultural families seek to cope with the problem of food insecurity associated with seasonal troughs in the agricultural production cycle? How do they cope with calamities such as drought and famine? How effective are the mechanisms they adopt? Does the burden of this coping fall equally on all family members or are some more equal than others?

Answers to questions such as these could be seen as necessary inputs in the designing of public policy, or other forms of external interventions for social security, on at least two counts:

  1. 1. To ensure that such interventions complement and strengthen rather than substitute for people’s own efforts in dealing with contingencies; that they are appropriate to actual and not assumed needs, and that people are seen as actors in the process of change rather than as passive recipients of aid and relief.

  2. 2. To address explicitly any intrafamily inequalities in the impact of contingencies. Rooted implicitly (and sometimes explicitly) in public policies and programmes are certain assumptions about the family, the responsibility of its members towards each other’s well-being and welfare, and their capability and willingness to fulfil that responsibility. For instance, government programmes that direct resources or employment mainly at men assume implicitly that the associated benefits will be shared equitably with the women and children. Since the effectiveness of these and similar interventions can depend crucially on the validity of such assumptions regarding family behaviour and intrafamily relations, clearly we need to know much more about how the family in fact operates as a unit, especially under contingencies, and how certain interventions will affect particular members. Also, assumptions made about the family in times of normality may not hold equally under severe crisis situations as in a drought or famine.

Drawing upon existing studies, this chapter addresses some of these questions, focusing on the issue of social security essentially in terms of food security. In specific terms, in Section 1, a conceptual framework for understanding intrafamily relations is discussed and certain analytical propositions on the ways in which seasonality and calamity might impinge on these relations (p.172) are put forward. Sections 2 and 3 examine the nature and effectiveness of the mechanisms adopted by poor agricultural families (namely, those of agricultural labourers and small cultivators) to provide food security for their members, against seasonality and calamity; the ways in which family coping mechanisms in the two contexts might differ; and intrahousehold inequalities (if any) in the burden of coping. Section 4 analyses how the family itself may begin to disintegrate in an extreme calamity such as famine, using the Bengal famine of 1943 as a case-study. The concluding Section 5 highlights some aspects of appropriate external interventions.

The issue of food security has been considered here in the narrow sense of ensuring maintenance of food consumption levels in the context of specific circumstances and contingencies rather than in the wider sense of ensuring adequate food intake for all. (This last would imply raising consumption levels substantially above existing ones for a large part of the population, and would impinge on the entire domain of development policy and anti-poverty programmes.) Broadly, a household is seen as effectively coping with seasonality and calamity where it is able to go through such periods without irreversible damage to the productive capacity of its members, or to its net assets position. The terms household and family have been used here interchangeably to connote commensal units.

The focus of this chapter will be on India, although supportive material from other South Asian countries will also be brought to bear on these issues. In general, material on family coping mechanisms in this region is rather limited, and even more so on intrahousehold aspects of coping. India’s ecological and social diversity again cautions against easy generalizations where the evidence is typically region-specific. This chapter therefore essentially attempts to provide pointers, and to raise some issues for debate and further research.

1. The Family, Seasonality, and Calamity

In examining the relationship between seasonality, calamity and the family, Sen’s (1983) conceptualization of family relations as embodying both cooperative and conflicting elements, and his entitlement approach to famine and starvation (Sen 1981), prove useful starting points. Sen conceptualizes the family as a unit in which various household arrangements (who does what and gets what goods and services) are fixed on the basis of the implicit bargaining strength of its members. He argues that there are many co-operative outcomes possible—beneficial to all parties compared with non-cooperation—but the different parties have strictly conflicting interests in the choice among the set of efficient co-operative arrangements. Which co-pperative solution emerges depends on the relative bargaining power of the different family members. A member’s bargaining power, in turn, may depend on various parameters, (p.173) especially the strength of the person’s ‘fall-back position’1 if co-operation should fail, and on feelings of love, affection, concern, and so on. In other words, it could depend on both the economic and extra-economic (including ‘moral’) factors.

Further, Sen in his entitlement approach argues that a person’s ability to command food depends in particular on his/her ownership endowments (of land, labour, and so on) and exchange entitlement mapping (although he notes that entitlements may also operate through various forms of non-market processes). In the present context, the approach could usefully be extended to cover especially two types of entitlements which neither fall within the purview of ownership nor are specifically exchange entitlements, namely, those stemming from (1) traditional rights to communal resources (such as village commons, forests, and so on), and (2) external social-support systems (such as those of patronage, kinship, friendship) embodying relationships between social groups or persons in which considerations other than the mere economic take precedence, that is, they fall under the rubric of what some authors have termed the ‘moral economy’. These typically relate to non-market exchanges, but do not altogether preclude market exchanges such as those existing between shopkeepers and favoured customers who get privileged treatment in access to goods, or goods on credit (Scott 1976; Greenough 1982).

Taking these as analytical starting points one could propose that:

  1. 1. A person’s ‘fall-back position’ (and associated bargaining strength within the family, vis-à-vis, say, food) would depend, among other things, on his/her ownership endowments, exchange entitlements, and external (social and communal) support systems (for example, patronage, kinship, friendship, and rights to communal resources).2 In other words, factors which impinge on a person’s command over food in general would also impinge on intrafamily cooperative conflicts over food. And inequalities among family members in their ownership endowments, exchange entitlements and access to external support systems would place some members in a weaker bargaining position relative to others. Gender could be one such basis of inequality, age another.

  2. 2. Crises of seasonality and calamity can negatively affect ownership endowments and exchange entitlements as well as the strength of external support systems for both sexes, but in so far as men and women are affected unequally it would alter their relative intrafamily bargaining strengths as well. A crisis which leads to a total collapse of or even a large decline in the wife’s fallback position (as could happen for instance during a famine) while that of the husband sustains (in relative terms) could weaken her bargaining power even to a point where non-cooperation is found more beneficial by the man than (p.174) cooperation, creating a tendency towards the disintegration of families, and the abandonment of spouses. The same could apply to children and the aged. Whether or not an actual breakdown occurs would be subject of course to other considerations as well, such as ‘moral’ ones, and to the ties of love and affection which hold a family together.

An examination of the evidence on household coping mechanisms during seasonality and calamity would help throw light on how the relative strength of intra-household fall-back positions impinge on and are affected by such contingencies, and how the ‘co-operative conflicts’ resolve themselves in the two contexts. Also we might ask: are responses to a calamity qualitatively and/ or quantitatively different from those to seasonality? The differences may stem both from the relative predictability of the one contingency as against the other, and from the relative intensity of shortfalls experienced in the two contexts. Given their periodicity, seasonal fluctuations can be anticipated, giving households greater scope for undertaking protective measures, although, in actuality, not all households may be in a position to protect themselves effectively.

Drawing upon a diversity of cross-disciplinary evidence, the next two sections will focus on the family’s mechanisms for coping with seasonality and with calamity respectively. In this process the nature and extent of differences in the mechanisms adopted in the two contexts, as well as in the relative contributions of and burdens borne by male and female members of the family, will be examined.

2. Coping With Seasonality

Seasonal variations in climate and crop cycles are associated, among other things, with variations in employment, wages, and food prices, which in turn can have repercussions on a poor agricultural household’s immediate as well as long-term command over food. Agricultural employment in India fluctuates seasonally both in terms of labour-force participation rates and the number of days of work available during particular months, as revealed in national-level surveys3 as well as region-specific studies;4 and it is associated with high levels of involuntary unemployment among those dependent substantially or mainly on agricultural wage labour for a livelihood: for all-India, the Rural Labour Enquiry (RLE) of 1974–5 recorded 124 days of not working during the year due to want of work for female agricultural labourers and 76 for male (see Table 5.1 for a breakdown by State).5 The length and frequency of seasonal (p.175) peaks and slacks as well as the average level of employment show considerable and complex interregional differences—varying by cropping patterns, the intensity of cropping, and the availability of irrigation.6 However, while intensification of crop rotations and expansion of irrigation may significantly increase labour requirements and alter the patterns of seasonality, fluctuations persist in greater or lesser degree, and certain months may still usually be (p.176) identifiable as chronically lean periods over fairly wide geographic areas.7 Seasonal fluctuations in wages (linked to task specificity of labour use as well) are typically lower than those in employment,8 but seasonal variability in wage earnings (dependent both on days of employment and on wages) is still likely to be substantial.9 This especially affects the purchasing power of those dependent on casual labour, in full or in part, for a livelihood (the less productive among whom may even be weeded out during slack periods).10 For households living on the margin of subsistence, these troughs in income (in cash or in kind) can constitute crisis situations. At the same time, as noted, seasonal fluctuations can be anticipated by rural households which can seek to devise various means of coping with them.

Table 5.1. Employment, unemployment and annual real earnings from agricultural wage work of women and men of agricultural labour households, (1974–1975)

Region/State

Average annual full days of:

Annual reala earnings per person

Ratio of male to female earnings

employment

not working due to want

F

M

F

M

F

M

North-Western

Haryana

131

203

88

88

213.3

406.8

1.91

Punjab

170

233

111

64

239.5

618.1

2.58

Rajasthan

163

239

97

49

150.2

310.7

2.07

Uttar Pradesh

124

200

114

57

133.2

277.4

2.08

Western and Central

Gujarat

160

206

111

67

168.0

278.1

1.66

Maharashtra

180

221

90

57

114.3

241.2

2.11

Madhya Pradesh

125

198

141

70

114.1

162.9

1.43

Eastern

Bihar

114

186

155

90

119.7

229.5

1.92

Orissa

111

164

158

92

73.1

155.7

2.13

West Bengal

147

210

166

88

167.1

294.3

1.76

Southern

Andhra Pradesh

138

193

103

61

104.8

198.2

1.89

Karnataka

175

204

81

58

134.2

246.4

1.84

Kerala

108

138

162

126

159.4

286.5

1.80

Tamil Nadu

118

148

142

98

93.4

183.9

1.97

All India

138

193

124

76

121.9

243.3

2.00

(a) Money earnings have been deflated by the Agricultural Consumer Price Index with 1964–5 as the base.

Sources: GOI (1981: 140, 143, 206, 212; 1979: 102, 103, 162).

Existing literature which in one way or another throws light on the mechanisms adopted by a household for coping with seasonal shortages points to a range and variety of methods, which broadly fall into five categories:

  1. (1) diversifying sources of income, including seasonal migration;

  2. (2) drawing upon communal resources—village common lands and forests;

  3. (3) drawing upon social relationships—patronage, kinship, friendship— and informal credit networks;

  4. (4) drawing upon household stores (of food, fuel, and so on) and adjusting current consumption patterns;

  5. (5) drawing upon assets.

These are not mutually exclusive and are typically adopted in combination. Consider each in turn.

2.1. Diversifying Sources of Income

Diversification of income sources—seeking employment where available, growing a range of crops including mixed cropping where the household has some land, keeping a variety of livestock (cows, buffaloes, goats, pigs, sheep) and poultry, and trading are common ways of dealing with seasonality and more generally with the subsistence risk that traditional farming systems entail.11 Some village studies list over ten different earning activities undertaken during the year by poor agricultural households.12

(p.177) Seasonal migration could be seen as a form of spatial diversification, made possible by interregional variations in peaks and slacks, and is usually especially high from unirrigated semi-arid areas to more prosperous irrigated ones.13 While national-level surveys (National Sample Survey, Census) do not throw much light on the scale of seasonal migration in India (since this is not separated out from other forms of migration), they do indicate that among agricultural labourers (who are affected most by seasonal troughs) the migrants are typically individual males moving from rural to rural areas, and more commonly within the district than across districts and states (Bardhan 1977). Region-specific studies reveal a variety of migration patterns: by distances covered, such as circulatory migration within a limited region (as noted by Breman 1979 a, 1985, in South Gujarat) and long-distance movements as from Bihar to Punjab; by the periods of absence—some go for a single agricultural operation, others for an entire season, yet others for much of the year (see Chen 1988, for Gujarat); by form, that is whether as individuals or in groups; and so on. Those migrating with capital assets, such as bullock or camel carts that can be used to transport produce, can command higher rates. It is noteworthy though that while seasonal migration clearly helps in mitigating region-specific stresses it can also depress wages in the areas to which the migrants go, and generate considerable hostility from local workers (Breman 1985).

While individual migration of younger males appears to be the most common pattern,14 individual women also migrate seasonally, especially among tribal communities (Bardhan 1977;15 Banerjee 1988), as do entire families hired in groups through contractors (Breman 1979 a, 1985), or through moneylenders seeking to ensure that their loans are repaid (Banerjee 1988). Caste groups may be formed as well, with young men and women of a single caste constituting work teams under the overall charge of a male member (Chen 1988), or, somewhat more commonly, women alone comprising work teams (Breman 1985). The nature of the labour process may itself dictate group migration: in Gujarat, the labour hired seasonally for sugarcane harvesting, for instance, has to work in teams—the male cutter followed by a helper (usually a wife, sister, or daughter) to clean the stalks of leaves, followed, in turn, by a child to bind the cleaned stalks together (Breman 1979 a).16 Whole families are (p.178) also more likely to migrate among the landless than among the cultivators who tend to retain some members to take care of the farm (Salva 1973). Bardhan’s (1977) review indicates that migration over longer distances (interstate relative to intrastate, and interdistrict relative to intradistrict), and from villages to towns, features more of male individuals and less of entire families or female individuals.

Essentially, coping with seasonal variation in employment opportunities and diversifying income sources requires close family co-operation. The labour of women and children is a specially critical input both in terms of earnings from wage employment and in enabling families to explore additional earning sources. Livestock and poultry, for instance, are typically looked after by women and children (see Mitra 1985; Banerjee 1988; George 1988), and individual male out-migration again requires them to assume additional responsibilities (see Desai 1982; Jetly 1987).

Typically too, agricultural labour women: (1) face much greater seasonal fluctuations in employment and earnings than men due to the greater task specificity of their work,17 are noted to have sharper peaks and longer slacks than men in the irrigated rice regions of South India (Harriss 1977), and have a lesser chance of finding employment in the slack seasons;18 (2) are much more dependent on wage labour than men,19 have lower average days of annual employment (and more days of involuntary unemployment), and lower daily wages (often even for the same tasks), which makes for considerable gender differences in annual real earnings (Table 5.1);20 (3) within agricultural wage work, are usually only employed as casual labour, typically men alone being hired on long-term contracts, possibly because they substitute for family men in ploughing, night operations, and market transactions—work in which women are socially disadvantaged or (as in the case of ploughing) socially excluded; (4) are much less likely to have meal provisions built into their (p.179) contracts;21 (5) have more limited information on jobs due to lower literacy levels, and lesser access to mass media and to the market place;22 (6) have lower job mobility due to their primary and often sole responsibility for child care, the ideology of female seclusion, and their vulnerability to class/caste-related sexual abuse. Indeed given purdah ideology and the social defining of gender roles, there are clear limitations to women’s control over their own labour power and ability to exchange it as they please, limiting their overall exchange entitlements.

Nevertheless their contribution to household earnings is substantial, as evidenced by studies from across the country.23 In particular, Mencher’s (1987) detailed quantitative evidence for landless and near landless agricultural labour households in twenty sample villages in Tamil Nadu and Kerala brings out several important features. She finds (taking a weighted average for each village) that (see Table 5.2): (1) although the wife’s earnings from agricultural wage work were typically half or one-third of the husband’s, in absolute terms her contribution from her earnings towards household maintenance was greater than his in six of the twenty sample villages, equal to or close to equal in five others, and substantial in the rest; (2) the wives typically contributed 90–100 per cent of their earnings and the men rarely gave over 60–70 per cent of theirs, keeping the rest for personal use;24 (3) the minimum contributed by all household females was greater than by all household males in thirteen of the twenty villages.

Table 5.2. Relative contributions by males and females to household maintenance, Rs per year

State/village

Wife

Husband

Proportion of wife’s to husband’s

All females Max. Min.

All males Max. Min.

Ratio males to all females

E

C

C/E

E

C

C/E

E

C

C

C

C

C

C

Kerala

Cannanore 1

1,138

962

0.85

1,954

1,249

0.64

0.58

0.77

1,924

500

2,935

211

1:0.79

Palghat 1

854

645

1.31

1,394

361

2,799

113

1:1.29

Palghat 2

1,065

990

0.93

2,039

1,406

0.69

0.52

0.70

1,606

104

3,029

115

1:0.62

Malappuram 1

435

421

0.97

1,219

1,020

0.84

0.36

0.41

1,333

101

3,517

45

1:0.25

Trichur 1

467

377

1.24

1,585

313

790

56

1:1.20

Trichur 2

786

688

0.88

1,787

1,294

0.72

0.44

0.53

1,323

309

2,824

380

1:0.56

Alleppey 1

752

691

0.92

748

569

0.76

1.01

1.21

1,181

14

1,072

49

1:1.30

Alleppey 2

530

438

0.83

743

541

0.73

0.71

0.81

600

211

970

137

1:0.77

Trivandrum 1

1,027

938

0.91

2,214

943

0.43

0.46

0.99

1,371

370

1,518

544

1:0.97

Trivandrum 2

1,420

1,209

0.85

2,235

1,141

0.51

0.64

1.06

1,797

480

2,165

317

1:1.16

Tamil Nadu

Chingleput 1

301

155

1.94

1,223

140

614

27

1:1.20

Chingleput 2

265

216

1.23

368

100

540

36

1:0.86

South Arcot 1

699

693

0.99

1,449

1,226

0.85

0.48

0.57

1,040

164

1,885

225

1:0.52

South Arcot 2

587

566

0.96

935

667

0.71

0.63

0.85

907

61

1,330

41

1:0.71

Thanjavur 1

468

490

0.96

816

801

616

127

1:1.20

Thanjavur 2

759

756

1.00

1,247

901

0.72

0.61

0.84

1,510

80

1,544

263

1:0.80

Tirunelveli 1

1,173

1,099

0.94

1,653

1,478

0.91

0.71

0.74

1,997

428

4,651

289

1:0.63

Madurai 1

564

556

0.99

1,240

938

0.76

0.45

0.59

1,072

184

1,716

135

1:0.60

Kanya Kumari 1

369

365

1.01

577

204

1,463

174

1:0.85

Kanya Kumari 2

599

570

0.95

1,297

808

0.62

0.46

0.71

891

156

1,681

399

1:0.61

Notes: Districts within each State are listed from north to south. The dash indicates a village where data on earnings were not collected. E = earnings; C = contributions.

Source: Mencher (1987).

To the extent that male and female labour peaks in wage employment and other tasks do not entirely coincide (given the sex-typing of certain agricultural tasks), one might ask to what extent there is a seasonal switching of responsibilities (p.180) (p.181) (p.182) within the family, especially for child care and domestic work, as found, for instance, by White (1976) among Javanese agricultural labourers. He noted that in the interests of maximizing household incomes, during periods of the year when women’s labour was in highest demand, men remained at home to cook and babysit, and when both parents were out earning, children herded the family livestock, cut fodder, and even undertook domestic work and child care. In the Indian context there is little systematic information to indicate whether traditionally gender-typed tasks are switched between the sexes to adjust for differential seasonal labour demand, but data on women’s time allocation to various tasks by seasons relating to five States (Table 5.3) does not suggest such shifts.25 Rather there appears to be a combination of a rise in women’s work burden and a squeezing of domestic work and child care during periods of high activity—agricultural or non-agricultural. On average too, time-allocation studies indicate that women in poor rural households typically work longer hours than men when all activities including domestic work are taken into account.26

Table 5.3. Daily time utilization of women by season (in hours)

Name of village and State

Season

House-hold work

Cooking and food processing

Personal work

Recreation and social activities

Animal husbandry

Agricultural work

Non-agricultural work

Cottage craft

Fuel search

Collection of forest produce

Marketing

Total time

N. Suriyan (Himachal Pradesh)

Rainy

Winter

Summer

1.96

3.25

4.00

2.46

3.85

3.30

0.25

0.50

0.40

0.18

0.15

0.75

1.62

3.55

1.00

2.62

0.25

_

0.81

1.05

0.78

1.06

1.05

0.05

0.12

0.25

0.08

_

_

_

_

0.15

_

11.08

14.05

10.36

Sehar (Madya Pradesh)

Rainy

Winter

Summer

2.15

3.75

1.75

4.10

4.98

5.43

1.15

1.15

1.30

1.25

1.01

1.30

1.05

0.50

0.45

3.15

1.45

0.30

_

_

_

_

0.10

3.07

1.30

2.05

2.00

1.00

_

1.00

_

_

_

15.15

14.99

16.60

Malari (Uttar Pradesh)

Rainy

Winter

Summer

1.44

3.01

3.65

3.85

4.45

4.41

0.56

0.63

0.76

0.19

0.12

0.08

1.12

2.37

2.51

5.61

1.37

1.85

_

_

0.17

_

_

_

0.56

0.75

0.77

_

_

_

0.06

_

_

13.39

12.70

14.20

Deokhop (Maharashtra)

Rainy

Winter

Summer

2.25

2.60

1.41

3.00

2.85

1.66

0.81

0.75

1.00

0.37

0.30

_

_

0.50

_

2.50

0.95

_

_

0.90

6.16

_

_

_

1.75

3.85

2.83

0.25

_

_

0.93

_

0.16

11.86

12.70

13.22

Rajapara

(Assam)

Rainy

Winter

Summer

3.29

2.37

2.91

3.90

3.75

4.41

1.10

1.12

1.52

_

_

_

0.55

0.50

2.05

5.21

6.00

2.33

_

_

_

_

1.25

0.33

0.17

_

_

_

_

_

0.07

_

0.16

14.29

14.99

13.71

Notes: Districts within each State are listed from north to south. The dash indicates a village where data on earnings were not collected. E = earnings; C = contributions.

Source: Mencher (1987).

2.2. Drawing upon Communal Resources

Access to village common property resources (CPRs) and state forests plays a critical role in enabling poor rural households to obtain essential items for daily use, to diversify income sources, and to increase the viability and stability of traditional farming systems by allowing a more integrated and diversified production strategy involving crops, trees, livestock, and so on. Communal resources serve as a source of various types of food, medicinal herbs, fuel, fodder, water, manure, silt, small timber, fibre, house-building with handicraft material, resin, gum, spices, and so on for personal use and sale, especially for the landless and land-poor. A study of the food habits of tribals in Madhya Pradesh lists 165 trees, shrubs, and climbers that they use as food in various forms (Tiwari, quoted in Randhawa 1980); and an other in Kerala lists eighty different forest items used as medicine (KFRI 1980).

Jodha’s (1986) study covering semi-arid regions in seven States of India shows that while all rural households use CPRs in some degree, for the poor (identified by him as the landless and those with less than 2 ha of dry-land equivalent) they account for as much as 20 per cent or more of total income in seven out of the twelve districts examined on this count, and 9–13 per cent in the remaining, while providing 1–4 per cent only of the incomes of non-poor households (Table 5.4). In six of the twelve districts the average income per year from CPRs for poor households exceeded Rs 700 even without counting (p.183) (p.184) their contribution towards the productivity of private property resources, such as owned land. The dependence of the poor on CPRs is especially high for fodder and for firewood which constitutes the single most important source of cooking fuel in rural India, providing 67–69 per cent of the domestic energy in the hills and desert areas of the north (NCAER 1981). In semi-arid regions, CPRs supply 91–100 per cent of firewood, 66–84 per cent of all domestic fuel, and 69–89 per cent of the grazing needs of the landless and land-poor (Jodha 1986); and some 8–9 per cent of total dietary requirements in many of these villages (Ryan et al. 1984). Even in the green revolution belt of Delhi and Uttar Pradesh, in two villages studied by Dasgupta (1987), agricultural labourers derived some 17 per cent and 24 per cent of income from free collection, including collection both from CPRs and other people’s fields; and for most poor households items collected free formed the bulk of dietary supplements to cereals bought or earned in kind through wage work.

Table 5.4. Average annual household income from CPRs in study villages of selected districtsa

Districtsb

Per household annual average income (Rs)

Poor householdsc

Other householdsd

Number of households

Value of CPR products collected

CPR share in live stock income

Total valuee

Number of households

Value of CPR products

CPR share in live stock income collected

Total valuee

Mahbubnagar

15

382

152

534 (17)

10

109

62

171 (1)

Mehsana

26

421

309

730 (16)

24

88

74

162 (1)

Sabarkantha

35

432

336

818 (21)

19

111

97

208 (1)

Mysore

26

534

115

649 (20)

11

112

58

170 (3)

Mandsaur

23

400

285

685 (18)

18

113

190

303 (1)

Raisen

37

568

212

780 (26)

15

283

185

468 (4)

Akola

16

342

105

447 (9)

9

85

49

134 (1)

Aurangabad

22

405

179

584 (13)

21

110

53

163 (1)

Sholapur

17

443

198

641 (20)

9

143

92

235 (2)

Jalore

24

447

262

709 (21)

27

170

217

387 (2)

Nagaur

32

473

358

831 (23)

25

143

295

438 (3)

Dharmapuri

30

530

208

738 (22)

11

112

54

164 (2)

(a) Based on field work during 1982–5. CPRs are common property land resources in which all village members have rights, such as village forests, community pastures, uncultivable and barren land, and so on.

(b) Number of villages covered was one each in Mahbubnagar, Akola, and Sholapur, and two each in other districts.

(c) Includes landless households and those owning less than 2 ha. dryland equivalent.

(d) Larger farmer households.

(e) Figures in brackets give percentage of CPR income to total household income.

Source: Jodha (1986).

Forests serve a similar function, especially for tribal populations—providing a basis for swidden cultivation, and a source of minor forest produce (MFP). Nearly 5 million persons (half of them in the north-east and the rest in central and eastern India) are assessed to be involved in shifting cultivation, covering an area of about 0.7 million ha (Srivastava 1977). In addition, MFP accounts for an estimated 13.6 to 38 per cent (varying by region) of total tribal income in Madhya Pradesh (a state containing in absolute numbers the largest concentration of tribals in the country), 10 to 55 per cent in Andhra Pradesh, and 35 per cent in parts of Gujarat (GOI 1982). In Orissa, 13 per cent of the forest population is estimated to subsist exclusively on MFP and for another 39 per cent it is an important source of secondary income (CSE 1985–6:91). Roughly, some 30 million people in the country are estimated to depend on MFP for a livelihood (Kulkarni 1983). An indication of some of the diverse forest-based activities undertaken by tribal families for survival is provided by Table 5.5 which also highlights the criticality of intra-family co-operation towards this end.

Table 5.5. Division of family labour in a Garo village, Meghalaya

Activity

Males

Females

Children

Home-based

(1) Making of basketry material

(2) Construction of house and bringing materials for the same

(3) Repairing of the house

(4) Fishing

(5) Hunting

(6) Propitiation for small ailments

(7) Decanting rice beer for guests

(8) Looking after domestic animals

(9) Threshing paddy

(10) Husking paddy and other grains

(1) Cooking

(2) Bringing firewood

(3) Fetching water

(4) Preparing beer

(5) Looking after small children

(6) Accompanying husband when he goes to fish

(7) Going to jungles in search of vegetables, mushrooms, crabs, snails, etc.

(8) Digging up of wild roots and tubers in times of scarcity

(1) Looking after infants during absence of parents

(2) Helping the mother to look after domestic animals

(3) Helping the mother in fetching water

(4) Helping the mother in washing cooking utensils

Shiftingcultivation

(1) Clearing of jungles

(2) Burning of plots

(3) Construction of the field house

(4) Making roads to the plots

(5) Transporting house-hold articles to field houses and back to village

(6) Keeping watch over the field at night

(7) Helping in harvesting

(8) Transportation of crops from field to the granary

(1) Planting and sowing

(2) Weeding

(3) Transporting house-hold articles to field houses and back to village

(4) Harvesting

(5) Transportation of crops to the granary

(1) Scaring away small animals and birds in day time

(2) Looking after small children when the parents are busy

Market-related

(1) Working as wage labourer

(2) Selling of cash crops

(3) Selling of fuel wood

(4) Selling of beef etc.

(1) Selling of vegetables

(2) Selling of fire wood in the town

(3) Preparation of rice beer for selling

(4) Selling of milk, fowl, pigs, etc.

(5) Selling of rice beer/distilled liquor in the town

(1) Working for daily wages

(2) Selling of vegetables etc.

(3) Selling of fodder

Source: Majumdar (1978: 88–89).

Seasonality impinges on this overall dependence in two ways: first, the degree of dependence on communal resources gets intensified during slack seasons; second, the availability of products from village commons and forests is itself subject to seasonal variations (see Banerjee 1988; Jodha 1986; Dasgupta 1987; Briscoe 1979). To the extent that periods of slack in crop production do not entirely coincide with lean periods of CPR and forest output, this helps sustain the poor when there is a trough in agricultural employment. Among tribals in central India, during normal pre-harvest seasonal shortages, gathered food provides 12 per cent of energy intakes compared with 2 per cent in the post-harvest period (Pingle 1975) For tribals in Bihar, forests are noted to be the only means of survival in the lean seasons, when the undergrowth of trees is picked for edible herbs, mushrooms, tubers, and so on. Wild millets, edible leaves, and fruits are found to be crucial sources of food for the poorest in the (p.185) lean periods in some north Indian villages (Dasgupta 1987). Irrigation tanks and river beds provide land for cultivation during the off-season. A recent study of six villages in semi-arid parts of South India (Ryan et al. 1984) also questions the universal validity of Chambers’s (1981) argument that nutritional distress is the most severe in the rainy season by pointing out that in many areas this is the time when wild leafy vegetables are most readily available. Again, Jodha’s (1986) estimation of employment provided by CPRs based on village level surveys in 5 semi-arid districts of four States in 1982–3 indicates that for the poor (1) CPRs provided exclusive employment for an estimated 43–49 days per household or 18–31 days per adult worker during the year, that is marginally more than the days worked on their own farms, in addition to part time employment when CPR-based activity was undertaken casually while doing other jobs; (2) in most districts CPRs were the only source of employment for 23–30 per cent of the total days for which the adults of poor households would otherwise have been involuntarily unemployed; (3) in some areas the total employment provided by CPRs was higher than that created by various anti-poverty government programmes.27

It is therefore during the periods when food or income are available through neither source, that poor households would be most vulnerable. Unfortunately, there is virtually no quantitative cross-matching by season of the employment and income generated via crop production with that generated via CPRs or forests, to identify such periods. A broad descriptive calendar of activities among the tribals of Orissa indicates that June to August is the leanest period when new crops are being sown, old stocks of grain have been depleted, and little food is available from the forest (see Table 5.6). Banerjee (1988) in the context of tribals in West Bengal similarly identifies July as the leanest month when there is a slack both in the crop calendar and in the availability of MFP, so that families have to resort to seasonal migration.

Table 5.6. Seasonal activities in three tribal villages in Orissa

Month

Settled cultivation

Shifting cultivation

MFP collected

VI

V2

V3

VI

V2

V3

VI

V2

V3

January

Harvesting winter crops

Harvesting wintercrops, sowing late varieties of millet and paddy

Harvesting winter crops

Nil

Harvesting late crops

Harvesting late crops

Siali leaves, datuna

Siali leaves, datun,abroomsticks

Not much collected

February

Nil

Sowing horsegram if rain; planting vegetables chilli, etc.

Nil

Nil

Nil

Harvesting late varieties and preparation for storage

Mahua, amla, behada, harada

Tamarind

Mahua

March

Nil

Some vegetable harvesting

Nil

Nil

Burning plot for bogodo

Clearing and burning plot

Mahua, mango, kendu fruit

Raw mango, jackfruit, processing of tamarind for storage (deseeding and salting)

Mahua

April

Nil

Clearing of plot

Nil

Nil

Burning and clearing plot or digging

Digging

Kendu leave sand fruit, mango, mahua, char, jackfruit

Jackfruit, ripe mango

Siali fruit, kendu, tamarind

May

Preparation of field; manuring

Preparation of plot

Preparation of field; ploughing

Clearing and burning; softening mud

Sowing cereal

Sowing cereal

Kenduc seeds and leaves, char

Mango, jackfruit, cashew

Mango, jackfruit, kusumo, tulo, mahua seeds

June

Ploughing and sowing

Sowing cereals; rice and mandia

Guarding of plot

Planting cerealin the latter half

Weeding

Weeding

Not much collected, kanda roots, shoots, and tubers

Mango, jackfruit

Mango processing (drying), jamun

July

Sowing, transplanting, ploughing again to soften mud; weeding

Sowing and transplanting vegetables; cultivation on dry land

Sowing

Sowing continued

Sowing pulses

Sowing pulses

Same as June

Not much

Not much, some bamboo shoots, etc.

August

Same as July

Transplanting, weeding, clearing

Weeding and clearing

Same as July

Weeding, clearing

Harvesting early varieties and vegetables

Same as June

Same as July

Same as July

September

Harvesting early varieties

Weeding, planting more cereals, harvesting mandia in the plains

Clearing bunds, sowing ragi

Harvesting early vegetables

Weeding, planting

Harvesting, guarding

Same as June

Sitaphal, Sialileavesa

Not much,

October

Harvesting, bunding, planting

Late planting of pulses, first mandia harvesting

Watering guarding

Harvesting, making bundles

Late planting, mandia harvesting, vegetable growing: pumpkin, bottlegourd, etc.

Nil

Sitaphal, bamboo, kendu,

Sitaphal, Sialileavesa

Same as Sept.

November

Harvesting paddy, mandia

Harvesting ragi

Harvesting paddy

Harvesting paddy, mandia

Harvesting makka, ragi, chana, etc.

Nil

Same as Oct.

Siali leavesa

Not much only honey

December

Same as Nov.

Late planting, harvesting other crops

Harvesting ragi, mustard

Same as Nov.

Same as Nov.

Nil

Same as Oct.

Same as Nov.

Salap (for liquor)

(a) Siali leaves and datun are available throughout the year but are collected mainly in lean months like January when no other minor forest produce is available.

Note: VI = Bhawanipatna; V2 = Kerandimals; V3 Mohana. These are the villages studied

Source: Fernandes and Menon (1987: 64).

It is necessary to note, though, that the rights to CPRs are not unrestrictedly granted to all comers. Typically they are extended to members of the village community by birth or marriage (Dasgupta 1987) and have been described by some scholars as part of the ‘moral’ claim of village members upon the village community (Scott 1976). Rights to forest produce are today also severely restricted by forest laws and a good deal of gathering of MFP is done illegally and under constant threat of harassment from forest guards (see Fernandes and Menon 1988; CSE 1985–6; Agarwal 1986a).

Typically it is women and children who play a primary role in the collection of CPR and forest produce (see Dasgupta 1987; Agarwal 1986a; Brara 1987), (p.186) (p.187) (p.188) (p.189) (p.190) (p.191) except in some tribal and scheduled caste communities dependent solely or mainly on forests, where men may be found to contribute substantially as well (Fernandes and Menon 1987). Women are also noted to have a more detailed knowledge of cultivated and wild crop varieties than men, among some tribes;28 and more generally it has been argued that poor peasant and tribal women as the main foragers and gatherers have a specially detailed reserve knowledge of edible forest produce that can help tide over prolonged shortages.29Of particular note too are the many examples of divergence between male and female priorities in CPR and forest development—with men typically favouring commercial varieties and women opting for trees and plants that fulfil the subsistence needs of food, fuel, and fodder.30 Essentially, CPRs and forests provide women, children, and even the aged with an independent source of subsistence, unmediated by dependency relationships on young male adults. Indeed the significance of CPRs and forests as food security systems for the poor needs particular underlining.

In this context, the rapid decline in and degradation of forest and CPR area would be a cause of particular concern. Area under forests, for instance, declined from 55.5 million hectares in 1972–5 to 46.4 million in 1981–2—an annual fall of l.3 million hectares.31 Again Jodha’s (1986) analysis for twenty-one semi-arid districts in seven states reveals that over the past three decades the area under CPRs has declined from 15–42 per cent (varying by region) in 1950–2 to 9–28 per cent in 1982–4; in other words, a decline ranging from 26 per cent to as much as 58 per cent in some regions (Table 5.7). In addition, the productivity of CPRs has been falling rapidly (Jodha 1985a), manifesting itself in shortening periods of assured supplies of fuel, food, fodder, and timber, a deterioration in the botanical composition of vegetation, the silting of water sources, and so on.

Table 5.7. Extent and decline of CPR land in study villages of selected districts

State and District

Number of villages

Area of CRP landh)

CPRs as of total village area

% decline in CPR area since 1950–2

1982–4

1950–2

Andhra Pradesh

Anantapur

2

221

15

24

36

Mahbubnagar

5

408

9

16

43

Medak

3

198

11

20

45

Gujarat

Banaskantha

5

167

9

19

49

Mehsana

5

224

11

17

37

Sabarkantha

5

198

12

22

46

Karnataka

Bidar

3

297

12

20

41

Dharwad

3

242

10

18

44

Gulbarga

3

291

9

15

43

Mysore

3

335

18

27

32

Madhya Pradesh

Mandsaur

4

327

22

34

34

Raisen

6

770

23

42

47

Vidisha

4

338

28

38

32

Maharashtra

Akola

5

192

11

19

42

Aurangabad

4

304

15

21

30

Sholapur

4

422

19

25

26

Rajasthan

Jalore

5

639

18

29

37

Jodhpur

3

591

16

38

58

Nagaur

3

619

15

41

63

Tamil Nadu

Coimbatore

4

187

9

17

47

Dharmapuri

3

225

12

26

52

Note: Based on village-level records and field work during 1982–5.

Source: Jodha (1986).

The decline in CPRs is attributed by Jodha partly to population growth and to the physical submersion of land under large irrigation projects, but mainly to land privatization, as a result of three causes: illegal encroachments by larger farmers made legal over time; government distribution of CPRs to individuals under various schemes officially intended to benefit the poor, such as the land reform programme implemented in the 1950s and the twenty-point programme in the 1970s; and the auctioning of parts of CPRs by the government to private contractors for commercial exploitation. In practice, much of this privatization favoured the larger farmers: in six of the nineteen districts surveyed, the poor received between 0.8 and 1.6 hectares per household, and the larger farmers 1.5 to 4.9 hectares per household (Table 5.8). In the Jodhpur district of (p.192) Rajasthan, 62 per cent of the privatized commons went to farmers already owning 10–15 hectares or more, and included 90 per cent of the good-quality land—while the landless got only 13 per cent, much of it of poor quality (Jodha 1983). Hence the collective loss of the poor was not made up by the private gain of a few of them.

Table 5.8 Distribution of privatized CPRs in study villages of selected districtsa

State, district, (Number of villages)

Total land given (ha)

Total households (No.)

%of land to

%of recipients among

Hectares per per household

Hectares per per household

Poor

Others

Poor

Others

Poor

Others

Poor

Others

Beforeb

Afterb

Beforeb

After

Andhra Pradesh

Mahbubnagar (3)

418

343

50

50

76

24

0.8

2.6

1.3

0.9

3.0

5.1

Medak (3)

75

58

51

49

59

41

1.1

1.5

1.0

2.2

3.1

4.6

Gujarat

Banaskantha (3)

75

29

18

82

38

62

1.3

3.4

0.8

2.0

5.4

8.8

Mahsana (2)

85

63

20

80

36

64

0.7

1.7

1.0

1.7

8.0

9.8

Sabarkantha (3)

127

74

23

77

55

45

0.9

2.8

0.5

1.1

7.0

9.8

Karnataka

Bidar (3)

89

55

39

61

64

36

1.0

2.8

1.0

2.0

6.4

9.2

Gulbarga (3)

112

50

43

57

60

40

1.6

3.2

0.8

2.4

4.5

7.7

Mysore (3)

161

98

44

56

67

33

1.2

2.9

0.9

1.9

4.1

11.6

Madhya Pradesh

Mandsaur (2)

120

55

45

55

75

25

1.3

4.7

1.2

2.5

7.7

12.4

Raisen (4)

115

72

42

58

68

32

1.0

2.9

1.3

2.2

6.2

9.0

Vidisha (4)

123

77

38

62

48

52

1.3

1.9

1.3

2.5

4.9

6.8

Maharashtra

Akola (3)

101

100

39

61

58

42

0.7

1.5

1.0

1.6

3.1

4.6

Aurangabad (2)

83

55

30

70

42

58

1.1

1.8

1.1

2.2

6.4

8.3

Sholapur (3)

132

72

42

58

53

47

1.5

2.3

0.7

2.2

3.4

5.6

Rajasthan

Jalore (2)

83

27

14

86

37

63

1.4

4.9

0.3

1.7

7.2

12.5

Jodhpur (2)

405

318

24

76

35

65

0.9

1.5

0.4

1.3

2.3

3.8

Nagaur (3)

147

81

21

79

41

59

1.2

3.1

1.3

2.5

2.4

5.2

Tamil Nadu

Coimbatore (4)

206

145

50

50

75

25

1.1

2.9

0.8

2.5

3.8

5.8

Dharmapuri (3)

241

127

49

51

55

45

0.9

2.1

1.0

1.9

4.6

7.5

(a) Based on field work during 1982–5.

(b) Before and after receiving CPR land.

Source: Jodha (1986).

(p.193) 2.3. Drawing upon Social Relationships and Informal Credit Systems

2.3.1. Patronage

Patron-client relationships as forms of traditional social-security systems for poor rural households have been widely commented on.32 For instance, the jajmani system in northern India, and variations therein all over the subcontinent, of labour and service relations between upper-caste landlord-patrons and their clients—including share-croppers, agricultural labourers, household servants, and the service castes (priests, artisans, barbers, scavengers, and so on)—traditionally involved a set of reciprocal (though asymmetrical)33 obligations between patrons and clients. On the side of the clients was the obligation to provide labour, rent, or services (as the case may be). On the side of the patron was the obligation to provide a fixed payment in grain, annually or semi-annually (although the amounts differed across regions),34and to honour a variety of rights (which varied by the type of client and the region). These rights might include gifts of food on ceremonial and festive occasions, a site for a house, the collection of weeds and leaves from standing crops, asking for old clothing from time to time, gleaning the fields for grains after a harvest, free dung, and in general a right to assistance such as interest-free loans during periods of dire need. These obligations between patrons and clients, often defined in moral and ritual terms, assured (among other things) a stable labour supply for the dominant cultivator castes, by tying the client to the patron and thus limiting his mobility, and provided the client some guarantee of a minimum subsistence. The privileges of collection and so on could be seen as providing clients with access to the private property resources of patrons supplementing their access to CPRs, for fulfilling seasonal needs.

Agriculture-related patron-client relationships were essentially those between landowners and men who worked as permanent farm servants or share croppers (although their wives and children could also be called in to provide domestic help). While Wiser’s (1936) contention that the jajmani relationships were essentially harmonious and symmetrical is disputable, since various forms of exploitation and coercion of the clients were in fact to be found in the system,35 there is general agreement that the system did provide some security (p.194) (p.195) (p.196) against seasonal troughs and the guarantee of a bare subsistence minimum to poor agricultural households during ‘normal’ years, even though it is clear that it did not sustain under acute crisis and generalized shortages as during a drought or famine (of which more later).36

However, much of the literature also indicates an erosion of these relationships over time, although not their disappearance, attributable to several factors impinging on both demand and supply. On the demand side there has been an increase in the possibility of hiring in cheaper labour seasonally from outside without involving patron obligations (Breman 1985); the competition from factory products which has affected the demand for items made locally by the service castes; the possibilities of making a profit by selling the goods earlier distributed free to clients with improved transportation and storage facilities, and so on (Dasgupta 1987).37. On the supply side, clients are noted to (p.197) opt out of such a relationship where more secure or profitable alternatives (especially urban jobs) can be obtained (Chen 1988). Table 5.9 which traces patron-client support available to thirty-five families in Rajasthan in 1963–4 and again in 1982–4, reveals a decline in all forms of support: for instance, the percentage of households resorting to off-season borrowing of foodgrains from patrons declined from 77 to 26 over the two decades. Among the halis of South Gujarat, Breman (1985) argues that this decline is associated with a slide from poverty to pauperization.

Table 5.9. Decline in patron-client relationships in Rajasthan

Indicators

% households during

1963–6

1982–4

Households with one or more members working as attached/semi-attached labour

37

7

Households residing on patron’s land/yard

31

0

Households resorting to off-season borrowing of food-grain from patrons

77

26

Households taking seed loans from patrons

34

9

Households marketing farm produce only through patrons

86

23

Households taking loans from others besides patrons

13

47

Note: Data relates to 35 households whose per capita annual income (at 1964–6 prices) declined during 1982–4 compared to 1964–6.

Source: Jodha (1985 b: 12).

2.3.2. Kinship and Friendship

Social relationships with kin, and with villagers outside the kin network, can serve as a means of support in various ways (although the information available on this is essentially descriptive). For instance, women draw upon other women for borrowing small amounts of food stuffs, fuel, fodder, and so on (Maclachlan 1983; Jetly 1987; Chen 1988); similarly, ‘helper’ relationships among men, with friends, neighbours, and local kin, can provide for reciprocal labour, a sharing of irrigation water, and the loaning and renting of agricultural implements, draft animals, and machinery, thus enabling the enhancement of small farm productivity (as noted by Montgomery 1977 in Tamil Nadu). Also, group help for peak operations and various labour exchange arrangements have been characteristic of tribal communities (Kar 1982; Majumdar 1978). Traditional support systems exist too along caste lines, and often it is easier for the poor to draw upon help from the richer peasants of their own caste than from outside it (Caldwell et al. 1986).

Women typically play a significant role in the cultivation of social relationships—especially through marriage alliances that they are frequently instrumental in arranging, and through complex reciprocal gift-giving systems (Maclachlan 1983; Vatuk 1981; Sharma 1980). At the same time, they are often less able to seek the help of their natal kin than men. Patrilocality, which greatly facilitates male ability to call upon the support of consanguinous kin, can at the same time restrict women’s ability to do so—the degree of such restriction depending on the physical and social distance between the natal and marital homes.38 Especially in north-west India, marriages into villages at substantial distances from the woman’s natal village and home, taboos against marrying close kin, limited social contact between parental and marital (p.198) families, discouragement of extended stays by married daughters in their natal home, and the considerable formality that marks their relationship with their natal kin after marriage, may all prevent women from seeking their support in times of crisis.39 These factors impinge with less severity where village endogamy and cross-cousin marriages are allowed, and social contact between the two sets of in-laws is more frequently possible, as, for instance, in the south and north-east of India, or in Muslim communities there and elsewhere.40Here instances of women seeking frequent help from parents when in need are not uncommon (see Caldwell et al. 1986 for Karnataka, and Nath 1979 for Bangladesh, where women are observed to increase the frequency of their visits to their parental homes during the ‘hungry season’).

Existing, rather fragmentary evidence (from India and neighbouring countries) suggests, however, that among the poor, even such social-support systems as existed are getting increasingly eroded. In one Bangladesh village study only 54 per cent of all widows were found to have the security of being integrated members of their sons’ households (Cain et al. 1979). Jansen (1986) likewise found several cases in the Bangladesh village he surveyed of old parents, widowers, and widows complaining that they had been abandoned by sons. In the Indian context, the erosion of family support, and support from kin and friends, has been documented especially in the context of tribal communities, traditionally characterized by team work in agriculture and mutual labour-exchange and community support arrangements. With growing land privatization, monetization of production, and class differentiation, there has been a general decline in such arrangements. Fernandes and Menon (1987: 115) give a graphic description of this in relation to the tribals in Orissa:

the earlier sense of sharing has disappeared. This has affected women’s condition more than of men, since they are responsible for the day to day running of the household. . . . Earlier they could rely on their neighbours in times of need. Today this has been replaced with a sense of alienation and helplessness. . . The trend is to leave each family to its own fate.

They note too that community support is no longer forthcoming for women if they want to separate from their husbands due to ill-treatment, or for widows neglected or exploited by relatives, where such support was customary, earlier. The overall neglect of widows by kin is also noted by Drèze (1988a). However, (p.199) much more systematic research is needed on the decline of family support systems, especially with marital breakdown and old age.

2.3.3 Other Informal Credit Systems

While there is considerable literature on interlinked credit-labour systems, much of it does not distinguish between traditional patronage (which also provided informal credit) and emerging new forms of dependency linkages associated with the development of capitalist farming in agriculture (Bhalla 1977; Breman 1985). Yet such a distinction appears warranted in that the new relations are found to contain most of the exploitative features of the old without the guaranteed subsistence that patronage was said to have provided (Breman 1985; Banerjee 1988). What is clear is that consumption credit from landlords and traders, whether extended within interlinked credit-labour systems (Bardhan and Rudra 1981),41 or independent of such linkages (Bliss and Stern 1982), continue to be critical in tiding over seasonality, although such borrowings at usurious rates of interest may lead to long-term bondage (Mundle 1979). Outside the context of tied relationships, however, such loans may not be forthcoming readily for poor households. Binswanger and Rosenzweig (1986) note on the basis of ICRISAT (International Crop Research Institute for the Semi-Arid Tropics) data for South India that the probability of getting credit from money-lenders depends critically on wealth endowments. Landless households often get short shrift from moneylenders, and small farmers are granted crop loans usually in kind, to be repaid also usually in kind, with a 25 to 50 per cent interest over the crop cycle, that is, at annual interest rates of 75 to 100 per cent.

2.4. Drawing upon Household Stores and Adjusting Current Consumption

Drawing upon stored items—grain, animal products, dung cakes, dried berries, grasses, and so on—many of them built up by women and children, is another way of coping with seasonal fluctuations.42 Although there appears to be little quantitative evidence for India to indicate the extent of this depletion by households (most surveys on this count are confined to the effects of drought and famine), data from Bangladesh (Table 5.10) clearly show the seasonal depletion in average cereal stock associated with troughs in agricultural wage employment and peaks in grain prices, especially during the lean months from August to October.

Table 5.10. Seasonal variations in prices, wages, and cereal stocks in Matlab Thana, Bangladesh

Month

Wholesale price of medium rice (taka/mound) 1976–7

Agricultural labour wage (taka/day) 1977–8

Average household cereal stocks (mounds) 1976–7

March

116

10.0

2.1

April

111

8.0

2.9

May

122

7.5

5.1

June

118

7.5

3.1

July

123

6.0

2.8

August

123

-

2.1

September

123

-

1.0

October

125

6.0

0.7

November

105

7.5

3.2

December

111

7.5

5.3

January

113

7.0

2.8

February

118

n.i.

1.1

Notes: Data relate to 205 households in Matlab Thana. The dash indicates extremely limited agricultural work opportunities, n.i. = no information given in the original table

Source: Chaudhury et al. (1981: 55)

However, stock depletion and other measures adopted do not prove adequate for all times of the year, and the hungry season—true to its description— typically (p.200) calls for consumption adjustments by the poor in one or more of the following ways:

  1. 1. change in the content of the diet such as shifting to coarse grains, drawing on wild vegetables, berries, and so on from the village commons and forests, with some tribal groups in West Bengal reportedly even consuming bats during the lean season (Banerjee 1988);

  2. 2. reduction of total intake: missing meals, reducing the number of meals cooked per day, stretching available food supplies by, say, eating rice gruel instead of boiled rice every day, and so on (Mitra 1988);

  3. 3. adjusting intra-household distribution of available food. A part of this adjustment reflects perceptions about needs, especially arising out of differential activity levels, but a part also appears to reflect a gender bias. Fernandes and Menon (1987) note, for instance, that among the tribals of Orissa, during months of low food availability, children get first preference, then men, and finally women. During the month when ploughing has to be done food is allocated in favour of men, and women either borrow from neighbours or go hungry. There is no mention, however, of any reverse adjustments by the men when women have a heavy load of work, as during the rice transplanting season. Also, working adults are favoured over non-working elders (typically old women). Fernandes and Menon emphasize that observed gender inequalities (p.201) are recent: ‘In the traditional tribal society they used to eat together. Today the woman has to starve quite often. Because of food shortages, she gives all that is available to the man and keeps nothing for herself.’ Caldwell et al. (1986) similarly find that women in rural Karnataka deny themselves food during seasonal shortages.

In general, the extent and causes of gender biases in food allocation within the household have been the subject of considerable discussion in recent literature, and are still being debated.43 However, there is a broad agreement that while anti-female biases may not exist as a universal feature across India, there is undeniable discrimination in food allocation against females in certain regions, such as the north-west of India, and in certain social groupings, especially the poor (including the poor in parts of South India).44 Also Sen’s (1981) and Kynch and Maguire’s (1988) work suggests that gender differentials among children tend to increase during crisis. And Behrman (1986), on the basis of a modelling exercise using data on child nutrition for South India, concludes that ‘parents display male preference, at least during the lean season, particularly in lower caste households’ and that ‘the nutritionally most vulnerable—especially females—may be at considerable risk when food is scarcest’.

Further, in so far as there is a closer positive link between children’s nutritional status and the mother’s wage employment than the father’s, among agricultural labour households (as found, for instance, in Kerala by Kumar 1978 and Gulati 1978), seasonal slacks in female employment may deprive the children as well. Kumar observes (1978: 43, 46) that ‘when mothers are not in the labour force, an increase in (the household’s) wage income shows absolutely no incremental effect on child nutrition’, but ‘for those mothers who are in the labour force, it is their own wages that primarily account for the positive wage income effect in child nutrition’.45 And while the greater length of the mother’s employment (and therefore presumably less time devoted to child care), especially during peak periods, was negatively related to child nutrition, this did not outweigh the positive income effect of the mother’s earnings.46 (p.202) The net effect of maternal wage incomes was thus positive during both peak and slack seasons, although more so in the latter. Gulati (1978) again found that in agricultural labour households, daughters, in particular, were left much worse off than sons on the mother’s non-working days, although she does not give details of seasonal fluctuations and whether this would be so at all times of the year.

All this suggests that periods or days of lowest food consumption during the year need not coincide for all household members, but may vary by age and gender, and be conditional upon the combined effects of several factors, including the woman’s employment and earnings, actual activity levels, as well as perceptions about needs. For the household on an average too, not all slack periods, and not all parts of given slack periods, need be periods of hunger. Usually there is a lagged effect as earnings in periods of peak labour demand help tide over the initial part of the following slack. The worst time is usually noted to be the start of a busy season after an extended slack when stored items have been depleted and the agricultural peak calls for additional energy inputs.47

Where variations in energy requirements do not match with food availability and intake levels, these may be reflected in seasonal changes in bodyweights as found in several African studies,48 as well as in Asia: a Bangladesh study of mothers noted that fluctuations were highest for the landless among whom stocks were entirely depleted during the critical lean month of October, while landowning families had sufficient stocks to maintain adequate family consumption throughout the year (Chaudhury et al. 1981). However the extent of these fluctuations could vary by region, age, and gender. A six-village ICRISAT study for Andhra Pradesh and Maharashtra found no significant differences in the weights of adults by seasons (Ryan et al. 1984) but a village study in Uttar Pradesh did find such changes, the seasonal loss of weight among women being greater than among men (Kynch and Maguire 1988).

Where seasonal changes in intake are not observed to affect bodyweight it could mean either that food intakes are being adjusted to activity levels, and/ or, as is being argued by some, that individuals are able to ‘adapt’ over certain ranges of reduced intakes, by increasing the efficiency of energy utilization, without any change in bodyweight. On the issue of efficiency—which has provoked considerable discussion—Osmani (1987), in an excellent unravelling of the various threads in the debate and a careful examination of the available evidence, notes that: There is . . . some evidence to show that phenotypic adaptation of a kind may take place in the efficiency of energy utilisation, but it (p.203) seems almost always to be accompanied by alteration of body weight’, and further that ‘what is relevant is "pure" adaptation. But the existing evidence does not indicate that such adaptation is possible’ (1987: 41; emphasis mine).

It has of course also been suggested that a seasonal decline in bodyweight need not impair functioning efficiency, since the body evolves effective storage mechanisms for various essential nutrients that are able to smooth out seasonal peaks and troughs in supply (Longhurst and Payne 1981). But this would still leave at least certain sections of the population at risk: the very poor, and among them typically pregnant and lactating women and pre-school age children, are found to be especially vulnerable to bodyweight changes and seasonal malnutrition.49 Childbirth, breast-feeding and weaning practices, and morbidity, also show a seasonal pattern, especially among agricultural workers, the burden of which falls mainly on women and children.50

Finally, social consumption such as celebrations of marriages and festivals reflect seasonal adaptations as well, often clustering around post-harvest periods.

2.5. Drawing upon Assets

As noted earlier, to judge the effectiveness of a household’s coping mechanisms we would need to know whether or not at the end of the year the household has suffered a reduction in the productive capacity of its family members or in its assets, especially (but not only) productive assets such as land and draught animals. On nutrition, as noted, available evidence is limited and provides no clear direction. Unfortunately, there also appears to be little quantitative information on seasonal changes in a household’s asset position in the Indian context. However, the detailed Bangladesh study by Hossain (1987) is indicative. It examines, among other things, seasonality in the distribution of employment, earnings, expenditure, borrowings, and the sale of assets in net terms, and observes noteworthy seasonal variations (see Table 5.11). Consumption expenditure exceeds earnings in virtually every season, though the deficits are especially high in the July-August slack. Deficits are made up by net borrowings, mostly from relatives (30 per cent), merchants (21 per cent), and moneylenders (11 per cent); the sale of assets (though it is not indicated what type of assets are sold, to which family members the sold assets belong, and whether any are mortgaged); and the transfer of receipts of various kinds (from friends, relatives, neighbours, political leaders, and so on).

Table 5.11. Seasonal distribution of employment, earnings, and expenditure in six villages, Bangladesh

Season (weeks)

Corresponding date

Income-earning work (Takas)

Earnings (Takas)

Transfer receipts (Takas)

Net assets sale (Takas)

Netborrowing (Takas)

Value foodconsumed (Takas)

Food-grain consumption (Seersa)

Total consumption expenditure (Takas)

Slack (1–5)

1 july- 4 Aug. 1984

58.15

133.70

15.07

2.97

32.52

98.51

15.09

176.86

Busy (6–10)

5 Aug.- 8 Sept. 1984

65.44

179.72

25.89

16.49

30.06

106.50

15.69

193.91

Slack (11–20)

9 Sept.- 17 Nov. 1984

55.92

155.85

22.09

3.05

35.93

104.04

14.50

177.74

Busy (21–31)

18 Nov. 1985 -2 Feb. 1985

64.75

160.18

21.63

−5.24

29.45

103.01

14.49

175.93

Slack (32–40)

3 Feb.- 6 Apr. 1985

64.60

168.09

16.37

−5.49

17.49

99.76

14.77

170.44

Busy (41–50)

7 Apr.- 15 June 1985

68.12

239.22

16.89

0.19

−1.56

100.62

16.01

184.96

Slack (51–52)

16 June- 29 June 1985

57.51

192.48

24.76

−1.37

12.04

102.51

17.60

220.16

Average for year (1–52)

1 July 1984 -29 June 1985

62.83

176.36

19.80

0.41

22.40

102.08

15.13

180.59

(a) 1 seer = 900 grams.

Source: Hossain (1987: 148).

(p.204)

(p.205) The overall deficits suggest the operation of what Chambers (1981a) has termed ‘poverty ratchets’ where seasonal troughs may leave poor households worse off each year in relation to the previous year, to the extent that they are unable to replenish the assets or repay the loans incurred by year’s end, as opposed to what he terms the operation of ‘screws’ which create pressures in certain seasons but from which the household can recover in other seasons.

In broad overview therefore, our examination of the mechanisms adopted by the poor agricultural family for coping with seasonality indicates the following:

  1. 1. The inputs of all family members are critical in tiding over the troughs, although the burden of coping in terms of work and consumption adjustments appears to fall disproportionately on women (see summary below):51

  2. 2. CPRs, income diversification strategies, and seasonal migration possibilities are significant for cushioning seasonal distress, and need strengthening. In contrast, traditional patron and kin support and informal credit systems provide limited help, are not always dependable, and (especially the former two) are eroding over time.

  3. 3. The effect of seasonability on poor households is likely to vary between households—especially cross-regionally—causing some to suffer temporary hardships and others to face a downward slide with each sharp trough. The latter households are more likely to be those with least assets, belonging to low (p.206) castes/tribes, and located in semi-arid regions. However, much more systematic research is needed on this in the Indian context (and in fact the overall South Asian context) by region, community (tribal/non-tribal, scheduled caste and other, and so on), gender and age, both on coping mechanisms and on the effectiveness of these mechanisms, before more generalizations are possible.52

Coping mechanism

Relative gender contributions (W=women M=men)

Seasonal migration

W 〈 M

Keeping livestock, and other forms of income diversification

W 〉 M

Drawing on communal resources

W 〉 M

Drawing on social relationships:

patronage

W 〈 M

husband’s kin

W 〈 M

wife’s kin

W 〉 M

neighbours/friends

W = M

Adjusting work loads

W 〉 M

Drawing on household stores

W 〉 M

Consumption adjustments

W 〉 M

3. Coping With Calamity: Drought and Famine

The literature on drought and famine (as discussed in detail below) indicates that disasters and severe food crisis situations bring few radical departures in coping mechanisms from those noted under seasonality, but they accentuate the degree of dependence on the same. In so doing they also bring out the critical limitations of these arrangements in providing subsistence security and crisis relief under calamity. Mechanisms of coping may then shift to measures of desperation as the family itself begins to disintegrate, vagrancy and destitution grow, and conflicts over food may even take the form of looting and violence. Indeed seasonality, drought, and famine may be seen as three points in a continuum, representing increasingly severe threats to the food security of the family. This section considers strategies and experiences of coping during drought and famine.

3.1. Communal Resources: Intensified Use and Conflicts

Dependence on communal resources which are already strained in their ability to cushion seasonal fluctuations increases significantly during severe food shortage situations. All studies of drought and famine without exception note shifts to ‘famine’ foods not consumed normally and many of which could be toxic.53 The Indian Famine Commission of 1880 lists a vast number of berries, grasses, and roots consumed in such contexts, identifying those that are toxic. In addition, dependence on CPRs for fuel, fodder, and water, increases. In Reddy’s (1988) Andhra Pradesh village study, in the pre-drought years an estimated 25–30 per cent of the incomes of landless households and 15–20 per cent of the income of the whole village came from CPRs; during the 1983–7 (p.207) drought period, 33–45 per cent of the income of the landless (for six families the percentage was 60) and 30–35 per cent of the village income came from CPRs.

At the same time, both the availability of CPR products and access to them for the poor declines, and conflicts over access intensify. Reddy (1988) notes that during the drought period, farmers blocked off rivulets flowing through their land to get a larger share of the water, and also to create pools for raising the water table of their private irrigation wells; people were denied access to neem trees located on private land to which earlier there was open access; and the poor were no longer allowed to collect tree leaves essential for roofing. In Gujarat during the 1987–8 drought, some of the fights among graziers over access to village grass patches ended in physical violence (Chen 1987). In this context, the panchayat (village council) is usually found to favour the better-off.54

3.2. Patrons, Kin, Friends, and Informal Credit Systems: Decreasing Reliability

Patron-client relationships, already weak, are found virtually to snap during periods of severe food stress, especially during a generalized food shortage, say caused by a drought. This decline in patronage can take several forms:

  1. (1) reduction of payments for traditional services (usually enforced through the panchayat), or a failure to meet obligations altogether, in some cases leading to a break in the relationship (Lewis 1958; Reddy 1988; Jodha 1978);

  2. (2) denial of access to free gifts (Reddy 1988);

  3. (3) denial of loans—the very poor being considered bad risks (Caldwell et al 1986; Reddy 1988; Rao 1974; Borkar and Nadkarni 1975; Singh 1975).

Greenough (1982) identifies a hierarchy of patron-client relationships—noting that some ties snap more easily and earlier than others. For instance, during the Bengal famine of 1943, the more ‘fixed’ clients—attached to the family for generations, providing caste-linked ritual and other services—held on the longest; and those essentially market-linked, such as between shopkeepers and favoured customers, were the first to snap; semi-market relationships as with share-croppers came in-between.

However, the failure of patrons to meet their obligations to clients is attributed by many authors to their inability to do so in a period of overall (p.208) shortages. Reddy (1988) notes that in 1984, the second year of the drought in his Andhra village, most of the landowning families ‘lost their capacity to meet the obligations of their clients in matters of annual payments in relation to services’. Epstein (1967: 24) argues that the jajmani system only broke down ‘during extended periods of crisis…in extremely bad harvests, when the total produce was not sufficiently large to provide subsistence to all members of the society’. Greenough (1982) observes likewise that: ‘In my view, the really difficult period of famine begins when the rural patrons exhaust their resources and are no longer able to succour their dependents.’ The assumption here is that the failure of patrons to keep obligations is dictated by adverse circumstances shared by all sections of the population, including the better-off, even if not in equal degree, and, by inference, once these circumstances improve the relationships will be restored. However, the collapse of patronage in calamities, even in the absence of a serious generalized food shortage, as noted in the Bengal famine of 1943 when there was in fact large-scale speculative holding back of food-grains by surplus cultivators and organized traders (Sen 1981), reflects more than an ‘inability’ on the part of patrons to fulfil their obligations —it indicates, as noted in Section 2, also a long-term irreversible erosion in these relationships.

At the same time, Prindle’s (1979) description of a village community’s response in Nepal during the drought of 1971 indicates that where the local population can exert group pressure, even a generalized shortage can be more equitably shared. He describes how the combined pressure of the less well-off brahmins and bhujels (the latter being hereditary ploughmen for the richer brahmins) succeeded in (1) preventing surplus farmers from selling their grain outside the village at a profit and making it available instead to villagers on interest-free credit; (2) preventing wealthy brahmins from reducing their daily wages for agricultural work; (3) persuading the wealthiest and most influential brahmin village patron to secure a loan of Rs 15,000 from shopkeepers in the market centre in the nearby town and re-lend it to almost every villager, thus enabling the villagers to buy grains at lower rates than those prevailing later when the drought worsened; (4) getting this patron to secure government credit assistance for them, as well as obtain cloth on credit from cloth merchants in the town and provide the cloth to the villagers on credit; and (5) getting the rich to maintain the system of sponsoring festival feasts at which poorer brahmins and bhujels were fed. In this co-operative pressure, the fact that many of the brahmins were also poor, appears to have helped create a tie of solidarity between them and the bhujels; at the same time the poorer brahmins could draw upon their caste connections to persuade the wealthy brahmins to cede to the demands that benefited not only them but also the bhujels. Here group pressure did not merely help to maintain traditional patron-client obligations but also to extend them, and so distribute the burden of food shortages more equitably. The significance of such group solidarity and (p.209) pressure clearly has a lesson to offer in drought management (of which more later).

Typically, however, as with patron-client relationships, friends and relations provide only limited support during severe food shortages (especially those caught up in a similar crisis). In the initial phases, relatives located in non-drought-affected regions or urban areas, and the better-off friends and neighbours may help out. In particular, several South Asian studies mention the family’s ability to take help from the wife’s parental home. In Nepal every family in the village studied by Prindle (1979) visited the wife’s kin at least once to secure loans during the 1971 drought. Caldwell et al. (1986) found that 43 per cent of the families they studied in nine Karnataka villages received such help. However, for reasons spelt out in the previous section, this kind of support is unlikely to be regionally widespread in India. Also the amounts of loans given are typically small and form only a limited percentage of total borrowings by needy families—the rest being obtained from money-lenders and institutional sources (ostensibly for production but often redirected to consumption). In near-famine conditions the limited help from friends and relatives can virtually dry up: in the Bihar drought of 1966–7, 88.8 per cent of those surveyed in Palamau district said that they could not or did not help their neighbours (Singh 1975). Small borrowings of food items and cash, possible during seasonal troughs, are also no longer possible. Experiences of African famines point to a gradual shrinkage of the food-sharing arrangements common among women in normal times, the cessation of hospitality in terms of sharing food with visitors, and various other manifestations of atypical social behaviour (Vaughan 1987).

Demand for consumption credit from various sources other than friends, relations, and patrons also increases sharply during calamity. Again, while the better-off can get loans from the government agencies, the poor are usually forced to depend on money-lenders or to disinvest.55 During the 1983 drought in Karnataka those who got least loans were either the upper-caste landlords and merchants or the harijan agricultural labourers—the former because they did not need credit, the latter because they could not get it. This bias was especially apparent in loans from banks and co-operatives (Caldwell et al. 1986). A similar pattern was observed by Borkar and Nadkarni (1975) in their study of two villages during the Maharashtra drought of 1970–3.

(p.210) 3.3. Work and Demographic Adjustments

3.3.1 Short-Term Work Shifts

Both long and short-term work adjustments by families have been noted as a result of calamity. In the short term, changes in work patterns include:

  1. (1) increased time spent in searching, foraging, gleaning, and the processing of food; in looking for fodder, fuel, and water; and in the more careful grazing of cattle—typically women’s and children’s time getting so extended (Jodha 1975; Caldwell et al. 1986; Chen 1988);

  2. (2) shifts in traditional caste and gender divisions of labour: for instance, people from the Reddy caste were found selling vegetables during the 1971–3 drought in Medak district, Andhra Pradesh—a task that they considered below their status in normal years (Rao 1974); and women not belonging to the shoemaker caste were found taking up shoemaking during the 1966–7 Bihar drought (Singh 1975). This flexibility can prove crucial for survival: in the Nepal drought of 1971, the bhujels who willingly took up whatever work was going survived the drought better than the brahmins who held on more rigidly to caste status norms and many of whom were finally forced to emigrate (Prindle 1979);

  3. (3) migration in various forms: Jodha (1975) describes the typical patterns in Rajasthan following the drought of 1963–4.56 However, his emphasis that migration is undertaken only as a last resort after adjustments in current consumption, depletion of household stocks, and the mortgage or sale of assets has been undertaken, appears surprising given that seasonal migration is common in Rajasthan—clearly what changes is the scale on which migration takes place. Also, reports from Western India (quoted in McAlpin 1983) suggest that in the nineteenth-century famines, migration took the form of aimless wandering, but in the twentieth century a few men were typically sent out from each village to scout for work as soon as the failure of early rains made harvest losses inevitable. Accounts of recent famines, however, do not suggest such community co-operation.

3.3.2. Long-Term Work and Demographic Shifts

While short-term adjustments of work during crisis are noted in several studies, few have examined long-term adjustments resulting from an experience of severe or recurrent crisis. Among these few are Jodha’s (1975) study of fifteen poor, upper-caste families in Rajasthan, and Caldwell et al.’s (1986) and Maclachlan’s (1983) studies in (p.211) Karnataka. In Rajasthan, due to recurrent droughts in the 1930s, the families studied permanently adopted the work patterns and life-styles of the lower castes—essentially a situation of downward economic and social mobility for survival. The Karnataka examples, however, indicate greater success in adaptation. Caldwell et al. (1986) note that the farming families in the nine villages they studied were extremely conscious of the need to diversify future income streams as risk insurance against seasonality and drought, and sought to do this by some members systematically seeking non-farm employment, investing in the children’s education as an entry into urban employment, and looking for sons-in-law with urban jobs.

Maclachlan (1983), by contrast, documents the case of a farming community that built up drought resistance through farm-based diversification itself. He compares the experience of the village during the severe drought of 1965 when no deaths occurred as a result, with that of 1876–7 when a large number died in the area, and asks the question: why didn’t they starve in the 1960s? His answer: agricultural intensification, especially the cultivation of gardens made possible by deploying substantial amounts of family labour, first to dig a large number of open surface wells and then to tend the fields and gardens. The long-term strategy included the building up of large joint families with a careful selection of brides from poorer hard-working families, especially first daughters (presumably used to managing a number of siblings), a preference for cross-cousin marriages, and a strong ideological emphasis on the joint family as the ideal and most productive family form. The ideological under-pinnings of efficient labour deployment were provided by the concept of dharma (duty) used to order a certain sexual division of labour and norms of behaviour.

On testing the relative productivities of households with more than one adult male (loosely classified by him as ‘joint’ families)57 and others, Maclachlan found that the former did indeed have higher yields per acre and per capita than those with only one male adult (Table 5.12). Joint family households were favoured by landlords when seeking tenants, derived economies in the division of farm tasks, and had higher savings, a more intensive use of equipment, the advantages of interactive decision-making, and the advice of the elders. Although the number of women is not found to affect agricultural productivity much, households with more than one woman had distinct advantages in terms of diversification, with livestock, dairy, and sericulture being looked after (p.212) mainly by them. Less time was also needed by women for domestic work, since work-sharing was possible.

Table 5.12. Economic performance of sixteen households with the poorest ratio of land to household size

Type of household

Yield/acre (Rs)

Per capita income (Rs)

Sample family labour used/acre (hr)

Distribution of labour (%)

Family lands

Kuli

Tenant

Others

One man (N=7)

119.0

160.9

132.8

43.1

52.6

1.9

2.3

More than one man (N=9)

283.8

355.1

192.6

37.6

35.8

25.2

4.2

Source: Maclachlan (1983: 147).

However, the factors which keep families joint are clearly weaker than those pulling them apart; and Kolenda's (1987) complex and detailed classification of household structures on the basis of existing studies from across India suggests that the typical family in India is nuclear, not joint (as defined in note 71). This is corroborated by more region-specific studies.58 And while there appears to be no clear-cut evidence of a decline over time in the incidence of joint families,59 there is found to be a noteworthy variation in this incidence by region (for instance, joint families are more common in the north-west than in central India), and by economic class (jointness is more prevalent amongst the rich than the poor), although less so by caste.60 Kolenda (1987), who finds a (p.213) lesser tendency towards jointness among the untouchable castes relative to the ‘twice born’, argues that among high-caste groups social opinion still discourages break-ups as long as the parents are alive (ownership of unpartitioned land by the father also serves to postpone the split), but among the untouchable castes, sons do not consider the obligation to their parents binding. Be that as it may, in so far as low-caste households are also often poor, even economically they would be in a less favourable position to fulfil such obligations.

Unfortunately, Maclachlan focuses almost entirely on the largely self sufficient cultivating households and gives us little idea of how agricultural labour households survived the drought, what strategies they followed, whether they gained from the agricultural intensification undertaken by the farmers, and what their relationship with the cultivating households was during their drought. For instance, did the farmers help sustain the labourers?

Nevertheless, there are lessons to be learnt from Maclachlan’s study: small irrigation works and agricultural intensification as ways of countering drought conditions; joint cultivation for increasing agricultural productivity and labour use efficiency; and work-sharing arrangements within the household for greater income diversification and for increasing the efficiency of domestic labour. The question is: what institutional arrangements can re-create cooperative labour deployment networks successfully, outside the context of the family?

3.4. Consumption Adjustments: Growing Hunger

Apart from increased dependence on a variety of ‘famine foods’ noted earlier, drought conditions or the onset of famine lead to shifts in the types of foods eaten—for instance, from fine to coarse grains such as from rice and dal to ragi and jowar (sometimes even that used normally as cattle feed) (Caldwell et al. 1986; Reddy 1988). As scarcity worsens there is also likely to be a decline in the consumption of protective foods such as milk, meat, fruits, vegetables, and so on;61 the stretching of food to make it last longer; a reduction in the quantity eaten by cooking fewer meals a day, or going hungry for several days.62 In severe shortages yams and roots may be eaten to kill the appetite—the tribals of Gujarat use a bitter root called kand for this purpose (Rangaswami 1985). Most studies record a fall in consumption per capita during drought years.63 In (p.214) addition, there is a decline in expenditure on clothing, children’s education, socio-religious ceremonies, a postponement of marriages, and so on.64 The passage from severe drought to famine is a progressive worsening in consumption as entitlements to food collapse.

There are, however, noteworthy gender differences in food-sharing— essentially an accentuation of gender differences noted in normal times, especially but not only in northern India. In Caldwell et al.’s (1986) study of the 1983 Karnataka drought, male heads of households were asked who bore the greatest burden of food shortages—43 per cent reported that all shared equally from the cutbacks; 35 per cent said adults (but none said ‘men’) and 22 per cent identified mostly women, children, and the old. (Possibly the same questions addressed to the women would have made the gender bias more explicit.) On morbidity, 60 per cent of the families reporting illness identified only women, children, and the aged; 22 per cent listed adults and 17 per cent mentioned a broad division by age and sex. Campbell and Trechter (1982), who explicitly examined differences in gender responses to varying degrees of food shortages in North Cameroon, found that under severe shortage situations, while women’s actions commonly included going hungry for the whole day, men’s more typically included migration.

3.5. Depleting Assets

Jodha (1975) observes in a much commented-on debate with Morris (1974, 1975) that the primary concern of rural families when faced with extreme food shortages is not the protection of current consumption but protection of productive assets—which are disposed of as a last resort, since their loss is likely to affect the household’s long-term prospects of recovery from the crisis. He argues, on the basis of his data on the Rajasthan drought of 1963–4, that, to begin with, households reduce their consumpiton and deplete stocks of fodder, fuel-wood, grain, animal products, dung-cakes, and so on, which have been stored for the lean period. Subsequently household goods such as ropes, cots, utensils, and jewellery—the non-productive assets—are disposed of (also see Table 5.13). Some small livestock may also be sold. It is noteworthy that unlike the price of livestock which plummets during drought (Jodha 1978), that of jewellery may show little decline (Borkar and Nadkarni 1975). Finally, draught animals and land are disposed of—the former being sold outright, the latter first mortgaged where possible. The disposal of productive assets thus becomes a barometer of the degree of distress and is typically associated with famine conditions.

Table 5.13. Monthwise foodgrain consumption, depletion of inventories and assets, and outmigration in Jodhpur district (Rajasthan), 1963–1964 (drought year) and 1964–1965 (post-drought year)

% of households consuming

300–450 g/day

451–600 g/day

601–750 g/day

D

PD

D

PD

D

PD

October 1963

7.7

_

21.2

67.3

71.1

32.7

November 1963

21.3

_

25.0

73.0

53.8

27.0

December 1963

34.6

1.9

38.5

69.2

26.9

28.9

January 1964

48.1

5.7

34.6

74.9

17.3

19.4

February 1964

57.7

3.8

35.5

76.8

5.8

20.4

March 1964

60.5

5.7

32.7

78.7

5.8

15.4

April 1964

69.2

7.6

25.0

78.7

5.8

13.9

% of households who

Sold inventories

Sold assets

Mortgaged assets

Outmigrated

October 1963

22.9

_

0.7

_

November 1963

40.9

1.4

_

_

December 1963

55.0

_

2.8

_

January 1964

50.1

_

5.6

_

February 1964

52.2

2.1

14.5

1.4

March 1964

35.3

0.7

26.4

4.2

April 1964

36.7

4.2

22.2

21.5

TOTAL

100.0a

8.4

68.8

27.5

Notes: D = drought year; PD = post drought year. Data for consumption relate to a sample of 52 farming households in one village. Data for activities relate to a sample of 144 farming households.

(a) Some households took some steps more than once.

Source: Jodha (1975: 1613, 1615).

According to the Mahalanobis et al. (1946) estimates, during the Bengal famine of 1943, between April 1943 and April 1944, one-fourth of all families (p.215) (numbering nearly 1.6 million) owning paddy land before the famine had either sold it in full or in part, or mortgaged it. Sales in full were highest among those with less than two acres: these, constituting 6.1 per cent of all families, thus lost their main source of livelihood. Those buying the land were largely based outside the village, many of them urban dwellers. Thirteen per cent of plough cattle was also lost (due to sales or deaths) and only 25 per cent of it could be replaced later. Losses during droughts which do not escalate into famines are usually less severe but there is also found to be a considerable variation in their extent in different drought contexts. In the Bihar drought bordering on famine in 1966–7, for instance, of the total value of property lost in Palamau district, the sale and mortgage of land accounted for 24 per cent, of cattle 50 per cent, and of jewellery and household utensils 25 per cent. During (p.216) other droughts, however, the losses recorded, especially of land, were less severe.65 Both droughts and famines, however, typically leave in their wake considerable increases in inequalities in asset distribution.66

Once mortgaged or sold the chances of these assets being redeemed in full in the post-calamity period are often slim, especially for the small farmers and landless agricultural labourers. Jodha’s (1978) comparision of the asset position of farmers in the pre-drought, drought, and post-drought years in three states shows only a partial recovery in productive assets, and a further depletion in non-productive assets in virtually all cases in the post-drought years (Table 5.14), suggesting lagged effects of the calamity.

Table 5.14. Indices of asset positiona in drought and post-drought years relative to pre-drought years (pre-drought year = 100)

Farm size group/year

Jodhpur

Barmer

Banas Kantha

Aurangabad

Sholapur

(Rajasthan)

(Rajasthan)

(Gujarat)

(Maharashtra)

(Maharashtra)

1963–64b

1969–70b

1969–70b

1972–73b

1972–73b

P

NP

P

NP

P

NP

P

NP

P

NP

Small Farms

Drought year

63

72

82

98

59

81

74

50

69

50

Post-drought year

73

68

95

99

79

73

NA

NA

60

29

Large Farms

Drought year

75

89

40

46

87

97

79

62

95

88

Post-drought year

83

91

42

47

93

97

NA

NA

88

108

All Households

Drought year

63

69

66

79

84

85

77

87

84

69

Post-drought year

83

64

76

77

97

85

NA

NA

78

66

(a) All assets are valued at 1972–3 prices.

(b) Drought years.

Notes: P = productive assets, which include livestock, farm implements, machinery, etc., but exclude land. NP = non-productive assets, which include jewellery, financial assets and consumer durables. NA = Not available.

Source: Jodha (1978: A-39).

Also, detailed case-studies collected by him of seven families in Rajasthan hit by successive droughts over twenty-five years, although inadequate for generalization, are illustrative of the process of impoverishment that can set in:the land possessed by these families decreased from forty-three acres to less than five acres per family over this period. Similarly, an investigation of land transfers in three drought-prone districts in Rajasthan over ten years showed (p.217) that 85 per cent of all land disposed of was due to drought-induced mortgages. The absence of any significant land transfers during the Maharashtra drought of 1970–3 is thus seen as one of the important indicators of the success of government relief programmes (Drèze 1988 b).

However, in this literature on asset disposal during drought and famine two aspects need specific comment: first the loss of certain types of assets has gone virtually unrecorded—one such is trees. Chen (1987) who monitored the 1987–8 drought in Gujarat found that in one village where virtually no trees were cut in the pre-drought year, some 150 trees were cut and sold in 1986 when the monsoon rains failed, and an additional 300 or more were cut in the second year of the drought—most of them being slow-growing species such as mango and neem. Reddy (1988) likewise notes that between 1983–7 almost 2,000 trees were cut and sold in the village he surveyed.

Second, in the noted sequencing of asset sales, what appears to have been missed in the literature is that the assets which are the first casualty—namely, household utensils and jewellery67—also happen to be those typically owned and controlled by women. Underlying the sales is of course an economic rationality, especially in the case of jewellery, in that in addition to the importance of holding on to productive assets such as land and cattle, jewellery is a much more liquid asset than land, and unlike cattle less prone to price plummeting.68 Jewellery (and gold and silver in general) clearly serves as a store of value for crisis situations. However, such sales have a special significance when we note that usually these are the only assets possessed by women.

For instance, while there is no comprehensive data on land ownership by gender, several factors would support the view that few rural women own agricultural land or homestead plots.69 First, legally, even under post independence laws, among all communities in India (except essentially those in the matrilineal south-west which is governed by different laws), women have highly unequal rights to agricultural land. Second, customarily too, barring a few matrilineal groups in the south-west and north-east, and exceptional circumstances elsewhere (for example, of daughters in son-less families) women in most communities had virtually no recognized inheritance rights in land, and even less the right to control or alienate it. Usufructory rights were somewhat more common, but basically confined to tribal communities; and, over time, these too have been systematically eroded, especially with the privatization of communal land typically registered in male names. Third, while legislation today allows women in most communities to own, use, and dispose of land (though not on an equal basis with men), practices such as (p.218) patrilocal post-marital residence, village exogamy, female seclusion and restriction of movement, in many parts of India prevent women from claiming their legal share. Especially in northern India, as a result, women typically give up their claims in favour of brothers, and widows face considerable hostility in their assertion of claims. What applies to agricultural land applies equally to other immovable property, including homestead plots.

For assets other than land such as livestock it is difficult to establish individual ownership within the household, except in the context of cattle distribution under the government’s Integrated Rural Development Programme (IRDP) for poverty alleviation, where the person receiving the cattle loan is the registered owner, and in customary practices among some communities where animals are gifted to the daughter on marriage (Mitra 1985; Miller 1981). While livestock is typically looked after by women and children, this does not imply rights of disposal.

A woman’s assets, then, would consist at best of some small animals,70 and items obtained as gifts during marriage, such as jewellery and utensils, and even these may not always be under her control.71 Once these are disposed of, even if the household is able to protect its productive assets during the calamity, women would be left with nothing to fall back on if abandoned or if there is a drought recurrence, since in the cycle of mortgage-indebtedness sale, jewellery once sold is unlikely to be easily redeemed. Most households are often unable to make up even for cattle losses: as noted in Table 5.14, on small farms, the indices of productive assets rose in the post-drought year (although not up to the pre-drought level) while those for non-productive assets, including jewellery, typically fell further. This leaves women especially vulnerable during a severe calamity such as famine, when families may themselves begin to fragment and disintegrate.

4. The Disintegrating Family

Famines in a sense represent the extreme end of the spectrum in terms of food crisis. They also pose in the most stark terms the economic and moral dilemmas relating to intrahousehold food sharing, and provide a mirror to intrafamily relations that few other contexts can. The Bengal famine of 1943 serves as a specially poignant case-study in this respect.

A variety of evidence points to some of the specific disadvantages suffered by women and children of poor households during the famine. To begin with, Mahalanobis et al. ’s (1946) estimates of persons made destitute in rural Bengal, (p.219) revealed a predominance of young and middle-aged females: in January 1943, 55 per cent of all destitutes and 66 per cent of destitutes in the 15–50 age-group were females; women of this age-group also constituted the largest number of new additions to destitutes between January 1943 and May 1944 (and amounted to twice the number of new male destitutes of this age group). Children (of both sexes) in the 5–15 age category were the next largest in number. Occupationally, in absolute terms the largest number of destitutes were agricultural labourers, and in terms of the percentage affected in a given occupation, the worst hit were those dependent on fishing, transport, agricultural labour, paddy-husking, and craft, in that order (Tables 5.15 and 5.16) .

Table 5.15. Estimated number of destitute persons in rural Bengal by age and sex (100,000s)

Age group (yrs.)

Estimated number of destitute persons

Estimated number of new destitute persons between Jan. 1943 and May 1944

Jan. 1943

May 1944

Male

Female

Total

Male

Female

Total

Male

Female

Total

0–5

0.27

0.31

0.58

0.60

0.50

1.10

0.33

0.19

0.52

5–15

1.61

1.06

2.67

2.19

1.68

3.87

0.58

0.62

1.20

15–50

0.98

1.95

2.93

1.43

2.92

4.35

0.45

0.97

1.42

Over 50

0.50

0.77

1.27

0.59

0.85

1.44

0.09

0.08

0.17

TOTAL

3.36

4.09

7.45

4.81

5.95

10.76

1.45

1.86

3.31

Source: Mahalanobis et al. (1946).

Table 5.16. Estimated number of destitute persons in May 1944 in rural Bengal (100,000s)

Family occupation

Infants

Children

Men

Women

Old persons

Total

% destitute persons in total population in that occupation

Agriculture

0.10

0.36

0.14

0.22

0.08

0.90

0.75

Agriculture and labour

0.13

0.15

0.05

0.09

0.07

0.49

0.82

Agricultural labour

0.21

0.70

0.27

0.58

0.23

1.99

3.02

Non-cultivating owner

0.03

0.11

0.05

0.06

0.02

0.27

0.98

All agriculture

0.47

1.32

0.51

0.95

0.40

3.65

1.30

Fishing

0.07

0.15

0.09

0.12

0.02

0.45

7.75

Craft

0.03

0.23

0.10

0.15

0.05

0.56

2.70

Husking paddy

0.01

0.01

-

0.04

-

0.06

2.73

Transport

0.01

0.03

0.02

0.02

-

0.08

7.06

Trade

0.02

0.18

0.05

0.12

0.03

0.40

1.31

Profession and service

0.04

0.17

0.08

0.11

0.04

0.44

1.30

Non-agricultural labour

0.01

0.05

0.03

0.02

-

0.11

2.00

Other occupations

0.02

0.04

0.02

0.02

0.04

0.14

1.83

Living on charity

0.42

1.69

0.53

1.37

0.86

4.87

73.86

TOTAL

1.10

3.87

1.43

2.92

1.44

10.76

3.39

Source: Estimated from Chattopadhyaya and Mukerjee (1946: 18).

Second, Greenough’s construction (based on the Bengal Famine Committee’s records) of the social profile of those who came to State relief centres during the most intense months of the famine, and were able to pay cash for food as compared to those who were absolutely destitute, is revealing (although the categories are overlapping). The cash-paying chief recipients of relief tended to be male adults; Hindus of the trading, cultivating, and high Sudra castes; and heads of households averaging 3.6 persons and willing to go some distance to seek out relief goods being sold at a discount. In comparison, those dependent entirely on gratuitous relief tended to be females (mostly adults) 84 per cent of whom got gratis aid compared with 43 per cent of the male destitutes (see Table 5.17) ; Muslims and Hindus of lower, fishing, and untouchable castes; widows and married persons often with absent spouses; heads of households averaging less than 2.5 persons; those in need principally due to a loss of income, and those able to secure relief only at their place of residence.72

Table 5.17. Heads of households receiving reliefa from the Bengal Relief Committee between October 1943 and June 1946b

Sex of chief recipients

Cash-paying recipients

Gratis recipients

Both

No.

%

No.

%

No.

% cash-paying

Male

965

84.4

726

44.1

1 691

(57.1)

Female

178

15.6

922

55.9

1 100

(16.2)

TOTAL

1 143

100.0

1 648

100.0

2 791

(41.0)

(a) Mostly food but also clothing and medicine.

(b) About 87% of these received relief between Oct. 1943 and June 1946.

Source: Greenough (1982: 190).

Third, consider the famine mortality figures in Table 5.18. The general pattern appears to be one of relative female advantage: for most age groups and for all ages taken together, survey-based estimates of the absolute number of deaths during the period from January to December 1943, as well as the absolute and proportionate increases in mortality rates due to famine (‘excess mortality’) were greater for males than females. But the exceptions to this general pattern, especially the age-grouping 20–40, are noteworthy in the context of the present discussion. For this age-grouping the estimated absolute number of famine deaths during the critical year of 1943 as well as the absolute increase in mortality rates were greater for women than men, though the proportionate increase showed gender parity. Men of this age-grouping showed the least excess mortality (in absolute terms) of all age groups of either sex. The greater mortality disadvantage in absolute terms suffered by women relative to men in the 20–40 age-grouping is particularly noteworthy as this is despite the reduction in child-birth and child-bearing associated risk due to famine induced reduction in birth rates. Also the bulk of women in this age-grouping (p.220) (p.221) (p.222) would be within marital relationships, that is women who may normally be expected to have the support of husbands (of which more later).

Table 5.18. 1943 Bengal famine mortality by age and sex

Age group (years)

No. of persons in sample survey who died over Jan.-Dec. 1943

Death rates in sample survey %

Death rates 1931 census calculations %

Increase in death rates attributable to 1943 famine (excess mortality)

Gender differences in excess mortality rates

In absolute terms (%)

In proportionate terms

In absolute terms (F-M)

In Proportionate terms

M

F

M

F

M

F

M

F

M

F

(1)a

(2)b

(3)b

(4)b

(5)b

(6)c

(7)c

(8)d

(9)e

(10)f

(11)g

(12)h

(13)i

1–5

189

168

8.43

7.99

5.9

5.2

2.53

2.79

0.43

0.54

+0.26

+0.11

5–10

366

330

6.00

6.57

1.6

1.4

4.40

5.17

2.75

3.69

+0.77

+0.94

10–15

195

137

4.92

3.95

1.0

1.2

3.92

2.75

3.92

2.29

−1.17

−1.63

15–20

146

148

4.20

3.89

1.2

1.7

3.00

2.19

2.50

1.29

−0.81

−1.21

20–30

190

287

3.21

4.49

1.8

2.5

1.41

1.99

0.78

0.80

+0.58

+0.02

30–40

223

226

4.94

5.58

3.0

3.4

1.94

2.18

0.65

0.64

+0.24

−0.01

40–50

282

191

8.77

6.77

4.2

4.3

4.57

2.47

1.09

0.57

−2.10

−0.52

50–60

277

169

14.08

10.78

5.5

5.0

8.58

5.78

1.56

1.16

−2.80

−0.40

over 60

225

122

17.10

12.66

9.8

9.1

7.30

3.56

0.74

0.39

−3.74

−0.35

All ages

2 093

1 778

6.40

5.89

3.1

3.2

4.18

3.21

1.35

1.12

−0.97

−0.23

(a) Infants under 1 year excluded.

(b) Mahalanobis et al. (1946) sample survey.

(c) Taken as the ‘normal’ period rates.

(d) Col. 4 less col. 6.

(e) Col. 5 less col. 7.

(f) Col. 8 divided by col. 6.

(g) Col. 9 divided by col. 7.

(h) Col. 9 less col. 8.

(i) Col. 11 less col. 10.

Source: Greenough (1982: 311); some taken directly, in other cases computed.

Fourth, a survey of 2,537 destitutes living on the pavements of Calcutta in September 1943 quoted by Greenough showed that 52.7 per cent of them were female, and among the married destitutes (excluding those widowed or divorced) 63.6 per cent were female. The pavement-dwellers had earlier belonged to 820 families; half of these had recently broken up, 70 per cent as a result of husbands and wives separating. The women, on being asked why they had left the village for Calcutta, said that their husbands had been unable to maintain them and had either deserted them or asked them to go elsewhere in search of food (Greenough does not say how many gave this answer).

In addition, Greenough provides anecdotal evidence on women being abandoned by husbands, and being forced into begging or prostitution; and quotes several reports of parents selling their children into bondage and especially their girl children (even in the ages 2–13) into prostitution. The observation of the Nari Seva Sangha, a private women’s service society set up in 1944 for sheltering abandoned women during the famine, is again telling: ‘Women were being thrown out into the streets and multitudes of them were being forced into a life of shame in a country which has always boasted about its purity and modesty’ (quoted in Greenough 1982: 225).

On the basis of such evidence Greenough argues that there was a striking fragmentation or disintegration of families during the famine, and further that this familial disintegration ‘did not occur randomly but seems to have been the result of the intentional exclusion of the less valued family members (women and children) from domestic subsistence’, by a decision to this effect on the part of the male head of the household (which he sees as parallel to the decision of the village patron to withhold support towards clients). He notes that the (p.223) Bengali family is not only a social but also a moral unit where the male head— the korta—controls all family members and their corporate interests; he is the annadata—the provider of sustenance for the family—and during the Bengal famine he increasingly favoured adults over children and males over females. However, Greenough argues that ‘decisions as to who will eat and who must starve are not taken suddenly or arbitrarily in the heat of the famine but draw upon an existing cultural arrangement that assigns a value to family members according to their roles’. It is a decision in keeping with the powerful existing ideal of family continuity, equated with the continuity of the patriline through the adult male, whose survival thus counts over that of women and children:

If the separated ‘master’ survives a crisis, he is enabled to marry again despite the death or violation of his spouse, and he may then engender children who can be expected to maintain the lineage under more prosperous conditions

(Greenough 1982: 224).

In other words, no contradiction is seen here between the male self-interest and the preservation of the moral order in the Bengal context.

The evidence on familial disintegration and the abandonment of women and children presented by Greenough (and as can also be gleaned from the surveys of destitutes mentioned earlier), while largely indirect (and from which the scale of this disintegration cannot be assessed), is nevertheless sufficiently indicative that this was not an isolated phenomenon, and could even have been fairly widespread.73 This is also a phenomenon which has been noted, although mainly in passing, in accounts of other famines. Alamgir (1980: 135) observes of the 1974 Bangladesh famine: ‘Besides, there were many cases of desertion. In Rangpur, special homes for deserted children were set up. In Dacca, there were many women who were deserted by their husbands among the inmates of vagrant homes.’ And in her reconstruction of the 1949 famine in Malawi, Vaughan (1987: 123) notes: ‘Women stress how frequently they were abandoned by men, how harrowing it was to be left responsible for their suffering and dying children.’74

What needs closer examination and some discussion, however, is Greenough’s explanation of what underlay the male decision to abandon. Appadurai (1984: 485) offers an alternative explanation to Greenough’s, suggesting that the abandonment of women and children could be seen as

an effort to maximise the life-chances of each and every member of the family in circumstances where co-residence was clearly not feasible. Thus the sale of children might be seen as an effort to construct a better set of life-chances than those of the existing family structure.

(p.224) (p.225)

(p.226) However, Appadurai’s explanation is difficult to uphold in a context where female chastity is held in very high regard (indeed loss of female life may be considered preferable to loss of chastity in many communities in India, including Bengal),75 and where the husband would know that the abandoned spouse would most likely fall easy prey to sexual exploitation. In fact, the women were doubly victimized, since even if they managed to physically survive, they were unlikely to be accepted back as wives by the husbands or their families; and in the aftermath of the famine, they filled the shelters opened to accommodate them. Similarly, the sale of girl children into prostitution could hardly be seen as a step taken by parents to improve their life chances.

An alternative way of viewing this process of family break-up and female victimization would be in terms of shifts in the relative male/female entitlements and fall-back positions—and so in their relative bargaining strengths within the family, along the lines discussed in Section 1 of this chapter. What Greenough vividly describes is a process of family disintegration during a calamity when the normal rules governing the sharing of resources and arrangements of reciprocity are beginning to shred. The rules that he emphasizes, however, are those stemming from a culture-specific moral code that ultimately favours male survival over female.

Without denying the persuasiveness of the moral explanation, it seems to me that the decision by the man to abandon the wife and children occurs at a point when the wife’s entitlements have collapsed completely while those of the husband are weakened but not entirely gone. It is a telling point that 57 per cent of the males compared to only 16 per cent of the female recipients of relief in the earlier mentioned survey were able to pay cash for the food received, and also that the least excess mortality (in absolute terms) occurred among men of 20–40 years of age.76 Also, one of the traditional (and few) fall-back occupations of women in Bengal—paddy-husking—clearly did not sustain beyond a point. Paddy-huskers ranked among the first four categories of those hardest hit by the famine; and the large number of new entrants to this occupation (noted by Sen 1981: 72) during the worst phase of the famine would not only be a symptom of the growing marginalization of women in other occupations but would also have depressed earnings in this occupation further.77 Again, it is noteworthy that in the three illustrative accounts of abandoned women presented by Greenough, one is of a woman of 18 found wandering in Calcutta (p.227) whose husband pushed her out and ‘decided to stay on the (one acre of) land to look after his coming betel-vine’ (1982: 218); in the second, the husband divorced and abandoned the wife saying he was going ‘to join the army’; and in the third, the husband left and was not heard of subsequently. In general, Greenough notes: ‘If there were still limited resources in the village and little likelihood of obtaining subsistence outside the village, the husband would stay behind. Alternatively, if a source of food or work was offered outside the village, and local resources had been exhausted, the husband would migrate, leaving the wife and children behind’ (ibid).

All this appears to suggest that the men’s fall-back position was stronger than of the women on at least two counts: first, their greater mobility and ability to migrate over longer distances for a job without the fear of sexual exploitation that women faced, and second, the possibility of left-over assets, while women’s jewellery, and so on, had already been disposed of.78 In other words, the women would have been left with virtually no bargaining strength within the family at a much earlier stage in the process of famine impoverishment.79 Within the bargaining view of the family, at this point non-cooperation by the husband would make sense in the interests of his individual physical survival, and do him no harm in terms of his social survival (he could marry again), whether or not this was in keeping with the moral order. If he additionally had an ideological (moral) justification that self-preservation was in keeping with the preservation of the moral order, it would merely have eased the decision.

Essentially, Greenough’s view of the family (and one which Appadurai implicitly accepts) comes close to what Sen (1987) has described as the ‘despotic’ family in which the male family head takes all decisions and others just obey (indeed ‘acquiesce even as victims’)—although Greenough sees the male head as essentially a benevolent despot, guided by moral considerations. What I am suggesting here is that the victimization process could fit in equally with the co-operative-conflict (bargaining) view of the family.

The point, however, is that what view we take of the family—the despotic one or the co-operative-conflict one—could point to basically different policy (p.228) conclusions. Under the former, one could, for instance, make a case for strengthening the economic position of the benevolent male head of household to enable him better to support his wife and children (the passive recipients of his bounty or rejection). Under the latter view where women are seen as active (if disadvantaged) agents in the arrangements of reciprocity within the family, one would make a case for strengthening their fall-back positions and bargaining power within the family. That this view of the family may be the more accurate is suggested not the least by the experiences of some of the non-governmental initiatives (discussed in the next section of this chapter), where the enhancement of poor women’s earnings opportunities, such as by cheap credit, has made a noteworthy difference to their position within the family.

5. On Policy and External Interventions

Our examination of a poor agricultural household’s coping mechanisms during periods of food shortages associated with seasonality and calamity, reveals several aspects which have a bearing on policy.

Firstly, seasonality reveals a face of the family which is essentially one of (unequal) co-operation; famine mirrors one of disintegration. In both contexts, the burden of coping falls disproportionately on female members within poor households—in terms of consumption adjustments, asset depletion, work burden, and, in extreme contexts, destitution and abandonment. Seen within the framework of families as arenas of co-operative conflicts, this unequal burden may be traced (among other things) to women’s weaker bargaining strength within the family.

Secondly, common property resources and forests play a critical role in cushioning the effect of seasonality and, although to a lesser extent, also of drought-related food troughs for poor households, especially for the more vulnerable members within such households whose ownership endowments are few and entitlement position relatively weak.

Thirdly, a poor agricultural household’s long-term capacity to survive the trough can depend crucially on whether it is able to pass through such periods without any permanent depletion in its assets, especially (but not only) in terms of livestock and land.

Following from this, policies that stem the ongoing depletion and erosion of village commons and forests and indeed help expand this communal base; alternative credit and employment systems which protect the household from having to reduce consumption, or mortgage or sell assets, including those owned by women; and any measures (in addition to those above) which strengthen women’s fall-back position by enhancing their entitlements base, would contribute to strengthening the coping mechanisms of poor households.

(p.229) The question then is: what would or could be the intervening forms and agencies for doing this? In much of the existing literature on seasonality or calamity there is a strong emphasis on direct State intervention especially in the form of public works, and relatively little on interventionist forms that do not directly depend on the State.

State success in preventing the Maharashtra drought of 1970–3 from escalating into a large-scale famine, and in significantly containing the extent of productive asset depletion, nutritional distress, and excess mortality through the provision of guaranteed employment on a massive scale through public works, appears to have been particularly influential in pointing policy in this direction. There are clearly significant lessons to be learnt about drought and famine management from the Maharashtra example (as brought out by Drèze’s (1988b) excellent and insightful review paper). Not the least is the fact that the State was able to protect entitlements to food not only of the most vulnerable households but also of the more vulnerable sections within the household, such as women, by providing them with direct (and not male-mediated) entitlements in the form of guaranteed employment. It is noteworthy that in recent years women have constituted at least 40 per cent according to official figures (and over 50 per cent according to field studies) of the labour force in the Maharashtra Employment Guarantee Scheme (EGS) (Dandekar 1983).

At the same time, in the context of the present discussion, several factors would caution against dependence on the State alone for providing food security in contingencies. First, although the government has been able to deal fairly effectively with the threat of large-scale famine since independence, and in situations verging on famine (for example, due to successive droughts) there may perhaps be few alternatives to substantial State interventions in the form of public works (as Drèze 1988b argues), the record of the Indian government in dealing with less acute situations has been poor.80 Even in the implementation of the EGS in Maharashtra there are shortcomings, and it has not been possible to absorb all those who seek work through it even in normal times;81while outside Maharashtra attempts at guaranteeing employment have had even less success: in Tamil Nadu they have been found to be largely ineffective (Guhan 1981).

Second, there are contradictions in State policies themselves. For instance, those relating to CPRs, forests, and environmental protection in general are noted to have systematically weakened the ability of the poor themselves to cope with food crisis situations.82

(p.230) Third, the degree to which the State responds to the demands of the vulnerable sections would depend not least on the degree to which these sections can make their demands heard—be it for minimum wage implementation, relief works, or for a guaranteed subsistence throughout the year. Sen (1984), for instance, lays particular stress on the role of a vigilant press and diverse political parties in giving voice to those affected by situations bordering on famine. At the same time, he notes the limitations of these agencies in dealing with regular and endemic malnutrition. It is in giving the poor a voice that the role of grass-roots social action groups in which the poor are direct participants is likely to have particular relevance (in complementing if not substituting for the role played by some of the political parties).83 The issue here is one not only of being entitled, but of being able effectively to enforce those entitlements.

To elaborate, consider the concepts of ‘enfranchisement’ and ‘empowerment’ in relation to that of entitlement. Sen’s entitlement approach ‘concentrates on the ability of people to command food through the legal means available in the society, including the use of production possibilities, trade opportunities, entitlements vis-à-vis the State, and other methods of acquiring food’ (Sen 1981: 45). However, access to what a person may legitimately or legally be entitled to is often mediated by political economy considerations. In seeking to incorporate these considerations, Appadurai (1984) introduces the notion of enfranchisement, which he defines as ‘the degree to which an individual or group can legitimately participate in decisions of a given society about entitlements’. He suggests that command over food during famine, for instance, would depend not only on entitlements but also on enfranchisement:

In the patterns of victimisation during the Bengal famine of 1943–44 we see those who are enfranchised—the controlling classes, the male heads of households in general— deciding to withhold the entitlement rights of dependents, women and children. Entitlement without enfranchisement, the fate of rural clients, women, children, slaves and household pets in many societies, is not a safe condition when famine sets in.

He argues that over the last century the rural poor in South Asia have traded a situation in which they were entitled (to receive help from village patrons, for instance) without being enfranchised, for one in which they are partly enfranchised (through the spread of universal franchise and mobilization of the poor by political parties) without being securely entitled.

One might, however, argue that the mere participation in the process of decision-making about entitlement (that is, enfranchisement), is not yet sufficient to guarantee entitlement. Here adding the concept of empowerment could take us a step further. This term has been used variously in recent years by several social action groups in India, and in the present context may be (p.231) defined as the ability of an individual or group legitimately to ensure that decisions relating to entitlement are taken in its favour (be it within the family, or of the family vis-à-vis the community or the State).

The Nepalese example may be recalled here of the effectiveness of group pressure from the poorer families in the village in getting credit and grain, and in maintaining wage rates and many of the traditional elements of patron-client support through a severe drought. However, it is the experience of grass-roots organizing in South Asia which has significant lessons to offer on this count. Group organizing has been one of the crucial means of empowering the vulnerable sections—the poor, the low caste, the women—not only the better to enforce their legal entitlements within the community and family, but also to expand the scope of these legal entitlements through agitating for changes in the laws themselves.84

A review of the nature, experience, and effectiveness of these groups— which vary considerably in their issue-focus, organizational form, and political outlook—is not attempted here; but for illustrative purposes some features of selected groups providing credit to the rural poor are worth highlighting, especially since credit has received a considerable emphasis in State policies on poverty alleviation, but very little success in reaching the poorest,85 and is seen to play an important role in the family’s coping mechanisms.

One such initiative is that of the Grameen Bank of Bangladesh launched as an experiment by an individual in 1976 and, after its proven success, institutionalized by the Bangladesh government in 1983 as a specialized credit agency for the rural poor.86 Today it has 250 branches in five districts, and a membership of 200,000 in 3,700 villages. Among its noteworthy features are the following: first, its clientele is essentially class-homogenous—it caters exclusively to the landless and those owning 0.5 acres or less of land. Second, it follows a group approach to loan disbursement: borrowers must form groups of five persons each, each group selecting its own chairperson and secretary and holding weekly meetings. While loans are given to individuals there is an implicit group pressure and responsibility for repayment.87 Third, the loans are given without collateral, for any viable income-earning activity of an individual’s choice, as well as for joint enterprises such as leasing in land for joint cultivation, investment in shallow tube-wells, and so on. Fourth, it recognizes the especially vulnerable position of women within the family, and (p.232) seeks to draw them in as members—the credit operation for women being entirely independent of the transactions with male borrowers and needing no mediation by husbands or male relatives. Fifth, the concept of empowerment by collective activity and solidarity is emphasized in meetings and in special training programmes. Sixth, repayment is made easy by enabling loans to be repaid in weekly instalments at a meeting in the village itself to which the bank worker comes for this purpose. This is especially helpful to women borrowers in a purdah society. Seventh, built into the system are various social-security schemes such as a group savings fund set up primarily to advance consumption loans; an emergency insurance fund against default, accident, and deaths; and various other schemes such as for house-loans, and so on. Eighth, to ensure that bank ownership stays with the members, each member has compulsorily to buy shares in the bank.

Since its inception there have been several evaluations of the Grameen Bank88 which indicate first of all that the credit indeed reaches the target group, spill-overs being minimal.89 Second, the loan recovery rate is about 97–99 per cent. Third, per-capita income per borrowing household has increased significantly.90 Fourth, the borrowing households have been able to build up some capital assets and also improve their consumption. Fifth, the bank has improved the bargaining power of the families vis-à-vis the landlords, enabling them to raise their wages.91 Sixth, women have been significant gainers: by recent figures they constitute 70 per cent of the borrowers, receive 55 per cent of the cumulative disbursement, and are considered by the Bank to be better credit risks (their repayment record is better), with a higher sense of social responsibility and making a more productive use of the loan than the men (women divert a smaller percentage of the loan to consumption). It is also noteworthy that most of the female borrowers manage the loans themselves (despite pressure from husbands),92 in general presenting a very different face from Greenough’s ‘acquiescing victims’. As a result of the investments, their incomes have increased multifold, especially since most of them were unemployed prior to taking the loans.93 Their standing within the family is also noted to have improved:

(p.233) its membership and access to institutional credit also means an access to some fixed assets and it endows those women with a special status within the family. Other family members become more conscious about the comfort and well being of the loanee women (Rahman 1986).

Another study found a significant reduction in verbal and physical abuse and threats of divorce by husbands since their wives joined the Bank (Ahmed 1985). Problems of replicability, of expanding the Bank’s geographic coverage without reducing efficiency in functioning, and of sustaining noted income increases over time, of course, remain.

An equally important example, in the Indian context, is that of SEWA (the Self Employed Women’s Association, started in 1972 in Ahmedabad city and since spread to several other cities and the rural areas of Gujarat), which focuses on the economic livelihood needs of poor women, through credit and other means.94 First of all, for providing credit it has set up its own cooperative bank mainly drawing on member deposits. Loans are advanced not only for production purposes without collateral but also for consumption needs, and for redeeming any assets such as jewellery or land that women may have had to mortgage to the money-lender. Large consumption loans are also given against jewellery as collateral. A fixed savings deposit scheme protects their savings from pressures for withdrawal from family members. Second, the association functions as a union for women in various trades, each trade organized into a group composed of anything up to 100 or more members, with group leaders elected from among them. Third, also like the Grameen Bank, SEWA operates a variety of social-security schemes including a group-based insurance fund, widowhood insurance, life insurance, maternal protection, and so on. Fourth, specifically in the rural areas, in order to reduce the burden of seasonal unemployment among women agricultural labourers (employment is available only for three months during the year in the area) and to increase their bargaining power vis-à-vis the employers, diversification of incomes is being promoted by reviving traditional crafts such as weaving, pottery, and so on, through craft co-operatives (and training), and through dairy co-operatives —a link being forged for this purpose with the National Dairy Development Board and the IRDP programme of the government. Like the Grameen Bank, SEWA too appears to have been successful in raising incomes (although there are no studies quantifying the exact income benefits), strengthening women’s position within the family, raising consciousness about the advantages of group solidarity, and, for the rural women agricultural workers, raising agricultural wages by strengthening their bargaining power vis-à-vis the employers.

(p.234) The Working Women’s Forum in South India,95 which has much in common with SEWA, again caters exclusively for poor women, loans being given to groups, with the additional feature that while the loans are in individual names there is group liability,96 and the elected group leader has to personally guarantee repayment.

Group-financed insurance schemes against food troughs are in fact a feature of several social-action initiatives. For instance, the Bhoomi Sena Movement of tribals in Maharashtra, as part .of its struggle for higher wages, minimum wage implementation, and access to land, set up a grain bank on the basis of crops threshed collectively and stored for the next season, as an insurance against seasonal fluctuations and harvest failures. In addition, a fund was set up to provide loans to small farmers for bullocks, seed, and consumption needs (Silva et al 1979).

All these are also potentially effective substitutes for the declining traditional arrangements based on feudal patronage and exploitation; and, where group ventures are undertaken, they offer the potential advantages and economies of labour deployment and team work that joint families are noted to have in the Maclachlan example. A more detailed review and evaluation of such initiatives from the point of view of social security for the poor, and a comparison with existing State attempts at providing similar services, are clearly needed. Equally useful would be a systematic review of the effectiveness of initiatives that have sought to provide direct relief against seasonality97 and droughts;98as well as of those which are explicitly agitational in nature, and which would include many groups and regional movements fighting for the enforcement of minimum wages, for a right to the land they have been tilling for generations, for the protection of forests and the environment, and working against intra-family inequalities and violence.99

Here it would suffice to note that underlying these and many like ventures (be they those providing services such as credit, or those with more explicitly political overtones) is a recognition that the effective implementation of existing State laws, programmes, and even relief measures, can be contingent (p.235) on group initiative and pressure from below; and that the group needs to be economically and often also socially (by caste, gender) homogeneous.100

It is difficult to say whether these ventures, by providing women with an alternative to the family for survival and so improving their fall-back position, will in the long run weaken the family as a unit or strengthen it. But either way, they offer the possibility of more egalitarian intrafamily gender relations.

(p.236) References

Bibliography references:

Agarwal, B. (1984), ‘Rural Women and the HYV Rice Technology’, Economic and Political Weekly, 31 Mar.

——— (1986a), Cold Hearths and Barren Slopes: The Wood fuel Cńsis in the Third World. Delhi: Allied Publishers; London: Zed Books, Maryland: Riverdale Press.

——— (1986b), ‘Women, Poverty and Agricultural Growth in India’, Journal of Peasant Studies, 13.

——— (1988), ‘Who Sows? Who Reaps? Women and Land Rights in India’, Journal of Peasant Studies, 15.

——— (1989), ‘Rural Women, Poverty and Natural Resources: Sustenance, Sustainability and Struggle for Change’, Economic and Political Weekly, 28 Oct.

Ahmed, M. (1985), Status, Perception, Awareness and Marital Adjustment of Rural Women: The Role of Grameen Bank. Dhaka: Grameen Bank Publication.

Alamgir, M. (1980), Famine in South Asia: Political Economy of Mass Starvation. Cambridge, Mass: Oelgeschlager and Hain Publishers.

Appadurai, A. (1984), ‘How Moral is South Asia’s Economy? A Review Article’, Journal of Asian Studies, 43.

Banerjee, N. (1988), ‘Women’s Work and Family Strategies: A Case Study from Bankura, West Bengal’, mimeo, Centre for Women’s Development Studies, New Delhi.

Bardhan, K. (1977), ‘Rural Employment, Wages and Labour Markets in India: A Survey of Research’, Economic and Political Weekly, 25 June, 2 July, 9 July.

Bardhan, P. (1983), ‘On Life and Death Questions: Poverty and Child Mortality’ in P. Bardhan, Land, Labour and Rural Poverty: Essays in Development Economics. Delhi: Oxford University Press.

——— and Rudra, A. (1981), ‘Terms and Conditions of Labour Contracts: Results of a Survey in West Bengal, 1979’, Oxford Bulletin of Economics and Statistics, 43.

Behrman, J. R. (1986), ‘Intrahousehold Allocation of Nutrients in Rural India: Are Boys Favoured? Do Parents Exhibit Inequality Aversion?’ mimeo, Center for Analysis of Developing Economies, University of Pennsylvania.

Bhalla, S. (1977), ‘New Relations of Production in Haryana Agriculture’, Economic and Political Weekly, Review of Agriculture, 27 Mar.

Binswanger, H. P., and Rösenzweig, M. R. (1986), ‘Credit Markets, Wealth and Endowments in Rural South India’, paper presented at the VIIIth World Congress of the International Economic Association, New Delhi, 1–5 Dec.

Bliss, C., and Stern, N. (1982), Palanpur: The Economy of an Indian Village. Oxford: Oxford University Press.

Borkar, V. V., and Nadkarni, V. (1975), Impact of Drought on Rural Life. Bombay: Popular Prakashan.

Boulding, E. (1976), The Underside of History: A View of Women Through Time. Boulder, Colo.: Westview.

Brara, R. (1987), ‘Shifting Sands: A Study of Rights in Common Pastures’, mimeo, Institute of Development Studies, Jaipur.

Breman, J. (1979a), ‘Seasonal Migration and Cooperative Capitalism: The Crushing of Cane and Labour by the Sugar Factories of Bardoli, South Gujarat’, paper presented at the ADC-ICRISAT Conference on Adjustment Mechanisms of Rural Labour Markets in Developing Areas, ICRISAT, Hyderabad, 22–4 Aug.

(p.237) ——— (1979b), Patronage and Exploitation: Changing Agrarian Relations in South Gujarat. Delhi: Manohar Publications.

——— (1985), Of Peasants, Migrants and Paupers: Rural Labour Circulation and Capitalist Production in West India. Delhi: Oxford University Press.

Briscoe, J. (1979), ‘Energy Use and Social Structure in a Bangladesh Village’, Population and Development Review, 5.

Brown, K., Black, R. E., and Becker, S. (1982), ‘Seasonal Changes in Nutritional Status and the Prevalence of Malnutrition in a Longitudinal Study of Young Children in Rural Bangladesh’, Abstract in American Journal of Clinical Nutrition, 36.

Burling, R. (1963), Rengsanggri: Family and Kinship in a Garo Village. Philadelphia: University of Pennsylvania Press.

Cain, M. T., Khanam, S. R., and Nahar, S. (1979), ‘Class, Patriarchy and the Structure of Women’s Work in Rural Bangladesh’, Working Paper No. 4, Center for Population Studies, The Population Council, New York.

Caldwell, J. C., Reddy, P. H. and Caldwell, P. (1984), ‘The Determinants of Family Structure in Rural South India’, Journal of Marriage and the Family, Feb.

——— (1986), ‘Periodic High Risk as a Cause of Fertility Decline in a Changing Rural Environment: Survival Strategies in the 1980–1983 South Indian Drought’, Economic Development and Cultural Change, 34.

Campbell, D. J., and Trechter, D. D. (1982), ‘Strategies for Coping with Food Consumption Shortage in the Mandara Mountains Region of North Cameroon’, Social Science and Medicine, 16.

Chambers, R. (1981), ‘Introduction’, in Chambers et al. (1981a).

——— Longhurst, R., and Pacey, A. (eds.) (1981a), Seasonal Dimensions of Rural Poverty. London: Frances Pinter.

——— (1981b), ‘Seasonality in Rural Experience’, in Chambers et al. (1981a).

Chandler, A. F. D. (1986), Participation as Process: What We Can Learn from Grameen Bank, Bangladesh. Oslo: NORAD.

Chattopadhyaya, K. P., and Mukherjee, R. K. (1946), A Plan for Rehabilitation. Calcutta: Statistical Publishing House.

Chaudhury, A., Huffman, S. L., and Chen, L. C. (1981), ‘Agriculture and Nutrition in Matlab Thana, Bangladesh’, in Chambers et al. (1981a).

Chen, M. (1987), ‘Update on Drought Situation in Devdholera Village’, mimeo, Harvard Institute for International Development, Harvard University.

——— (1988), ‘Women and Household Livelihood Systems’, mimeo, Harvard Institute for International Development, Harvard University.

Chinnappa, B. N., and Silva, W. P. T. (1977), ‘Impact of the Cultivation of High-Yielding Varieties of Paddy on Employment’, in B. H. Farmer (ed.), Green Revolution? (London: Macmillan).

Chowdhury, K. M., and Bapat, M. T. (1975), ‘A Study of Impact of Famine and Relief Measures in Gujarat and Rajasthan’, Research Study No. 44, Agro-Economic Research Centre, Sardar Patel University, Ahmedabad.

Clay, E. (1981), ‘Seasonal Patterns of Agricultural Employment in Bangladesh’, in Chambers et al. (1981a).

(p.238) Commander, S. (1983), ‘The Jajmani System in North India: An Examination of its Logic and Status Across Two Centuries’, Modem Asian Studies, 17.

Connell, J., and Lipton, M. (1977), Assessing Village Labour Situations in Developing Countries. Delhi: Oxford University Press.

——— Dasgupta, B., Laishley, R., and Lipton, M. (1976), Migration from Rural Areas: The Evidence from Village Studies. Delhi: Oxford University Press.

Copestake, J. (1987), ‘Loans for Livelihood? Government Sponsored Credit Schemes in India: Proposals for Reform’, mimeo, Department of Agricultural Economics and Management, University of Reading.

CSE (Centre for Science and Environment) (1985–6), The State of India’s Environment: A Citizen’s Report. Delhi: Centre for Science and Environment.

Currey, B. (1978), ‘Famine Symposium Report: The Famine Syndrome, Its Definition for Relief and Rehabilitation in Bangladesh’, Ecology of Food and Nutrition, 7.

Dandekar, K. (1983), Employment Guarantee Scheme: An Employment Opportunity for Women. Pune: Orient Longman.

Dasgupta, B., Laishley, R., Lucas, H., and Mitchell, B. (1977), Village Society and Labour Use. Delhi: Oxford University Press.

Dasgupta, M. (1987), ‘Informal Security Mechanisms and Population Retention in Rural India’, Economic Development and Cultural Change, July.

Dasgupta, S., and Maiti, A. K. (1987), ‘The Rural Energy Crisis, Poverty and Women’s Roles in Five Indian Villages’, Technical Co-operation Report, World Employment Programme, International Labour Organization, Geneva.

Desai, G. M., Singh, G., and Sah, D. C. (1979), ‘Impact of Scarcity of Farm Economy and Significance of Relief Operations’, CMA Monograph No. 84, Indian Institute of Management, Ahmedabad.

Desai, R. X. (1982), ‘Migrant Labour and Women: The Case of Ratnagiri’, World Employment Programme Research Working Paper No. WEP 10/WP28, International Labour Organization, Geneva.

Deshpande, V. (1982), Employment Guarantee Scheme: Impact on Poverty and Bondage among Tribals. Pune: Tilak Maharashtra Vidyapeeth.

Dhagamvar, V. (1987), ‘The Disadvantaged and the Law’, paper presented at a Workshop on Poverty in India, Queen Elizabeth House, Oxford, Oct.

Dirks, R. (1980), ‘Social Responses during Severe Food Shortages and Famine’, Current Anthropology, 21.

Dixon-Mueller, R. (1979), Rural Women at Work: Strategies for Development in South Asia. Baltimore: Johns Hopkins University Press.

Drèze, J. (1988a), ‘Social Insecurity in India: A Case Study’, paper presented at STICERD/WIDER Workshop on Social Security in Developing Countries, London School of Economics, July 1988.

——— (1988b), ‘Famine Prevention in India’, Discussion Paper No. 3, Development Economics Research Programme, London School of Economics; to be published in J. P. Drèze and A. K. Sen (eds.) (1990), The Political Economy of Hunger. Oxford: Oxford University Press.

——— and Mukherjee, A. (1987), ‘Labour Contracts in Rural India: Theories and Evidence’, Discussion Paper No. 7, Development Economics Research Programme, London School of Economics.

——— and Sen, A. (1989), Hunger and Public Action. Oxford: Clarendon Press.

(p.239) Epstein, S. (1967), ‘Productive Efficiency and Customary Systems of Rewards in Rural South India’, in R. Firth (ed.), Themes in Economic Anthropology. London: Tavistock Publications.

Fernandes, W., and Menon, G. (1987), Tribal Women and Forest Economy: Deforestation, Exploitation and Status Change. Delhi: Indian Social Institute.

——— (1988), Forests, Environment and Tribal Economy: Deforestation, Exploitation and Status Change. Delhi: Indian Social Institute.

Food and Nutrition (1985), ‘Special Issue on Minor Crops, Seasonality, Women’s Activities: Three Crucial Issues in Rural Food Production and Consumption’, 11/2.

Gaiha, R. (1988), ‘Are the Chronically Poor also the Poorest in Rural India’, mimeo, Faculty of Management Studies, University of Delhi.

Gangrade, K. D. and Dhadda, S. (1973), Challenge and Responses: A Study of Famines in India. Delhi: Rachana Publications.

George, S. (1988), ‘The Female of the Species: Women and Dairying in India’, paper presented at a Workshop on Women in Agriculture, Centre for Development Studies, Trivandrum, 15–17 Feb.

Ghodake, R. D., Ryan, J. G., and Sarin, R. (1978), ‘Human Labour Use in Existing and Prospective Technologies of the Semi-Arid Tropics of Peninsular India’, Progress Report, Economics Program-1, ICRISAT, Hyderabad.

Ghose, A. K. (1979), ‘Short-term Changes in Income Distribution in Poor Agrarian Economies: A Study of Famines with Reference to the Indian Sub-Continent’, World Employment Programme Research Working Paper, WEP 10–6/WP28.

GOI (Government of India) (1945), Famine Enquiry Commission Report on Bengal. Delhi: Government of India Press.

——— (1979), ‘Rural Labour Enquiry 1974–1975, Final Report on Wages and Earnings of Rural Labour Households’, Chandigarh: Labour Bureau, Ministry of Labour.

——— (1981), ‘Rural Labour Enquiry 1974–75, Final Report on Employment and Unemployment of Rural Labour Households’, Chandigarh: Labour Bureau, Ministry of Labour.

——— (1982), Report of Committee on Forests and Tribals in India. Ministry of Home Affairs.

——— (1987), Integrated Rural Development Programme and Allied Programme of Training of Rural Youth for Self-Employment (TRYSEM) and Development of Women and Children in Rural Areas (DWCRA): A Manual. Delhi: Dept. of Rural Development, Ministry of Agriculture, Government of India.

Greenough, P. R. (1982), Prosperity and Misery in Modern Bengal: The Famine of 1943–1944. Oxford: Oxford University Press.

Guhan, S. (1981), ‘Social Security: Lessons and Possibilities from the Tamil Nadu Experience’, Madras Institute of Development Studies Bulletin, 11.

Gulati, L. (1978), ‘Profile of a Female Agricultural Labourer’, Economic and Political Weekly, Review of Agriculture, 25 Mar.

Harriss, B. (1986), ‘The Intra-Family Distribution of Hunger in South Asia’, paper presented at a Conference on Food Strategies held at WIDER, Helsinki, July 1986; to be published in J. P. Drèze and A. K. Sen (eds.) (1990), The Political Economy of Hunger. Oxford: Clarendon.

Harriss, J. (1977), ‘Implications of Changes in Agriculture for Social Relations at the (p.240) Village Level: The Case of Randan’, in B. H. Farmer (ed.), Green Revolution? London: Macmillan.

Hill, P. (1980), ‘Joint Families in Rural Karnataka, India’, Modern Asian Studies, 14.

Hirway, I. (1985), ‘Garibi Hatao: Can IRDP Do It? A Discussion’, Economic and Political Weekly, 30 Mar.

Hossain, M. (1984), ‘Credit for the Rural Poor: The Experience of Grameen Bank in Bangladesh’, mimeo, Bangladesh Institute of Development Studies.

——— (1987), The Assault that Failed: A Profile of Absolute Poverty in Six Villages of Bangladesh. Geneva: UNRISD.

——— (1988), ‘Credit for Alleviation of Poverty: The Grameen Bank of Bangladesh’, Research Report No. 65, International Food Policy Research Institute, Washington, DC.

IDS Bulletin (1986), ‘Seasonality and Poverty’, 17/3.

Jansen, E. G. (1986), Rural Bangladesh: Competition for Scarce Resources. Oslo: Norwegian University Press.

Jetly, S. (1987), ‘Impact of Male Migration on Rural Females’, Economic and Political Weekly, 22/44.

Jodha, N. S. (1975), ‘Famine and Famine Policies: Some Empirical Evidence’, Economic and Political Weekly, 10/41.

——— (1978), ‘Effectiveness of Farmer’s Adjustment to Risk’, Economic and Political Weekly, 13/25.

——— (1979), ‘Intercropping in Traditional Farming Systems’, Progress Report 3, Economics Programme, ICRISAT.

——— (1981), ‘Role of Credit in Farmers’ Adjustment Against Risk in Arid and Semi-Arid Tropical Areas of India’, Economic and Political Weekly, 15/42–3.

——— (1983), ‘Market Forces and Erosion of Common Property Resources’, paper presented at the International Workshop on Agricultural Markets in the Semi-Arid Tropics, ICRISAT, Andhra Pradesh, October.

——— (1985aø), ‘Population Growth and the Decline of Common Property Resources in Rajasthan’, Population and Development Review, 11.

——— (1985bò), ‘Social Research on Rural Change: Some Gaps’, paper presented at the Conference on Rural Economic Change in South Asia: Differences in Approach and in Results between Large Scale Surveys and Intensive Micro-Studies, Bangalore, 5–8 Aug.

——— (1986), ‘Common Property Resources and Rural Poor’, Economic and Political Weekly, 5 July.

Kalpagam, U. (1987), ‘Working Women’s Forum: A Concept and an Experiment in Mobilisation in the Third World’, Offprint 6, Madras Institute of Development Studies.

Kapadia, K. M. (1966), Marriage and Family in India. London: Oxford University Press.

Kar, P. C. (1982), The Garos in Transition. New Delhi: Cosmo Publications.

KFRI (Kerala Forest Research Institute) (1980), Studies on the Changing Pattern of Man-Forest Interactions and Its Implications for Ecology and Management. Trivandrum: Kerala Forest Research Institute.

Kolenda, P. (1987), Regional Differences in Family Structures in India. Jaipur: Rawat Publications.

(p.241) Kulkarni, S. W. (1974), Survey of Famine Affected Sinnar Taluka. Poona: Gokhale Institute of Politics and Economics.

Kulkarni, S. (1983), ‘Towards a Social Forestry Policy’ % Economic and Political Weekly, 6 Feb.

Kumar, B. G. (1987), ‘Poverty and Public Policy: Government Intervention and Levels of Living in Kerala, India’, unpublished Ph.D. thesis, University of Oxford.

Kumar, G., and Stewart, F. (1987), ‘Tackling Malnutrition: What Can Targeted Nutritional Interventions Achieve?’, paper presented at a Workshop on Poverty in India, Queen Elizabeth House, Oxford, Oct.

Kumar, S. K. (1978), ‘Role of the Household Economy in Child Nutrition at Low Incomes’, Occasional Paper No. 95, Dept. of Agricultural Economics, Cornell University.

Kynch, J., and Maguire, M. (1988), ‘Wasted Cultivators and Stunted Girls: Variations in Nutritional Status in a North Indian Village’, mimeo, Institute of Economics and Statistics, Oxford.

Leslie, J. (1988), ‘Women’s Work and Child Nutrition in the Third World’, World Development, 16.

Lewis, O. (1958), Village Life in Northern India: Studies in a Delhi Village. Urbana: University of Illinois Press.

Lipton, M. (1983a), ‘Labour and Poverty’, World Bank Staff Working Paper No. 616, The World Bank, Washington, DC.

——— (1983b), ‘Poverty, Undernutrition and Hunger’, World Bank Staff Working Paper No. 597, The World Bank, Washington, DC.

Longhurst, R., and Payne, P. (1981), ‘Seasonal Aspects of Nutrition’, in Chambers et al. (1981a).

McAlpin, M. (1983), Subject to Famine: Food Cńsis and Economic Change in Western India, 1860–1920. Princeton, NJ: Princeton University Press.

Maclachlan, M. D. (1983), Why They Did Not Starve: Biocultural Adaptation in a South Indian Village. Philadelphia: Institute for the Study of Human Issues.

Madan, T. N. (1975), ‘Structural Implications of Marriage in North India: Wife-givers and Wife-takers Among the Pandits of Kashmir’, Contńbutions to Indian Sociology, 9.

Mahalanobis, P. C., Mukherjee, R., and Ghosh, A. (1946), A Sample Survey of After-Efſects of the Bengal Famine of 1943. Calcutta: Statistical Publishing Society.

Maįumdar, D. N. (1978), Culture Change in Two Garo Villages. Calcutta: Anthropological Survey of India.

Marum, E. M. (1981), Women at Work in Bangladesh. Dhaka: BRAC.

Mencher, J. (1987), ‘Women’s Work and Poverty: Women’s Contribution to Household Maintenance in Two Regions of South India’, in D. Dwyer and J. Bruce (eds.), A Home Divided: Women and Income Control in the Third World. Stanford: Stanford University Press.

——— and Saradamoni, K. (1982), ‘Muddy Feet and Dirty Hands: Rice Production and Female Agricultural Labour’, Economic and Political Weekly, Review of Agriculture, 25 Dec.

Mies, M. (1983), ‘Landless Women Organise: A Case Study of an Organisation in Rural Andhra’, Manushi, 3.

Miller, B. (1981), The Endangered Sex: Neglect of Female Children in Rural North India. Ithaca, NY: Cornell University Press.

(p.242) Minturn, L., and Hitchcock, I. T. T. (1966), Rajputs of Khalapur. New York: John Wiley & Sons.

Mitra, M. (1985), ‘Women and Work in the Livestock Economy: An Introduction’, in Women in Dairying: A Set of Case Studies and Recommendations. Delhi: Ford Foundation.

——— (1988), ‘Women’s Work and Household Survival Strategies: A Case Study of Santal Women’s Lives and Work’, draft, Centre for Women’s Development Studies, Delhi.

Montgomery, E. (1977), ‘Social Structuring of Nutrition in Southern India’, in L. Green (ed.), Malnutrition, Behaviour and Social Organisation. New York: Academic Press.

Morris, D. M. (1974), ‘What is a Famine?’ Economic and Political Weekly, 2 Nov.

——— (1975), ‘Needed: A New Famine Policy’, Economic and Political Weekly, 10/5–7.

Mundle, S. (1979), Backwardness and Bondage: Agrarian Relations in a South Bihar District. Delhi: Indian Institute of Public Administration.

Nath, J. (1979), ‘The Role of Women in Rural Bangladesh: A Study of Nantupur Village’, in Women for Women: Situation of Women in Bangladesh. Dhaka: BRAC.

NCAER (National Council of Applied Economic Research) (1981), Report on Rural Energy Consumption in Northern India, Environment Research Committee. Delhi: National Council of Applied Economic Research.

Noponen, H. (1987), ‘Organising Petty Traders and Home-Based Producers: A Case Study of Women’s Forum, India’, in A. M. Singh and A. Kelles-Vittamen (eds.), Invisible Hands. Delhi: Sage.

Osmani, S. R. (1987), ‘Controversies in Nutrition and their Implications for the Economics of Food’, Working Paper No. 16, WIDER; to be published in J. P. Drèze and A. K. Sen (eds.) (1990), The Political Economy of Hunger. Oxford: Clarendon.

Oughton, L. (1982), ‘The Maharashtra Droughts of 1970–73: An Analysis of Scarcity’, Oxford Bulletin of Economics and Statistics, 44.

Patel, V., Indu, R., and Patel, V. P. (1975), ‘Employment in Rural Gujarat: A Study of Four Villages in Anand Taluka, 1970–1971’, Agricultural Economics Research Centre, Sardar Patel University, Vallabh Vidyanagar.

Patil, S. (1978), ‘Famine Conditions in Maharashtra: A Survey of Sakri Taluka’, Economic and Political Weekly, 28 July.

PIDT (People’s Institute for Development and Training) (1982), Towards Participation: Case Studies of Small Farmers Development Programme Nepal. New Delhi: People’s Institute for Development and Training.

Pingle, V. (1975), ‘Some Studies of Two Tribal Groups of Central India. Part 2: Importance of Foods Consumed in Two Different Seasons’, Plant Food for Man, 1.

Prindle, P. H. (1979), ‘Peasant Society and Famines: A Nepalese Example’, Ethnology, 1.

Rahman, A. (1986), ‘Impact of Grameen Bank Intervention on the Rural Power Structure’, Working Paper No. 2, Grameen Bank Evaluation Project, Bangladesh Institute of Development Studies, Dhaka.

Rahaman, M. M. (1981), ‘The Causes and Effects oí Famine on the Rural Population, A Report from Bangladesh’, in J. R. K. Robson, (ed.), Famine: Its Causes, Effects and Management. New York: Gordon & Breach.

(p.243) Rahman, R. I. (1986), ‘Impact of Grameen Bank on the Situation of Poor Rural Women’, Working Paper No. 1, Grameen Bank Evaluation Project, Bangladesh Institute of Development Studies, Dacca.

Rajgopalan, S., Kymal, P. H., and Pu-ai, P. (1981), ‘Births, Work and Nutrition in Tamil Nadu, India’, in R. Chambers et al. (1981a).

Randhawa, M. S. (1980), A History of Agriculture in India. Delhi: Indian Council of Agricultural Research.

Rangaswami, A. (1985), ‘Women’s Roles and Strategies during Food Crisis and Famines’, International Workshop on Women’s Roles in Food Self-sufficiency and Food Strategies, 14–19 Jan., ORSTOM, Paris.

Rao, N. V. K. (1974), ‘Impact of Drought on the Social System of a Telengana Village’, Eastern Anthropologist, 27.

Rao, V. M., and Erappa, S. (1987), ‘IRDP and Rural Diversification: A Study in Karnataka’, Economic and Political Weekly, Dec.

Reddy, G. P. (1988), ‘Drought and Famine: The Story of a Village in a Semi-arid Region of Andhra Pradesh’, paper presented at a Workshop on Afro-Asian Studies on Social Systems and Food Crisis, India International Centre, 26–9 Mar.

Rosenzweig, M. R. and Schultz, T. P. (1982), ‘Market Opportunities, Genetic Endowment and Intrafamily Resource Distribution: Child Survival in Rural India’, Center paper No. 323, Economic Growth Center, Yale University.

Rudra, A. (1982), ‘Extra-Economic Constraints on Agricultural Labour: Results of An Intensive Survey of Some Villages Near Shantiniketan, West Bengal’, Asian Employment Programme Working faper, ARTEP, ILO, Bangkok.

Ryan, J. G., and Ghodake, R. D. (1980), ‘Labour Market Behaviour in Rural Villages in South India: Effects of Season, Sex and Socio-economic Status’, Progress Report, Economic Programme 14, ICRISAT.

——— Bidinger, P. D., Rao, N. P., and Pushmpamma, P. (1984), ‘The Determinants of Individual Diets and Nutritional Status in Six Villages of Southern India’, Research Bulletin No. 7, ICRISAT, Hyderabad.

——— and Wallace, T. D. (1985), ‘Determinants of Labour Market Wages, Participation and Supply in Rural South India’, Progress Report 73, Economics Group, Resource Management Program, ICRISAT, Hyderabad.

Salva, C. (1973), ‘Some Aspects of Out-Migration from Gujarat’, Ph.D. thesis, Bombay University.

Scott, J. C. (1976), The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia. New Haven, Conn.: Yale University Press.

Sebstad, J. (1982), ‘Struggle and Development Among Self-Employed Women: A Report on the Self-Employed Women’s Association, Ahmedabad’, Report submitted to USAID, Washington, DC.

Sen, A. K. (1981), Poverty and Famines: An Essay on Entitlement and Deprivation. Delhi: Oxford University Press.

——— (1982), ‘Food Battles: Conflicts in the Access to Food’, Coromandel Lecture No. 1, New Delhi, 13 Dec.

——— (1983), ‘Economics and the Family’, Asian Development Review, 1.

——— (1984), Resources, Values and Development. Delhi: Oxford University Press.

——— (1987), ‘Gender and Cooperative Conflicts’, mimeo, Department of Economics, Harvard University.

(p.244) Sen, A. K., and Sengupta, S. (1983), ‘Malnutrition of Rural Children and the Sex Bias’, Economic and Political Weekly, 19, Annual Number.

Sen, I. (1988), ‘Class and Gender in Work Time Allocation’, Economic and Political Weekly, 13 Aug.

Sethi, H. (1987), ‘Refocusing Praxis’, Report of the UNU/UNDP Interaction Programme of Senior Activists/Researchers in South Asia, held in collaboration with PIDA, Sri Lanka, and SETU, Lokayan, Delhi.

Shah, A. B. (1973), The Household Dimension of the Family in India. Delhi: Orient Longman.

Sharma, U. (1980), Women, Work and Property in North-West India. London: Tavistock Publications.

Shiva, V. (1988), Staying Alive: Women, Ecology and Survival in India. Delhi: Kali for Women.

Siddiqui, K. (1984), An Evaluation of the Grameen Bank. Dhaka: Grameen Bank Publication.

Singh, K. S. (1975), The Indian Famine 1967: A Study in Cńsis and Change. Delhi: People’s Publishing House.

Silva, G. V. S., Mehta, N., Rahman, M. A., and Wignaraja, P. (1979), ‘Bhoomi Sena: A Struggle for People’s Power’, Development Dialogue, 2.

Srivastava, R. (1977), ‘Shifting Cultivation in India’, Man in India, 57.

Torry, W. I. (1987), ‘Evolution of Food Rationing Systems with Reference to African Group Farms in the Context of Drought’, in M. H. Glantz (ed.), Drought and Hunger in Africa. Cambridge: Cambridge University Press.

Vatuk, S. (1981), ‘Sharing, Giving and Exchanging of Foods in South Asian Societies’, mimeo, final draft, University of Illinois at Chicago Circle.

Vaughan, M. (1987), The Story of an African Famine: Gender and Famine in Twentieth Century Malawi. Cambridge: Cambridge University Press.

Wade, R. (1987), ‘The Management of Common Property Resources: Finding a Cooperative Solution’, Research Observer, 2.

Watts, M. (1983), Silent Violence: Food, Famine and Peasantry in Northern Nigeria. Berkeley: California University Press.

White, B. (1976), ‘Population, Involution and Employment in Rural Java’, in G. E. Hansen (ed.), Agricultural Development in Indonesia. Ithaca, NY: Cornell University Press.

Wiser, W. H. (1936), The Hindu Jajmani System. Lucknow: Lucknow Publishing House.

Yunus, M. (1982), ‘Experience in Organising Grass-Root Initiatives and Mobilising People’s Participation: The Case of Grameen Bank Project’, paper presented at SID World Conference, Baltimore.

Notes:

(1) In his more recent writings, Sen (1987) substitutes this term by the term ‘break-down position’.I prefer to retain ‘fall-back position’.

(2) It could depend too on State and non-traditional support systems. These will, however, be discussed specifically in Section 4 below.

(3) e.g. National Sample Survey (NSS), 1977–8, conducted by the Government of India.

(4) See Lipton’s (1983a) review piece; also Patel et al. (1975) and Chinnappa and Silva (1977).

(5) Also see Jodha’s (1986) estimates of involuntary unemployment based on village-level data collected at the International Crop Research Institute for Semi-Arid Tropics (ICRISAT, Hyderabad) relating to four districts in three states.

(6) See the analysis in Ghodake et al. (1978) for six villages in Maharashtra and Andhra Pradesh; Dasgupta et al. (1977) for a review of village studies across India; and Clay (1981) for a comparison of four districts in Bangladesh.

(7) Clay (1981) notes, for instance, that across four geographically dispersed districts of Bangladesh, Sept./Oct. continues to be the slackest period, with little increase in employment in this period despite technological change.

(8) See Drèze and Mukerjee (1987) for a useful review of and discussion on the reasons for this lower variability.

(9) See especially Hossain’s (1987) six-village study of poor households in Bangladesh, which gives a breakdown of earnings by seasons among landless and land-poor households. Also see Lipton’s (1983a) review.

(10) Rudra (1982) notes this for West Bengal, and Drèze and Mukherjee (1987) for Uttar Pradesh.

(11) See e.g. Caldwell et al. (1986); Chen (1988); Jodha (1979); Majumdar (1978).

(12) See e.g. Majumdar’s (1978) study of tribals in Meghalaya, and Hossain’s (1987) study in Bangladesh.

(13) About the migrants from the arid parts of West Maharashtra to South Gujarat, Breman (1979a) notes that the limited employment opportunities and overall poverty cause labour to migrate out for a large part of the year, and particularly ‘in times of serious and substantial drought—as occurs once every three to four years on average—the supply is virtually inexhaustible’.

(14) Connell et al. (1976: 39) in their survey of migration patterns as revealed by village studies note: ‘Almost everywhere, migration concentrates extremely heavily on villagers aged 15–30.’

(15) Bardhan (1977: A-45) notes: ‘The phenomenon of tribal and scheduled caste women in paddy areas migrating in the busy season to take up agricultural work like transplantation in prosperous villages seems not to be an isolated one but rather of some general validity.’

(16) Similarly, Mencher and Saradamoni (1982) observed in some Thanjavur villages that harvesting operations are customarily done by teams of two—a man and a woman. This seriously disadvantages female-headed households, especially if they do not have a brother or son to team up with them.

(17) See Agarwal (1984) who gives the operation-wise use of labour by gender in three Indian States.

(18) Ryan and Ghodake (1980). Also, in a sample study of 432 West Bengal villages in 1972–3, women of agricultural labour households showed the greatest percentages of slack-season decline in both average wage-rate and average duration of work: ‘A female labourer’s weekly wage earnings dropped…during the slack quarter to 20% of the (levels) during the busiest quarter. In the case of a male labourer, they dropped to 67%…’ (quoted in Lipton 1983 a: 85).

(19) By the NSS ‘usual status’ criterion, in 1983, of the rural women workers of over 5 years of age, 35% were employed as casual labour on a daily or piece-rate basis; 62% were self-employed (in family enterprises) in agricultural and rural non-agricultural work, and 3% in regular employment (waged or salaried) especially in domestic service. In contrast, of the male workers, 29% were casual agricultural labourers, 61% self-employed, and 10% in regular employment.

(20) In the table, based on the Rural Labour Enquiry of 1973–4, the annual earnings from agricultural wage work of women agricultural labourers are less than half of the men in five states, and close to half in the rest.

(21) A Karnataka-related study of rural labour found that 70% of male labour contracts and only 20% of female labour contracts had meal provisions (Ryan and Wallace 1985: 24).

(22) The notion of ‘territorial purdah’ (prevailing in much of northwest India)—the effective segregation of village space by gender whereby there are clearly identifiable spaces which essentially constitute male areas, such as the bazaar (market-place), that women are expected to avoid—strongly disadvantages them in their search for employment and in managing land independently. Although seniority, age, whether she is a daughter or daughter-in-law, her class/ caste all affect a woman’s freedom of movement, so that older women with grown-up sons, village daughters, and women of poor and low caste families enjoy greater liberty, even they are supposed to avoid spaces of predominantly male presence (see especially Sharma 1980; also personal observation). These restrictions are even greater in countries such as Bangladesh and Pakistan.

(23) See Gulati (1978) for Kerala; Mencher and Saradamoni (1982) and Mencher (1987) for Kerala, Tamil Nadu, and West Bengal; and Dasgupta and Maiti (1987) for Himachal Pradesh, Madhya Pradesh, Uttar Pradesh, Maharashtra, and Assam. The last study finds that in terms of wage earnings alone women’s contribution ranged from 20% of household annual earnings in Assam to 26% in Himachal Pradesh.

(24) Other micro-studies, which have looked at male and female expenditure patterns within poor households with earners of both sexes, provide additional evidence to the effect that while female earnings typically go towards the family’s basic needs of food, fuel, etc., a not insignificant part of male earnings goes towards tobacco and liquor (see Gulati 1978; Mies, et al. 1983; Sharma 1980; Mencher and Saradamoni 1982). A study for rural Punjab also showed that an agricultural labour household with a heavy drinker on average spent 40 per cent less on food per capita and less on clothing than a ‘non-drinker’ household (Dass, quoted in Harriss 1986).

(25) See Dasgupta and Maiti (1987) who also note that a man rarely cooks except if the wife is seriously ill and in the absence of a daughter to substitute for her.

(26) Ryan and Ghodake (1980) find this in all six villages of Andhra Pradesh and Maharashtra in which they gathered time allocation data by gender. Also see Sen (1988) for Madhya Pradesh, and Agarwal (1984) for a review of other studies.

(27) Unfortunately Jodha does not provide an assessment of the relative productivity of or earnings from the different sources of employment. Also where a greater number of days spent in CPR collection reflects lower productivity (e.g. with deforestation women spend longer hours collecting the same quantum of firewood), the ‘days of CPR employment’ could be a misleading indicator in cross-time as well as cross-regional comparisons. However it is still useful as an indicator of some productive use of time as opposed to no employment at all.

(28) See e.g. Burling (1963) on the Garos of the north-east.

(29) See e.g. Boulding (1976); also Shiva (1988).

(30) See Brara’s (1987) village study in Rajasthan, and Agarwal (1986a) for several examples from the Chipko movement for forest protection and regeneration in the hills of Uttar Pradesh.

(31) Figures provided by the National Remote Sensing Agency, Government of India.

(32) See Wiser (1936); Lewis (1958); Epstein (1967); Vatuk (1981); Commander (1983); Breman (1985); Dasgupta (1987); Torry (1987). See also Chap. 4 above.

(33) With the possible exception of priests who constituted a somewhat distinct category. Commander (1983: 295), for instance, observes: ‘The purohit is thus donated grain in an act of symbolic obeisance, while, on the contrary, the bulk of kamins are given grain in return for labour…’

(34) Vatuk’s (1981) survey of literature on the jajmani system indicates that such fixed payments did not bear a direct relationship to the actual amount of work done per year, which was variable. The nature of tasks to be completed were, however, clearly specified. Also see Commander (1983) on interregional differences in payments.

(35) Wiser (1936), who was the first to conceptualize and describe the jajmani relationship as an interrelated system, observes: ‘Each serves the others. Each in turn is master. Each in turn is servant’ (p. xxiii). However others have questioned this apparent symmetry on the basis both of Wiser’s own ethnography and of other empirical evidence (see esp. the discussion in Commander 1983). Breman (1985), for instance, notes that the permanent farm servants (halis) in south Gujarat got from the master ‘every thing necessary for a bare living’ but at the same time ‘they were at their master’s beck and call not only during the day but in the evening as house servants and if necessary at night as well’.

(36) Epstein’s (1967) contention that the system worked well even during years of bad harvests in that the landlords, in order to keep lower castes alive, reduced their own consumption and shared available grain with clients, so that consumption between all village households tended towards equality in such situations, has, however, been disputed on both methodological and empirical grounds (see esp. Chap. 4 above; also Torry 1987).

(37) In the two north Indian villages studied by Dasgupta, it is now found profitable to sell milk and transport it out, substantially reducing the supply of buttermilk earlier transported free to needy villagers. Also dung is now increasingly valuable to owners as manure and fuel, and the village council has ruled that dung belongs solely to the cattle owners and cannot be collected free. Also see Commander (1983) for additional evidence on this decline.

(38) Post-marital residence is found to be patrilocal among virtually all caste and class groups in India, the exceptions being a few matrilineal tribes of the north-east and some communities (such as the Nayars) in the south-west, which traditionally practised matrilocality, although even these have been shifting to patrilocality over the past few decades (Agarwal 1988). Of course patrilocality need not in itself necessitate women’s leaving their parental village, if village endogamy is allowed and practised, as it is in many parts of India—here the constraints on women seeking natal kin support would be less severe, as discussed further on.

(39) During fieldwork in Kithoor village (Rajasthan) I found that most of the twenty-odd agricultural labour women with whom I raised this issue said that they preferred to borrow from the money-lender even at usurious rates or from other villagers, rather than from their natal kin for fear of ‘spoiling the relationship’ if they defaulted. Jetly (1987) found this as well in her study of the impact of male migration on rural females in central and eastern Uttar Pradesh. She notes that women avoided borrowing from relatives ‘since this brings loss of prestige’, and a neighbour or friends were more commonly approached for small sums.

(40) For a review and discussion of these cross-regional variations in marriage patterns based on village studies see Agarwal (1988).

(41) Such linkages could also restrict the labourer’s freedom to migrate seasonally.

(42) For instance, women in poor rural households in West Bengal and Bangladesh are known to put aside a handful of rice (mustischaal) when cooking the day’s main meal. This then becomes a store for use in lean seasons (see e.g. Siddiqui 1984).

(43) See e.g. Sen (1981); Sen and Sengupta (1983); Harriss (1986); Lipton (1983b); Behrman (1986); Agarwal (1986a, 1989); Bardhan (1983); Rosenzweig and Schultz (1982); Drèze and Sen 1989.

(44) See especially Harriss’s (1986) review of ICRISAT data.

(45) The likely reason for this (as already noted in n. 24 above) is that in such households women’s earnings in much greater proportion than men’s tend to go towards fulfilling basic needs. Also, Kumar notes: ‘since women do the shopping for food and preparation of family meals, it is possible that their participation in the generation of income gives them better control over its allocation and they might more equitably distribute assets among family members’.

(46) Leslie (1988) arrives at broadly similar conclusions on the basis of a fairly comprehensive review of existing studies for Third World countries on the relationship between women’s work, child nutrition, and infant feeding practices. While noting the methodological problems that confound clear-cut answers she observes that: (1) there is surprisingly little evidence of a negative effect of women’s work on either breast-feeding or child nutritional status; (2) few studies measured women’s income but most of those that did found a more positive effect of the mother’s income than household income on child nutritional status.

(47) Hossain (1987x: 90)Hossain from his six-village survey in Bangladesh notes: ‘the lowest point in per capita foodgrain consumption for households…is neither the summer trough nor the fall or spring slack time, but rather the winter busy season. Reserves are usually not sufficient to permit inhabitants to break out of the long hungry slump, beginning during the rainy season and continuing unabated into winter.’ (Also refer to Table 5.11.)

(48) Quoted in Longhurst and Payne (1981).

(49) See e.g. Chambers et al (1981b); Rajgopalan et al (1981); Ryan et al (1984); Brown et al (1982).

(50) For instance the peaking of births before the main transplanting season noted in Tamil Nadu negatively affected breastfed infants (Rajgopalan et al 1981). Of course in so far as smaller sized persons have lower BMR needs, women on average being of smaller build than men may ceteris paribus need less food to survive in a situation of nutritional constraint (Osmani 1987). But clearly much depends on the extent of nutritional reduction.

(51) This should be taken only as a very rough assessment.

(52) There appears to have been much more research on these aspects in the African context. See, for instance, various articles in IDS Bulletin (1986) and Food and Nutrition (1985).

(53) During the Bihar drought of 1966–7, children were found eating the skin of toddy palms, jungle berries, edible roots, mushrooms, and even these were not available in plenty (Gangrade and Dhadda 1973). In Andhra Pradesh during the drought years of 1983–7, women collected a cactus creeper to be mixed with lentils, wild fruit, vegetables, leaves, and so on (Reddy 1988); rats, and grains stored by rats, were eaten during the Karnataka drought of 1965 (Maclachlan 1983); ground nettles are reported to have been consumed in the 1987–8 drought in Rajasthan; snails plus a variety of food otherwise considered inedible during the 1943 Bengal famine (Greenough 1982); and wild arum, plantain saplings, leaves, and rice husk during the Bangladesh famine of 1974 (Currey 1978; Rahaman 1981).

(54) During the Andhra Pradesh drought, claims of the landless to dead tamarind trees from the community orchard were rejected by the panchayat secretary who chose to sell them to his followers at a nominal price (Reddy 1988). Again, denial of rights to collect free dung and the imposition of fines on those who persist is enforced by the panchayat on behalf of cattle owners in villages in the north (Lewis 1958; Dasgupta 1987).

(55) See e.g. Jodha (1978) who makes a strong plea for extending consumption credit to the poor as a drought relief measure. Also see Gaiha (1988) who, on the basis of all India panel data, compares the situation of the chronically poor in areas that faced three years of adverse weather conditions between 1968 and 1970, and finds that borrowings among this category of households increased sharply as a result, with 42% in 1970, relative to 19% in 1968, disinvesting.

(56) This includes migration to irrigated areas along with bullocks to work as share-croppers, or to commercial centres with bullocks and camel carts for transporting goods; gangs of youths moving to irrigated or non-drought-affected areas in the peak seasons; and cross-state migration with animals in search of greener pastures—the last being most important in West Rajasthan.

(57) Maclachlan’s descriptions suggest that when he speaks of ‘joint’ families, he has more than one married son living with the parents in mind, but he does not explicitly state this. Other scholars examining household structures in India have used more precise classifications (e.g. Kolenda 1987; Caldwell et al. 1984; Hill 1980). While there appears to be agreement among them that commensal units consisting of a conjugal couple and unmarried children constitute ‘nuclear’ families, and those with two married siblings living with or without the parents are ‘joint’ families, the classification of those with only one married child living with the parents differs. Such families are classified by Kolenda as ‘joint’ but by Caldwell et al. as ‘stem’. Shah (1973) also provides a useful review and discussion on the subject.

(58) Detailed case-studies in Karnataka by Caldwell et al. (1984) in 9 villages and Hill (1980) in 6 villages also indicate that nuclear families are in the majority. Caldwell et al. found that 7% of the households were ‘joint’ and 27% were ‘stem’, while Hill found that 10% were ‘joint’. However, both the studies note that several times the number observed to be living in joint families at any one point in time would have lived in such families at some point in their lives—typically during the first few years of marriage. A ‘developmental cycle’ pattern is thus noted to exist, and one which appears to have persisted over several generations in the past.

(59) Kolenda’s (1987) cross-country review suggests some slight tendency towards a decline in jointness in parts of India; while Caldwell et al. (1984) argue that their Karnataka evidence shows no apparent change. However their conclusion does not entirely tie up with some of their observations which suggest a possible decline. They note, for instance (see especially pp. 223–4), that the proportion of younger couples leaving to seek work elsewhere (which is one of the causes of joint family partition) ‘has been rising constantly’ and further that ‘certainly, the case studies of the causes of partition appeared to show that conditions that were tolerated without partition only a generation ago are much less likely to be tolerated now’.

(60) See Kolenda (1987) on the regional patterning of family structures as well as on cross-class and cross-caste variations. On class differences also see Caldwell et al. (1984) and Hill (1980) both of whom find a systematically lower incidence of jointness among the poorer households in the Karnataka villages. Caldwell et al. find a systematic decline in jointness with a fall in land-holding size. Also landed households tend to have larger houses, capable of accommodating larger families. However, neither Caldwell et al. nor Hill find variations in family structures by caste although they provide no satisfactory explanation of why this should be so, given the usually noted correlation between caste and economic situation.

(61) Jodha (1978) on comparing drought and post-drought years in Rajasthan and Gujarat found a decline in protective food consumption by 28% to 48% (varying by region) and a decline in per capita grain intake by between 11% and 23%.

(62) Reddy (1988); Currey (1978) on Bangladesh.

(63) For a quantification of declines in consumption during droughts see Jodha (1975, 1978); also see Chowdhury and Bapat (1975) for Rajasthan and Gujarat; Desai et al. (1979) for Gujarat.

(64) See e.g. Jodha (1978); Singh (1975); Caldwell, et al. (1986).

(65) See Caldwell et al. (1986) on the Karnataka drought of 1983; Borkar and Nadkarni (1975) on the Maharashtra drought on 1970–3; and Jodha (1978) on droughts in 5 districts of Rajasthan, Gujarat, and Maharashtra at various points in time.

(66) See e.g. Oughton (1982) on the Maharashtra drought of 1970–3; Mahalanobis et al. (1946) and Ghose (1979)on the Bengal famine of 1943; Alamgir (1980) on the Bangladesh famine of 1974.

(67) Jodha (1975); Singh (1975); Borkar and Nadkarni (1975); Greenough (1982); Currey (1978).

(68) Borkar and Nadkarni (1975) found that during the Maharashtra drought in 1972–3 the price of jewellery did not register any fall while that of livestock registered a steep decline. A sharp decline in livestock prices is common in such contexts as sellers begin to flood the market.

(69) See Agarwaľs (1988) review both of the legal position and of customary practices among 145 communities across India, based on village studies.

(70) In Africa, small livestock are typically owned by women, and again are among the first casualties during famines (Watts 1983).

(71) In some communities a portion of the dowry is customarily taken away by the in-laws (see e.g. Madan 1975; Miller 1981; Minturn and Hitchcock 1966).

(72) Famine relief in general was much delayed and, when it came, was grossly inadequate (Greenough 1982:127–38; GOľs Famine Enquiry Commission Report on Bengal 1945:96–102).

(73) A fair number of widows who had earlier been supported by relatives are also noted to have been turned out (Chattopadhyay and Mukerjee 1946: 15).

(74) Gangrade and Dhadda (1973) also mention cases of abandonment of women, but especially of children and the aged, during the 1966–7 Bihar drought. Also Dirks (1980: 30) in his crosscountry review of the social responses to famine, notes: ‘Both sale and abandonment of children have been recorded in a wide variety of cultural settings.’

(75) The concept of pativrata popularized by the Puranic writers implied absolute fidelity to the husband and demanded that the wife always remain chaste and pure. The Dharmashastras stressed this as the guiding principle of a woman’s life (Kapadia 1966).

(76) As noted, this is also part of the age-group within which the number of female destitutes was substantially higher than male destitutes in the Mahalanobis et al. (1946) survey.

(77) It is of note that during the 1943 Bengal famine, there was a virtual absence of government famine relief works, which are known typically to attract more women than men, and which would have provided an alternative source of earnings.

(78) Greenough (1982: 197) notes: One of the first signs of economic distress in rural Bengal was the sale of women’s jewellery and ornaments.’ As distress continued, household utensils followed, and cooking implements, brass pots, tin roofs all found their way to the market. In 1943, in Contai town, sales of utensils and jewellery were noted to be ‘brisk’. Only subsequently were cattle and land sold. Currey (1978) documents a similar sale of household utensils during the Bangladesh famine of 1974.

(79) In this context, Vaughan’s (1987: 124) observation on the experience of women (married uxorilocally) during the Malawi famine of 1949–50 is also striking: ‘it was the men who had the external social and economic linkages—both with other ecological zones and with the labour economy. These external linkages became more and more crucial as the famine increased in severity. The scarcer the food the smaller became the units of consumption. Beyond a certain point the advantages of co-residence with kin ceased to be apparent and women were increasingly dependent on husbands (if they had any) or on themselves’ (emphasis mine).

(80) See e.g. various articles on drought management in the Economic and Political Weekly, especially over the years 1987–1988.

(81) See especially Deshpande (1982) for a critique of EGS functioning in the adivasi (tribal) parts of Thane district in Maharashtra. See also Chap. 7 below.

(82) For a critique of these policies see especially discussions in CSE (1985–6); Agarwal (1986a); Shiva (1988); and various articles over the past 5 years in the Economic and Political Weekly.

(83) It is necessary to distinguish between different political parties in this respect—especially in terms of the extent to which they incorporate the concerns of the poor in their political agendas.

(84) See especially Dhagamvar (1987), and various issues of the Lawyers’ Collective (Bombay).

(85) See especially the recent literature criticizing the government’s IRDP programme, including Copestake (1987), Hirway (1985), Rao and Erappa (1987), Drèze (1988a), and Chap. 7 below.

(86) For details see especially Siddiqui (1984), Chandler (1986), and Hossain (1988).

(87) In contrast, the government’s IRDP loan subsidy scheme for the rural poor in India targets at individual families (GOI 1987). If the husband defaults, the wife is refused a loan. However, as a subscheme of the IRDP, a pilot project DWCRA (Development of Women and Children in the Rural Areas) has been initiated in 1983 in 50 districts to try out the group approach for loan disbursement to women.

(88) Ahmed (1985); Hossain (1984, 1988); Siddiqui (1984); R. Rahman (1986); A. Rahman (1986); Chandler (1986).

(89) In 1984, 86% of the borrowers were the landless, only 5.6% were found to be owning more than the specified 0.5 acres, and l.7% had over 1 acre.

(90) By one estimate the increase was on average 32% over 2 years (1980–2) after taking the loan, during which period the per-capita income of Bangladesh increased only by 2.6% (Hossain 1984).

(91) In some areas the increases are estimated to be as high as 25% (Yunus 1982).

(92) By R. Rahman’s (1986) evaluation of the scheme in 5 villages, of the female borrowers, 77.4% kept 75% of the loans under their own management, and only 12% surrendered the entire loan sum to the male head of household.

(93) The average income of female borrowers (who invest mainly in milch cattle, paddy processing, beef-fattening, and oil-pressing), in R. Rahman’s (1986) evaluation, was 5,140 Taka (constituting 33% of the household income) while that of a non-borrowing housewife in the target group of the project village was 92 Taka. Twenty-two per cent of the female borrowers contributed 50% or more of the family income, and some 30% earned 6,000 Taka per year, sufficient (according to Rahman) to maintain two persons above the officially specified poverty line.

(94) Description based on Sebstad (1982) and conversations with Ela Bhatt (who initiated SEWA).

(95) See especially Noponen (1987); and Kalpagam (1987).

(96) A pilot project launched under the Small Farmer’s Development Programme in Nepal followed a similar procedure of individual loans and group liability with considerable success (PIDT 1982).

(97) See Marum (1981) for an evaluation of several food-for-work programmes for women, initiated by non-governmental organizations in Bangladesh to deal with seasonal troughs in employment.

(98) Several such initiatives sprung up during the 1987–8 drought in various parts of India for fodder distribution, well digging, etc.

(99) For a brief review of the women-related initiatives see Agarwal (1989); on ecology movements see Shiva (1988); for an overview of these and other movements and initiatives see Sethi (1987); also various articles in the Lokayan Bulletin (Delhi) over the past two years.

(100) Social and economic homogeneity is typically found to be important in ensuring the successful functioning of such initiatives. See Dixon-Mueller (1979); PIDT (1982); and Wade (1987) whose examples of successful collective action by villagers for regulating the use of common property resources are of special interest in the present context.