Background Report on Gender Issues in India

Report No 32

Background report on gender issues in India: key findings and recommendations Short report
Report commissioned by the Overseas Development Administration (DFID), UK

February 1995 (revised)

The authors gratefully acknowledge support for the preparation of this report from the Overseas Development Administration (DFID), UK. However, the views expressed and any errors or omissions are those of the authors and not of the DFID, UK.

BRIDGE (development - gender) Institute of Development Studies University of Sussex Brighton BN1 9RE, UK Tel: [phone] Fax: [phone]/691647 Email: [e-mail] Website: http://www.ids.ac.uk/bridge ISBN: 1 85864 167 5 © Institute of Development Studies, Brighton

CONTENTS PAGE LIST OF ABBREVIATIONS GLOSSARY EXECUTIVE SUMMARY KEY RECOMMENDATIONS FOR DEVELOPMENT ASSISTANCE 1. 2. INTRODUCTION: contemporary contexts for the gender question THE ECONOMY, STRUCTURAL ADJUSTMENT AND GENDER 2.1 Women and men in the Indian economy 2.2 Employment status and wages by gender 2.3 Gender and poverty 2.4 Structural adjustment and gender: employment effects 2.5 Structural adjustment and gender: income and consumption effects CASTE, COMMUNALISM AND GENDER 3.1 Everyday caste in India 3.2 The emergence of caste as a national political issue 3.3 Caste and gender 3.4 Tribals and gender 3.5 Gender and communalism GENDER BIAS IN SOCIAL SERVICES: health and family planning, water, sanitation and housing 4.1 Social sector policy, funding and biases 4.2 Health institutions and health policy 4.3 Gender and health 4.4 Gender bias in access to health care 4.5 Family Planning 4.6 HIV/AIDS 4.7 Water, sanitation and housing GENDER BIAS IN SOCIAL SERVICES: education 5.1 Women in educational policy 5.2 Trends in enrolment, retention and literacy 5.3 Differential access to education by religion and caste 5.4 Constraints and opportunities for female education THE STATE, NGOs AND THE MAINSTREAMING OF GENDER 6.1 The women’s movement in India 6.2 Government policy on women and the national machinery 6.3 Women in development schemes 6.4 NGO approaches to gender issues 6.5 Gender training 6.6 Women’s participation in panchayati raj ANNEX 1: Terms of reference 1 3 3 5 6 8 12 13 14 15 16 18 19 21 21 23 25 28 29 30 32 34 34 35 39 40 43 43 46 48 50 53 55

3.

4.

5.

6.

LIST OF ABBREVIATIONS CEDAW DWCRA EGS GDP GER HIV/AIDS IDA IRDP ISST JRY LHV NCERT NFE NGO NIRD NRY OBC PHC SAMATA SIAAP STD TLC UMA UN WDP WHO WID Convention for the Elimination of Discrimination Against Women Development for Women and Children in Rural Areas (Government programme) Employment Guarantee Scheme (Maharashtra) Gross Domestic Product Gross Enrolment Ratio Human Immunodeficiency Virus/ Acquired Immunodeficiency Syndrome International Development Association Integrated Rural Development Programme Institute for Social Studies Trust Jawahar Rozgar Yojana (government wage employment scheme in rural areas). Lady Health Visitor National Council for Educational Research and Training Non-formal education Non-government organisation National Institute for Rural Development (Hyderabad) Nehru Rozgar Yojana (government wage employment scheme in urban areas) Other Backward Caste Primary Health Centre Women’s organisation South India AIDS Action Project Sexually Transmitted Disease Total Literacy Campaign Utsashi Mahila Abhyudaya (information and resource centre on women and panchayati raj in Bangalore) United Nations Women’s Development Programme World Health Organisation Women in Development

GLOSSARY agarbatti arrack bidi Dalit madrassa Mahila Samakhya Mahila Mandals Other Backward Castes panchayati raj purdah sarpanches sathin sati Scheduled Tribe Scheduled Caste incense stick Home made country liquor hand-rolled cigarette Self-chosen name for ‘untouchables’; literally ‘downtrodden’ Traditional Muslim religious school Empowerment and development programme aimed at women Women’s centres set up after by the government after Independence ‘Lowest’ groups within caste system; classified as Other Backward Castes by Mandal Commission in 1981 local self-government system of seclusion for women sometimes involving mandatory veiling Leader of panchayat village level animator of Women’s Development Programme (Rajasthan) Historical upper-caste ritual of self-immolation by a widow on the funeral pyre of her husband or son Tribal groups officially designated as disadvantaged in the post-Independence era for government programmes Those outside caste system (‘untouchables’) who were officially designated as a disadvantaged group in the postIndependence period for government programmes Government journal

Yojana

EXECUTIVE SUMMARY PREAMBLE 1. This Background Report on Gender Issues in India was commissioned by the Overseas Development Administration (ODA), UK, to help prepare a new Women in Development (WID) strategy for its India operations.1 It is intended to help technical, field and desk personnel within ODA increase the gender-sensitivity of the India aid programme. 2. Gender and development is becoming an accepted perspective in practical development work. A focus on gender issues means looking at both women and men, whilst recognising that it is women who suffer from gender inequality and discrimination. The importance of gender concerns to successful project and programme implementation and to promoting social welfare and economic efficiency as well as in addressing social, economic and political inequalities, is now widely acknowledged. ODA has a declared objective to ‘promote the social, economic, political and legal status of women’. 3. This Executive Summary highlights the principal findings of the report and makes key recommendations for future development assistance. The Short Report (56 pages) elaborates the main points and gives more detailed recommendations. The Full Report (173 pages) is an in-depth study of current literature and thinking on women and gender and issues in India today. PRINCIPAL FINDINGS The background 4. A rapidly globalising economy, an intensely communal culture and a rising caste consciousness are major forces shaping India’s development today. Recent events - the decision to implement the Mandal Commission recommendations on caste-based reservations in 1990; the announcement of a New Economic Policy in 1991; and the outbreak of communal violence in 1992 - will have lasting impacts which are as yet difficult to discern. These events will influence not only material gender inequalities, but also the construction of gender identities and the way in which questions of gender are debated in years to come. They have also fostered relations of conflict between women along caste and communal lines. Poverty, adjustment and gender 5. Female headship of households and well-documented gender biases in intra-household resource allocation in India suggest that women are specially vulnerable to poverty, but there
1

See Annex 1 for Terms of Reference

1

are complications. Gender biases within the household may be less severe in lower income groups and not all female-headed households are poor. It is not helpful to either poverty alleviation efforts, or interventions to promote gender equity, to see women and the poor as synonymous. 6. Poverty alleviation schemes have provided employment opportunities for the rural and, more recently, urban poor with limited success, especially in asset creation. Special schemes for women have focused on empowerment rather than poverty alleviation per se; their impact on livelihoods is uncertain. 7. The effects of India’s liberalising New Economic Policy on employment and personal incomes is not yet clear. The employment prospects for women are mixed but look particularly bleak for women who are not young or educated. Rising food grain prices and the rationalisation of the public distribution system may lead to lower real wages and increased vulnerability to food insecurity for low income households, particularly in urban areas. Urban poverty is becoming a major problem in India and may intensify under adjustment. There is thus a case for increased interventions towards the urban poor and especially poor urban women. 8. In the agricultural sector, some think increased agricultural prices will create growth and improved employment opportunities for the rural poor. Others fear that the supply response will be limited, partly due to the decline in state investment in rural infrastructure, and that few new labour opportunities will result, especially for women. Caste, communalism, tribal issues and gender 9. Recent events have pushed caste and communal issues to the forefront of national political life in India. Previously, dominant political interest groups have tended to be ‘caste-blind,’ representing the interests of upper caste Hindus. Minority religious groups, as well as lower castes, Dalits and tribals, may distrust existing political structures, parties and movements, including the women’s movement. Women are singled out for special attention and attack as bearers of their respective caste and communal identities. Dalit women have been subject to sexual harassment and attack with virtual impunity. 10. In spite of the reservations policy, Dalit and minority (Muslim, tribal) groups are poorer and have less access to formal employment, education and health care than the population as a whole. The limited impact of the reservations policy has mainly benefited men. On all indicators, women from these groups fare worst. 11. Although tribal communities are increasingly heterogeneous and integrated into the cash economy, they remain overwhelmingly concentrated in rural, often remote, forested areas. They are therefore particularly affected by deforestation and other forms of environmental degradation. This gives special opportunities for interventions to improve their situation. But the benefits of development programmes aimed at Scheduled Castes and Scheduled Tribes have tended to be captured by non-tribals.
2

3

Gender bias in social service provision 12. Gender bias in access to health care, education and other social services (e.g. water supply, sanitation and housing) is marked in India for two main reasons. There are inequalities within the household in food allocation and in spending on health care and education; and there are also clearly established gender biases in service delivery. 13. The combined result is that girls are less likely to receive medical care than boys, less likely to be admitted to hospital for treatment, and, in some instances, less likely to survive illnesses than boys. The poor nutritional status of girls and women means that their illnesses are of longer duration. Millions of children in India, but particularly girls, never attend school or drop out soon after initial enrolment. Considerable efforts are being made to increase girls’ access to education but the relative contribution of different interventions to increasing female enrolment is poorly understood. 14. Declines in spending on social sector services (except in education since 1985) have reduced the overall coverage and quality of public sector social services. Most households, including the very poor, use private health facilities. The increasing burden of social sector expenditure on poor households is likely to reinforce existing patterns of gender bias, unless incentives are created for investment in female health and education. 15. There is increasing emphasis on community participation and NGO involvement in social sector service delivery. This may create more accountable services and encourage greater input of local people, including women, into the design of services, but it cannot be assumed that NGOs (or private agents) will be any more efficient, participatory or gender-aware than state agencies. Women’s participation in community-based service provision is constrained by their existing work burden. To improve women’s participation, clear incentives, training and rewards for women need to be built in. The institutional context: the state, NGOs and mainstreaming 16. Women’s issues are increasingly taken seriously by the Indian Government (e.g. increasing references to women in Five Year Plan documents, the issue of special policy documents on women; and the setting up of various national and state-level bodies to handle women’s affairs). There has also been an increase in special development schemes for women and in gender training of government functionaries. 17. However, there is no consistent overall policy framework on gender issues to which government departments can be held accountable. The national machinery dealing with women and development lacks co-ordination, resources and power. The system needs reform. 18. The women’s movement in India has had considerable impact on policy but, with liberalisation and decentralisation, it needs to develop new strategies and uniform interests between women can no longer be assumed. There may be new opportunities at the state level
4

to demand more comprehensive policies on gender and funding commitments for women. Training and support for increased participation of women in local government through the one third reservations system ratified in 1993 is an area of considerable current activity for women’s organisations. 19. Although there are several thousand NGOs in India, the proportion of these explicitly targeting women is low and they are concentrated in particular regions. NGOs are under increasing pressure to become alternative service deliverers to government, which may limit their flexibility, capacity for innovation and advocacy role. Scaling up may pose particular problems for women’s organisations and their multi-functionality may also limit their access to resources where these are channelled on a sectoral basis. Some women’s organisations may wish to retain an independent, social movement character, rather than become professionalised and reliant on external funds. 20. There are pioneering areas of NGO activity with a gender perspective in India which may need support, for example, attempts to increase the asset base (land, forests etc.) and secure the livelihoods of poor rural women; and HIV/AIDS awareness programmes, linked to other health issues and to women’s empowerment.

5

KEY RECOMMENDATIONS FOR DEVELOPMENT ASSISTANCE Tackling the gender dimensions of poverty _ - It is important to understand what different groups of women themselves would consider to be an improvement in their lives, to inform the design of poverty alleviation programmes. - A better understanding of the processes through which female headship arises is needed to identify particular sub-groups of female headed households who face extreme hardship. - The participation of women in poverty alleviation schemes needs to be improved both quantitatively and qualitatively and urban coverage extended. Factors which would facilitate women’s participation need to be identified by drawing on the lessons from the most successful schemes. - Project and programme level monitoring of the impact of structural adjustment on the livelihoods of the urban poor could assist in informing the design of poverty alleviation programmes. - Improved targeting of the public distribution system is needed to ensure maximum benefits to both the rural and urban poor and specifically to women in these groups. Taking account of caste, communal and other inequalities - Interventions should be checked at appraisal and design stages to ensure that resources are not being delivered unduly to already privileged groups, given the concentration of poverty among Dalit and tribal groups, especially women, and also urban Muslim communities. - When working through local institutions, specific mechanisms are needed to elicit the opinions of low-caste, non-Hindu, Dalit and tribal groups, and particularly women from within these groups, who are under-represented in existing political structures. - In planning investments in social infrastructure, facilities should be appropriately located and made accessible to women of all communal and caste groups within a community. - Dalit women specifically are discriminated against in access to water. This may be a priority area of intervention.

6

- Specific measures are required to increase the attendance and retention of tribal girls in educational institutions. - In the planning of environmental interventions, the priorities of tribal women in relation to the natural resource base should be specifically examined and addressed. - Initiatives on violence against women will have to tackle sensitive issues of inter-caste and inter-communal tensions. Increasing gender-sensitivity in health programmes - Awareness raising about women’s, and specifically girls’, health problems is needed to promote early recognition and treatment of illness among girls. - Health delivery systems need to be designed to take account of limitations on women’s flexibility in timing and mobility in attending health facilities. - Family planning programmes need to promote a wider range of contraceptive methods and follow up care. Women are rarely independent decision-makers; it is vital to educate men regarding responsible family planning. - Gender-sensitive AIDS awareness and behavioural change requires decentralised peer group education, e.g. in existing women’s organisations in literacy programmes, and among groups of men, to complement mainstream public information campaigns. Addressing gender inequalities in education - Measures are needed to reduce, rather than accommodate, girls’ workload in order to improve their access to education. The provision of crèches addresses this problem but this needs to be co-ordinated with workplace and school locations and timings to make pre-school schemes more effective. - Non-formal education requires well conceived bridges into formal education provision to prevent ghettoization of disadvantaged groups, especially girls. - Areas of intervention in education which require further attention are: the ‘hidden curriculum’ and girls’ safety and vulnerability to harassment. - Attempts to increase the access and retention of girl students need to consider specific constraints to tribal, Dalit, and Muslim girls. Provision of single sex facilities, particularly at post-primary level, is important to increase the access of Muslim girls to education, especially in more conservative regions.

7

- Project-related research could help to identify what factors enable girls to remain in education and the relative impact of different factors on increasing female enrolment, retention and attainment. - Economic policies are needed to address gender bias in education. Improved employment opportunities are needed for educated women, especially in rural areas, to create incentives to female education, for example by relaxing recruitment criteria for women to public service positions. - Policies promoting female employment must counteract the possibility that women working outside the home may impact negatively on girls’ education. Promoting institutional change and support to NGOs _ - Training for women candidates and representatives in local government is needed for women from all communities, particularly Dalit, tribal and Muslim women. - Women’s effectiveness in local government could be strengthened by supporting increased links between local government representatives and local development NGOs; such alliances have proved effective in other countries in challenging vested interests. - Alongside investment in enhancing the capacity of NGOs for expanded service delivery where appropriate, there is a need to provide mechanisms for continued support to smaller NGOs, particularly those working with women and/or not directly engaged in service provision. - The resources and skills base of existing women-focused NGOs and networks needs to be drawn on to support the development of NGOs in areas of the country where they currently have little presence. - Gender training needs to be integrated into mainstream training programmes; incentives to institutions are needed to invest in training and follow-up. It should assist in supporting long-term links between institutions (e.g. between government and non-government; across sectors) and should aim to become self-sustaining by training trainers. - Gender training must be systematic, analytically coherent and address other axes of inequality (e.g. caste, class). It must support government (and non-government) personnel in promoting long-term change within their institutions, while recognising the constraints to this. Finally, gender training must address male needs without losing sight of the politics of gender. Rigorous impact assessments of gender training programmes are needed.
8

KEY ISSUES AND RECOMMENDATIONS This is a summary of the key findings of the more extensive and detailed Background Report on Gender Issues in India, with an emphasis on the policy and operational considerations arising from these findings. The analysis provided in the Background Report and summarised here, attempts to situate changes in gender relations and gender inequalities in the larger context and to give an idea of what broader changes will mean for different women in years to come. Issues of particular importance to policy and operations or specific recommendations are highlighted in the text. Since the various dimensions of gender inequality are interconnected, there is some overlap between sections. The main report is divided into six chapters, which are summarised here. The numbered headings follow the chapter sequence in the main report; sub-headings do not. Chapter One sets out the approach of the report and the context within which gender issues are being examined. Chapter Two examines the likely impact of the New Economic Policy from a gender perspective, considering gender divisions and differentials within the Indian economy. Chapter Three looks at how issues of caste and communalism interact with gender discrimination. Chapters Four and Five look in depth at gender biases in the provision of social sector services, specifically, health, water and sanitation and housing (Chapter Four) and education (Chapter Five). Finally, Chapter Six looks at the institutional context within which gender inequalities are being addressed, particularly government and NGOs, reviews the impact of various programmes on gender equality, and looks at the experience of and future possibilities for ‘mainstreaming’ gender in the Indian context. 1. INTRODUCTION: Contemporary contexts for the gender question Consideration of gender issues is now common currency in government, NGOs, the private sector and mainstream politics in India, at least at the rhetorical level. This awareness has come about as a result of the efforts of the women’s movement, and the influence of international development and feminist debates, as well as aid initiatives with a focus on gender. However, this increased awareness does not imply that gender issues are being dealt with on their own terms. The range of actors with a stake in promoting ‘women’s’ issues each have their own underlying agendas which influence the way in which gender questions are addressed. Moreover, analysis of gender issues cannot be separated from broader political and economic developments. Gender relations are being continuously reshaped by contemporary developments so that it is necessary to look at the complex ways in which gender relations are being recast today. A rapidly globalising economy, an intensely communal culture and a rising caste consciousness are major forces shaping India’s national development today. Events in the early 1990s - particularly the decision to implement the recommendations of the Mandal Commission on caste-based reservations in 1990; the announcement of a New Economic Policy in 1991; and the rise of communalism culminating in a major outbreak of communal
9

violence in 1992 - will have lasting impacts which are as yet difficult to discern. Moreover, these trends and events are not isolated but interconnected. It is in this context that issues of gender have to be considered, since these events and their effects will influence not only material gender inequalities, but also the construction of gender identities and the way in which questions of gender are framed and debated in years to come. One influential analysis of gender issues in India is that offered by Bennett in Gender and Poverty (1991), which characterises gender relations in India in terms of an ‘inside/ outside’ dichotomy, where the former represents the domestic and reproductive sphere to which women are largely confined and the latter the public domain of fields, markets, government institutions, seen as arenas of male power and control. Whilst this is a useful distinction, it may be limiting in a number of respects. Firstly, there are in fact relatively few women in India who are exclusively associated with the ‘inside’ sphere and their numbers are decreasing. Poor women are, on the whole, deeply enmeshed in the ‘outside’ sphere, on highly unfavourable terms. Moreover, women who do operate mainly from the ‘inside’ sphere are not necessarily worse off than those exposed to discrimination and harassment in the ‘outside’. Finally, men as a broad category are not in control of the ‘outside’; for men from social groups disadvantaged by poverty caste or minority status, the degree of their control over the outside needs to be qualified. Bennett’s analysis leads to the conclusion that the solution to gender discrimination is India lies mainly in the provision of inputs in order for women to realise their productive potential. A strategy is proposed of organising women into ‘demand groups’ in order to gain access to the necessary inputs. Here, it is argued that the provision of inputs alone may be a limited solution and under certain circumstances may simply increase women’s burdens and responsibilities. If greater resources are channelled to women, increasingly perceived as more efficient and socially responsible users of resources than men, the situation of poor men increasingly considered irresponsible - may be overlooked. A gender-aware strategy needs to go beyond simply directing more resources at women rather than men. There is also a need to support transformatory change in gender relations at individual and institutional levels. Further, whilst group formation and collective action will clearly form a key part of any strategy to counter gender discrimination, divisions and conflicts between different groups of women, which may provide obstacles to this strategy, also need to be recognised. The biggest single implication of the analysis offered here is that it is necessary to pay attention to the specific circumstances of different women, since caste, communal, class and other divisions mediate the way in which women suffer from gender discrimination. Moreover, recent events have encouraged relations of conflict between different groups of women particularly along caste and communal lines, with implications for development interventions with a participatory approach to working with women, or targeting women as beneficiaries.

10

2. THE ECONOMY, STRUCTURAL ADJUSTMENT AND GENDER For at least two reasons, the gendered impact of structural adjustment in India is difficult to assess. Firstly, India’s New Economic Policy only began in 1991 and thus its impact is as yet limited and uncertain. Secondly, any assessment of gender-differentiated impact of economic policy is complicated by the limitations of existing data, and the difficulty of relating apparently gender-neutral macro-economic categories and policies to their impact on genderdifferentiated labour markets and households at micro-level. In order to discuss the likely impact of adjustment by gender in India, an overview of where women are positioned relative to men in the economy is required. 2.1 Women and men in the Indian economy Overall adult labour force participation rates have remained stable over the 1970s and 1980s but this apparent stability masks considerable inter- and intra-sectoral shifts for both men and women. There has been an overall decline in the labour force participation of children in the 1970s and 1980s but this has been more pronounced for male children such that the rates of child labour force participation (as conventionally measured) are equalising. However, these do not include girls’ domestic labour responsibilities. According to official estimates, child labour still affects up to one quarter of rural children. In certain sub-sectors, such as factory employment, there is some evidence that girl child labour is increasing. Overall, women, as well as men, are concentrated in rural areas and the agricultural sector. Within agricultural employment, however, there are wide variations, both between regions and forms of employment. Women may be landless labourers or part of cultivating households; they may be working on family land or confined to work in the compound; they may also be engaged in non-agricultural work. Dalit and tribal women make up over half of female agricultural labourers. There is evidence of an increase in agricultural wage labour over the 1960s and 1970s, with growing numbers of non-Scheduled Caste or Scheduled Tribe workers. However, the proportion of agricultural employment in total rural employment is declining for both men and women, as the rural non-agricultural sector has been growing more rapidly than agriculture. Men, particularly, are moving into more diversified forms of nonagricultural employment in rural areas. Women are also increasing their employment in rural industries, but are overwhelmingly concentrated in household-based manufacturing (where they often serve as unpaid family labour) and in specific rural industries, such as dairying, fisheries, small animal husbandry, khadi cloth and village industries, handlooms, handicrafts, sericulture and social forestry. In manufacturing, there have been retrenchments in the 1980s in the private formal sector (textiles, pharmaceuticals, food processing) particularly affecting women, with a decline of 16 percent in female employment in private formal sector manufacturing from 1981-88.
11

Whereas in 1977-8, 15 percent of the female manufacturing labour force were employed in textiles, this percentage fell to 11 percent by 1987-8. There is also a growing informalization of manufacturing employment and a decline in household manufacturing, with a sharp drop of male employment in this sub-sector. Overall, women now form a greater share of the household manufacturing sector in particular (26.9 percent compared to 9.3 percent in nonhousehold manufacturing) and overall are more concentrated in informal sector manufacturing than in the 1970s. The major avenue of growth in employment in the formal sector has been in services, particularly in the public sector which has the highest proportion of service sector employment. In general, services sector employment has been the major growth area for both men and women in the 1980s. Trade and community services has registered particularly high growth; on the other hand, female employment in domestic service is declining. Overall, formal sector employment (which is mainly urban based) constitutes only 9.6 percent of employment and 5.6 percent of female employment (1991 data). Women form only 14 percent of the formal sector workforce but the benefits and rights of women workers in this sector have acted as a benchmark for women in the informal sector struggling for improved labour rights and conditions. By contrast, 94.4 percent of women are in informal sector occupations, including home-based workers, piece rate workers, casual labourers and petty traders. The informal sector is the dominant form of employment in most sectors, but particularly in agriculture (99 percent) and manufacturing (78 percent) and especially for women; in manufacturing, 88 percent of women are in informal sector jobs. In other areas such as transport and community services, the formal sector has a much greater presence. In general, there is a high level of insecurity in informal sector employment. Pay and employment conditions in the informal sector are poor relative to the formal sector and pay differentials by gender are greater. Moreover, women in this sector are also vulnerable to discrimination, harassment, sexual exploitation and violence; they also lack access to productive resources for self-employment. The share of services in GDP in the Indian economy has overtaken that of agriculture or industry and it is a sector of growing importance in terms of employment. Services constituted 82 percent of public sector employment and 29 percent of private sector employment in 1991. The presence of the formal sector in different service industries varies considerably. Service sector employment is mainly concentrated in urban areas where it provides around 50 percent of employment for both men and women. Urban services employment for women, in particular, has been steadily growing since the late 1970s (e.g. in retail trade, hotels); for males, less so, through from a higher base. Male employment in rural services (e.g. retail trade, transport) has been growing steadily over the same period, that of females less so.

12

2.2 Employment status and wages by gender Overall employment rates are much higher for men than for women and for women in rural than urban areas, when domestic work is not considered. When domestic work is included, the female employment rate overtakes that of males. Women also have much higher rates of subsidiary employment than men. Overall, women form a low proportion of those in regular salaried or waged employment compared to men in both rural and urban areas, but the gender gap in this regard is particularly pronounced in urban areas, where most regular waged work is located. Table 1 gives a breakdown of persons by work status, caste, gender and rural/urban residence.
Table 1 Distribution of persons by work status, caste group, gender, and rural/urban residence (NSS Data 43rd Round, 1987-88)
RURAL SCH. CASTES Male Female 19.0 9.9 5.4 1.5 0.5 0.5 26.9 15.7 51.8 27.5 53.8 35.8 0.2 14.2 0.4 17.7 1.3 1.4 OTHER CASTES Male Female 32.9 13.6 5.5 1.0 0.3 0.3 12.4 7.0 51.2 21.9 53.5 29.4 0.3 22.2 0.3 15.1 1.7 0.8 URBAN SCH. CASTES Male Female 13.6 4.9 18.9 4.7 0.3 0.1 14.2 6.9 48.1 16.6 49.2 21.3 0.5 28.8 0.1 10.4 3.5 1.1 OTHER CASTES Male Female 21.4 4.4 22.4 4.0 0.1 0.0 5.9 2.3 49.8 10.7 50.9 14.0 0.4 35.4 0.1 9.7 3.2 1.2

Self Employed Regular Waged/salaried Casual(P.Works) Casual (Other) TOT. EMPLOYED Tot.+Subsidiary Domestic (92) Domestic (93) Unemployed

Caste and communal differences interact with gender differences in terms of employment status. Dalit males and females are more concentrated in casual forms of employment than non-Scheduled Caste groups with approximately double the participation rates in casual labour of other groups in both rural and urban areas. Dalit females are less likely than other women to be exclusively engaged in domestic work. Overall, Muslim women have lower labour force participation rates than both other women and Muslim men. Muslims in general also have low rates of regular waged employment and high rates of self-employment in comparison to other groups. Economic support to Dalit groups - particularly women would need to be tailored to their heavy involvement in casual wage labour; similarly, for Muslim groups (including women), their high level of involvement in selfemployment should be recognised in interventions aimed at including this group among beneficiaries. Greater detail of the gender-differentiated conditions of wage labour and self-employment among Dalits and Muslims respectively would be of value in this regard. Conventional unemployment data are of limited value in the context of India, because the majority of the population cannot afford not to work. Also, under some circumstances, women are less likely than men to report unemployed even where they cannot find work. Nevertheless, unemployment rates tend to be higher for women than for men and higher in

13

urban than rural areas. Those with high school and college education also tend to have higher unemployment rates than less educated groups (who cannot afford to be ‘unemployed’). Among educated women, unemployment is higher in rural than urban areas. Underemployment is also a growing problem for both men and women. Investigation of the specific barriers to employment of female high school and college graduates would seem to be justified given the gender differential in unemployment rates and the questions it raises about the value of investing in female education. Gender discrimination in wages and differentials in earnings are widespread in India and particularly in the informal sector where equal pay legislation is not applied. Moreover, a large proportion of women and children working in the informal sector are doing so as unpaid family labour. In agriculture, wage rates vary widely across regions and by season; but rates for women, children and bonded labourers are particularly low. Female wages as a proportion of male wages range from around half to over three quarters depending on the state. There is no consistent trend in and little recent data on gender differentials in agricultural wages. In most industrial categories, women’s earnings in the informal sector were around half those of men. Even in the formal sector, however, inequalities in earnings are marked because women are concentrated at lower occupational levels. In regular waged employment, women’s earnings are on average 83 percent of men’s. 2.3 Gender and poverty India contains one of the largest concentrations of poor people in the world and thus poverty has been an areas of extensive debate, measurement and policy intervention. There remains considerable controversy over what measures of poverty and/ or methods of poverty assessment are most appropriate and this is reflected in differing data and assessments. In spite of all the attention to poverty, differential experiences of poverty according to other caste and community, and the interactions of these with gender, are relatively under-analyzed. Whilst women’s relationship to poverty is shaped by the wider context, there are also gender specific processes of impoverishment. Here, intra-household processes and the incidence of female headship are particularly considered. There is currently increased attention to the perceptions of the poor of their own poverty and increasing use of participatory methods of poverty assessment. One study of poor villagers in India found that reduced dependence on landlords, greater mobility, changes in consumption patterns and the opportunity to purchase consumer durables were valued by villagers themselves but these priorities may differ by gender and by residence, caste and community. It is important to understand more about what different groups of women themselves would consider to be an improvement in their lives, to inform the design of poverty alleviation programmes. Overall, poverty has been declining in India since the late 1970s but, according to one recent survey, more rapidly in rural than in urban areas. Whilst the absolute numbers of the rural poor are the largest, they are declining; by contrast, the absolute number of the urban poor has
14

increased in recent years. Whereas in the early 1970s the urban poor constituted around one sixth of the total poor, by the late 1980s they constituted more than one quarter of the total. However, the relative importance of rural and urban poverty varies considerably according to the state. Intra-state variations in poverty levels are also considerable. Poverty estimates based on expenditure class data show that in rural areas, tribals then Dalits are the poorest social group; in urban areas, Muslims are the poorest followed by Dalits then tribals. Women and girl children suffer from gender discrimination in the allocation of resources within the household, in spite of their considerable labour and often cash contributions. This discrimination is particularly marked in the allocation of food and health care resources, resulting in imbalances in the sex ratio for most states, discussed further in the health section. The relationship between household wealth and income and gender discrimination is not straightforward; there is some evidence that in the Indian context, gender discrimination within the household may be less in poor than well-off households. In situations of upward mobility, women often see less of the benefits than men and gender discrimination certainly does not disappear in better-off households; in fact it may intensify. In general, where women’s productive work is not visible, or where gender differentials in earnings are high, women may be particularly prone to discrimination in the household. There may also be socio-cultural factors involved, since Muslim women’s apparently lower participation rates do not seem to play out in terms of significantly lower sex ratios in aggregate, for example. In most households, women’s relationship to and uses of income are quite different from those of men. Although women frequently manage household budgets and consumption, they may have little direct control over income and often do not even know what husbands earn. When women do earn cash their income is often entirely absorbed in family expenditure; men, on the other hand, will tend to retain personal income for spending on luxuries (tea, alcohol, bidis), irrespective of poverty, consumption of which helps them maintain some sense of manhood in the face of inability to support their families; to be ‘good’ husbands and fathers. In this respect, poor men are increasingly seen as irresponsible and shiftless, particularly from a middle class perspective. The relationship between female headship and poverty is not clear and there is considerable debate over the definition of female headship. In India, female headed-households tend to be concentrated in the lowest expenditure classes and among cultivating households, they own smaller than average plots of land. They are also over-represented among casual labourers. At the same time the average size of female-headed households is smaller than households overall, in most cases due to the absence of a male spouse. Official estimates put the incidence of female headship at around ten percent, but the actual incidence may be considerable higher, possibly as high as 30 percent, especially in rural areas, where ruralurban migration of males may be a significant contributing factor. Variation across states in the incidence of female headship is considerable. In some accounts, the higher incidence of female headship in particular states (e.g. Kerala) relates to traditions of a more elevated status of such women, where matriliny has been prevalent. Other explanations link the phenomenon to high rates of male outmigration, including to work
15

overseas, which may indicate that such female-headed households are comparatively well off, at least in material terms. The proportion of female-headed households increases in older age groups and the majority are widowed, divorced or separated. The process through which female headship arises is important in terms of what forms of social support are available; this probably differs considerably across social groups or communities. It would be of interest, for example, to know whether female headship through widowhood remains a predominantly Hindu phenomenon (where Hindu widows are not expected to remarry) and how the social support and living arrangements for such women varies by caste group. In general, a more nuanced picture of the processes through which female headship arises and the implications of these for the extent of available social support, may assist in identifying particular sub-groups of female headed households who face extreme hardship. 2.4 Structural adjustment and gender: employment effects Structural adjustment programmes are explicitly aimed at inter and intra-sectoral reallocation of resources for greater efficiency as well as increased openness of the economy to bring the production structure in line with global comparative advantage. Shifts in production, consumption and employment patterns affect the incomes of various social groups differently, so that there are losers and gainers in the process. For example, regional disparities are likely to increase under adjustment, although there are conflicting pressures: low wages may attract investors to certain regions but in other cases, investors may tend to concentrate in areas with an existing economic infrastructure and labour skills. Some relatively developed urban centres, such as Bangalore, have been growing rapidly, particularly with the development of ‘high tech’ industries. Other backward regions, such as Orissa, may see limited gains, since they lack infrastructure and a developed skills base. Any gains here are likely to be concentrated in development of the natural resource base (logging, fisheries, mining and quarrying) for export, areas unlikely to provide new employment opportunities for women. Under structural adjustment, prospects for women’s employment in the formal sector are uncertain and on balance not very good. Private sector manufacturing industries shed considerable female labour in the 1980s, partly on the basis that employers considered women to be less flexible (in terms of hours available) and higher cost (in terms of e.g. sickness absences) employees than men. However, these may be justifications for shedding labour rather than a real assessment of the relative costs to employers. Given the less positive overall economic prospects for the Indian economy over the 1990s compared to the 1980s (in terms of growth and on indicators of capital investment for example), it is unlikely that there will be a major expansion of female employment (except perhaps, in specific sub-sectors) and women workers may be increasingly vulnerable to retrenchment with fluctuating market demand.

16

In the public sector, there have not been significant retrenchments to date of either men or women; labour force rationalisation has relied more on voluntary retirement and on a freeze in new recruitment. Whereas in the 1980s, the public sector and specifically service occupations in this sector represented a major source of growth in employment for women, due to stagnation in public sector employment under adjustment, it is unlikely that this trend will continue in the 1990s. There is some evidence on trends in specific areas of public services employing large numbers of women, e.g. nursing, where vacant posts are no longer being filled. Export-led development is a key feature of structural adjustment programmes. One area of potential growth in employment for women in other adjusting countries has been in new export-oriented manufacturing industries, such as electronics, as well as more established export industries such as textiles. In India, there are a number of factors which may limit this potential source of new employment for women. Firstly, much of the recent growth in exports has been concentrated in sectors such as gems and jewellery (especially diamonds), plastics and steel, which are not major employers of women. Even in export processing zones, which, in other parts of the world, employ a high percentage of women workers (70 percent plus in many cases), the proportion of women workers in export processing zones in India is relatively low (under 50 percent in most instances) and may be declining. In the export garment industry, the proportion of women employees was only 20-30 percent in 1985. Both the long-term potential of export industries currently in growth and the quantity, quality and impact of the employment they are providing for a small number of women, are questionable in this case. Nevertheless, employment in manufacturing export industries has brought into the manufacturing labour force a section of lower middle-class women, with moderate education, in semi-skilled jobs. Previously, employment for women was mainly concentrated either among poor illiterate women who had no option to work in unskilled occupations, or among middle-class educated women in professional and white collar jobs. A number of studies have been done of female employment in export industries generally and in export processing zones specifically (see Box 1). However, they tend to lack systematic comparison with male employment patterns, with alternative forms of employment for women and assessment of trends over time. Broadly, such studies have found that wage rates for women in export industries are quite variable, partly depending on whether they are employed in large-scale or informal sector units. Female employees in manufacturing export industries, as elsewhere, tend to be young and unmarried with at least primary education. Their employment prospects may be fairly short term; earnings are often used to support brothers’ education, or to pay for their dowry; thus, even for those women who do have access to such opportunities, the longer-term benefits are not clear in terms of improvement of their status.

17

Box 1: Women’s Employment in Export Industries in India
In 1985, the electronics zone in Bombay employed 5500 men and 1500 women (21 percent women). The women, between 17-25 years and largely unmarried, earned less than Rs. 500 per month, saw hardly any wage increases, and were overwhelmingly concentrated in monotonous and tedious assembly work in contrast to male workers who were predominantly in software preparation. Their earnings often went towards seeing a brother through college, helping to tide over the unemployment of the main breadwinner, or towards their own dowries (Sharma and Sengupta, 1984). There are some indications that the proportion of women employed in the zone has declined since then (Shah et al, 1994). Wage rates for women employed in electronics units in other parts of Bombay showed much greater variation - while smaller units paid less than Rs.500, a substantial proportion of women in larger units earned about Rs.750-1000 (Pore, 1991). In the Kandla Free Trade Zone in Gujarat, the share of women employed in 1985 was much higher (3400 females to 4750 males, or 42 percent), and they were distributed across ready made garment units, electronics, pharmaceutical, and brush and bristle making units. They were largely below 20 years and unmarried, with a primary education, and earning between Rs.300-600. Since such women were also involved in household work, their work day often began at 4 a.m. and did not end until 10 p.m.. Unlicensed contractors recruit women from rural areas in Gujarat by bussing them to and from work. (Visaria and Minhas, 1991) The export garment industry encompasses factory units and smaller workshops, to all kinds of subcontracting and home-working arrangements. Smaller units are frequently closing down and relocating themselves to cope with fluctuating demand. The proportion of female employment in garments in 1985 was between 20 and 30 percent, i.e., unlike other countries, this industry is not female dominated. As in other export industries, women workers in the garment industry tend to be young and largely unmarried, with level of education and caste being a determining factor in terms of whether they get factory or informal jobs. Interestingly, women are more concentrated in the larger units that use assembly line production methods, although often as temporary or casual labour, earning less on average than men. In Delhi, Madras and Calcutta, there is a clear hierarchy of jobs by gender but this was less the case in Bombay. (Krishnaraj, 1985; Banerjee, 1991; Kalpagam, 1994; Rao and Hussain, 1987.) As Krishnaraj (1985: 19) points out,unlike South-East Asia where women are "cheaper than machines", here both men and women are cheaper, except that in some sectors women can be used as reserve labour much more easily.

In specific sectors, e.g. textiles, competition from exports following liberalization, and from larger units set up to take advantage of export opportunities, is likely to have a negative impact on the handloom sector of textiles, where women’s jobs are concentrated. Often, women are concentrated in smaller, lower technology units and as scaling up of production processes occurs, the proportion of male employment tends to increase. Most urban women work in the informal sector. Moreover, trends in the 1980s show a growing informalization of manufacturing industry, with increasing use of sub-contracting and home-working arrangements. Evidence is limited, but broadly speaking, as mentioned

18

above, wages and employment conditions are relatively poor in the informal sector, particularly for women and gender differentials in pay are greater than in the formal sector. Whilst in other contexts, there has been much discussion of feminisation of the labour force, and specifically of the informal sector labour force, it is not clear that this is the case in India, in the specific sense of women being employed in occupations previously dominated by men. Women in informal sector manufacturing remain concentrated in very specific industries such as food processing, garments, embroidery and bidi and agarbatti rolling; their employment opportunities do not seem to have greatly diversified. At the same time, increased crowding of informal sector labour markets and falling demand under stabilisation measures may mean declining incomes for those working in the sector; the net outcome is difficult to determine however, because of possible substitution as well as income effects on demand and because of the heterogeneous nature of informal sector activity. In other contexts, it has been found that women tend to be concentrated in the non-tradable petty services end of informal sector activity and are thus very reliant on demand from low income groups who may suffer most from income declines under adjustment. In agriculture and the rural economy, there is considerable debate and disagreement about the likely effects of structural adjustment in India, including among feminists. Agriculture is key because of its major role in agricultural employment (especially of women) and in providing for basic consumption needs. Policies in agriculture include: removing price controls; removing agricultural subsidies; a withdrawal of the state from the provision of rural infrastructure, and encouragement to private sector operators in this area. A particular feature of rural development policy in India is that expansion of special employment schemes under structural adjustment, as part of the social safety net arrangements. Some see these policies as reversing decades of bias against the agricultural sector, providing the potential for increased agricultural growth and thus increased incomes not only for farming households but also for the landless or land-hungry poor, including women, through more and/ or better remunerated agricultural labour opportunities. Others are more doubtful, seeing the likely response of agricultural output to price increases as limited because of the reliance on rain-fed agriculture and also due to declining investment in rural infrastructure and services, such as irrigation, communications etc., which may be necessary for a significant supply response. In this view, the new agricultural policies are more likely to favour large-scale commercial agriculture, providing limited employment opportunities for women. These are both are rather crude positions. More detailed analyses of the impact of adjustment on agricultural employment are required, relating farming systems and cropping patterns in particular regions (including distribution of landholdings and extent of wage labour), the extent and pattern of involvement of women (and men) in agricultural labour, to the likely impact of specific price and other policy changes. As part of the New Economic Policy, there is a strong focus on the development of rural industries, in recognition of the need to diversify rural employment and as a means of creating labour intensive employment in the rural population. This policy is seen as specifically benefiting women who form a significant percentage of rural industry employees in India, particularly in certain sub-sectors. Productivity in rural industries is to be raised through the
19

development and dissemination of new technologies. At the same time, it is recognised that in the short term, liberalization may lead to significant job losses in this sector. No specific measures to protect or improve women’s employment in rural industries are evident in this policy and a joint household production model tends to be assumed. There is also considerable emphasis on poverty alleviation schemes under adjustment, as a social safety net. The experiences of and lessons from these are summarised in Chapter Six (see below). Key lessons are: the participation of women in poverty alleviation schemes needs to be improved both quantitatively and qualitatively, as well as urban coverage; more attention is also needed to asset creation, rather than simply short term employment; in self-employment programmes, there is a need to go beyond credit provision to provide complementary packages of support and market access. 2.5 Structural adjustment and gender: income and consumption effects Beyond direct employment effects, stabilisation and structural adjustment will also have broader effects on incomes and consumption. Overall, it is probable that real wages and incomes will decline under structural adjustment, particularly for casual workers and the selfemployed poor, due to demand restraint, slowed growth, and shifts in the prices structure, especially in urban areas. Dalits, tribals and Muslims, including women, may be overrepresented in these groups. As managers of household consumption, there are specific implications for women of declining real incomes, in terms of the need for ‘expenditure saving’ activities and/ or pressures to earn additional income. Moreover, intra-household inequalities in resource allocation will mediate the income and consumption effects of adjustment by gender. Poverty alleviation efforts under adjustment need to address the specific disadvantage of Dalit, tribal and Muslim groups in their design, as well as monitoring intra-household resource allocation processes among these different groups. There is a shift in policy away from large-scale subsidisation of consumption through control of food grain prices and the public distribution system, towards increasing the purchasing power of the poor through special employment schemes and ultimately improved employment opportunities under adjustment. Concerns have been raised, however, about the impact of liberalisation and agricultural price rises on the rural poor, of whom 40 percent are reliant on casual and seasonal wage labour and who are therefore particularly vulnerable to increases in grain prices and to fluctuations in agricultural prices in general. Whereas in the 1980s, food grain prices rose more slowly than agricultural prices as a whole, so that real agricultural wage rates were rising, this trend seems to have reversed in the 1990s, implying possible declines in the real agricultural wage rate. In some views, the famine prevention capacity which India has built up over the last two to three decades through its food security arrangements and public distribution system, may be jeopardized by liberalised agricultural trade. Terms of trade appear to be moving against urban areas, under adjustment, where safety net measures including special employment schemes are at present thinly spread and function
20

poorly. Poor urban households are particularly vulnerable to price increases because of their total dependence on market purchases, whereas sections of the rural poor at least have some recourse to subsistence production. Whilst to date the public distribution system may have been biased towards urban consumers, within the urban population, relatively well-to-do households have benefited proportionally more from this system. Urban poverty, according to recent estimates, is increasing in absolute numbers and certainly declining at a slower rate than rural poverty. Retrenchments in certain industries, increased labour crowding in the informal sector and declining real incomes are likely to concentrate ‘new’ poverty associated with structural adjustment in urban areas. In the longer term, there may even be a reversal of rural to urban migration trends. Under structural adjustment, then, there may be a particular case for increased interventions towards the urban poor and especially poor women. Project and programme level monitoring of the impact of structural adjustment on the livelihoods of the urban poor could assist in informing the design of such interventions. There is also a need for measures to improve the targeting of the public distribution system to both the rural and urban poor. 3. CASTE, COMMUNALISM AND GENDER Caste and communal issues have not featured prominently either in analyses of gender relations in India, nor, more generally, in considerations informing the design of aid programmes and projects. This is partly, perhaps, because of the dominance of upper caste and/ or middle-class groups in national level political life generally, including within the women’s movement. The ‘invisibility’ of caste in the post-Independence secular politics of India, at least from an upper caste perspective, has, until recently, meant that the complex ways in which caste and gender interact have been largely ignored. Also, perhaps, much in the way that external actors (particularly men) are often reluctant to tackle gender inequalities, since these are often consigned to the domain of ‘culture’ or ‘tradition’, caste is perhaps seen in a similar light, as a peculiarly Indian tradition which should be left well alone. The communalization of Indian political life over recent years also has specific gender dimensions: as with caste issues, women are often singled out for attention or attack as bearers of the respective communal identities. On the other hand, some women have a stake in advancing communal projects, or in maintaining caste divisions. Relations of gender inequality in India, as well as, importantly, differences between women in India, cannot be understood without analysing how they interact with caste relations. Moreover, in terms of development policy and practice, there are particular implications arising from an incorporation of caste and communal issues into a gender analysis. In general: Given the paucity of data, particularly gender-disaggregated data, on caste (and communal) issues, interventions should be preceded by both qualitative and quantitative analysis of the target or participating populations, to ensure that resources are not being concentrated among already privileged groups. This is particularly important where there is regional and spatial separation of different caste (or communal) groups.
21

Where participatory approaches to development are being promoted, it is important to understand the dynamics of inter-caste and inter-communal relations in a given context, particularly where there is a local history of conflict. It must not, conversely, be assumed that such a history of conflict is always present, nor that different communal or caste groups are unable to work together. Certain social and religious groups may dominate local political structures and the voices of lower caste, non-Hindu, Dalit and tribal groups, and particularly women within these groups, are likely, in many instances, to be under-represented. Specific mechanisms may therefore be needed to elicit the opinions of under-represented groups. 3.1 Everyday caste in India Caste is a system of hierarchical social relations peculiar to Indian society (although with important parallels elsewhere), with economic, social and ideological dimensions. Embedded in caste ideology are concepts of purity and pollution which restrict contact and intermarriage between different caste groups and which exclude lower caste groups (and particularly Dalits) from access to particular occupations, economic and cultural resources and certain social spaces. Caste is not, as some academic studies suggest, a traditional, village-based feature of India life, vulnerable to erosion under the impact of ‘modernising’ trends. Indeed, the argument here is that caste is very much a contemporary issue, through which new forms of inequality, power and conflict are being created, as much in urban as in rural areas. Because control over women’s sexuality is vital to the maintenance of caste identity, women are caught up in caste relations and conflict in very particular ways. Recent studies of the practice of untouchability, for example, find it highly prevalent in relation to traditional spaces such as temples and wells, but it is also evident in ‘modern’ spaces such as the cinema and schools. Moreover, it is not just a rural phenomenon. For untouchable (or Dalit) women, untouchability may prevent their access to vital resources for daily life (e.g. water) and confine them to separate living spaces, but on the other hand offers them little protection from harassment and attack of upper caste men. In planning investments in social infrastructure, such as schools, medical facilities, water supply etc., questions of the location of facilities and their accessibility to all members of a community must be addressed with specific reference to caste (and communal) issues. Studies have shown that Scheduled Castes/ Scheduled Tribes suffer a higher incidence of poverty than the remainder of the population. Studies of intra-household inequality in these communities are lacking. Nevertheless, it is likely that lower caste, and particularly Dalit, women are over-represented among the poor. There is a case for the consideration of
22

specific interventions and investments to counter the concentration of poverty among Dalit and tribal groups, particularly women within these groups, as well as, possibly, Muslim communities, especially in urban areas. Further investigation of gender relations among Dalit and tribal communities, and particularly the intra-household dynamics of such communities, would be valuable in order to inform such interventions. Because caste is not a purely economic phenomenon, economic advancement alone will not cancel out patterns of caste discrimination. On the other hand, power relations between caste groups may be affected by the cultural and political advancement of lower caste groups (e.g. through education), even in the absence of material advancement. Given limited quantitative data, and the complex nature of caste relations (which are different from, though correlated to, issues of class and income), a qualitative understanding of how contemporary caste relations are evolving in particular contexts is required, as well as improved quantitative data. 3.2 The emergence of caste as a national political issue Since Independence, when a new ‘caste-blind’ Constitution was established, there has been little attention to caste as a national level political issue. Reservations for Scheduled Castes and Scheduled Tribes (representing 15 percent and 7.5 percent of the population respectively), introduced to give these ‘disadvantaged’ groups proportional access to government employment and education, is the main policy through which caste has been addressed. By 1981, 18.7 percent of all Government jobs were held by Scheduled Castes/Scheduled Tribes, slightly lower than their overall representation in the population (22.5 percent), but nevertheless representing considerable progress. It is not known what proportion of reserved positions are held by Scheduled Caste/ Scheduled Tribe women. Indications suggest that to the extent that reservations have promoted employment for Scheduled Castes and Scheduled Tribes, men have gained more than women. From the perspective of Indian upper caste Hindus, who have dominated national political life, caste has until recently been an ‘invisible’ or private issue. Official statistics on caste issues are limited and not well disaggregated. Scheduled Castes and Scheduled Tribes are grouped together in government data collection; other caste groups did not, until recently, have data collected on them. Importantly for this study, gender-wise disaggregation of data by caste is largely absent, rendering difficult assessments of the relative position of low caste women, either in relation to higher caste women or men within the same category. It is only recently that caste issues have come to prominence in national political life, in particular following the announcement in 1990 that the government was going to implement the recommendations of the 1981 Mandal Commission. This Commission had recommended additional reservations for ‘Other Backward Castes’ (OBCs) constituting 52 percent of the population (although the reservation percentage was limited to 27 percent since total reservations cannot exceed 50 percent). In 1981, the representation of OBCs in government jobs was only 12.6 percent (and less than five percent in Class I jobs) compared to their
23

overall proportion of over half the population. Since data are not disaggregated for upper caste groups (who along with Muslims and Christians make up the ‘residual’ category which is neither Scheduled Caste/Scheduled Tribe nor OBC), their exact representation in government jobs is difficult to gauge, but is, in any case, clearly disproportionate to their representation in the population and increases at higher levels of the system. The decision by Government to implement the Mandal recommendations, sparked off nationwide protests, primarily among upper caste Hindus (including women) who perceived it as a threat to their privileged access to government positions (either directly or through marriage) as well as a challenge to the dominant ‘caste-less’ ideology which had prevailed since Independence. 3.3 Caste and gender The connection between caste and gender is most evident in the differential control men exercise over women’s identity, labour and sexuality. Whilst upper caste women have gained access to privileged employment and other resources through their caste status, lower caste women are generally confined to work in the non-formal sector. In as much as reservations have promoted lower caste employment, this had tended to favour men. On the other hand, the sexuality of upper caste women is highly controlled through marriage, as the means of ensuring their caste status. When marriage is being considered for upper caste women (and men), caste considerations - which may otherwise feature little in daily life - becomes prominent. Whilst upper caste practices (such as dowry) have spread to some extent among lower caste and tribal groups, their sexuality may be less controlled and women from these groups may have greater autonomy though in conditions of greater poverty. Lower caste women are nevertheless highly vulnerable to sexual harassment and attack, particularly from upper caste men. Dalit women have arguably suffered most from physical and sexual violence against women in India, with little protest or protection offered either by the state or other sections of society. In a situation of increasingly public inter-caste tension and conflict, women are being singled out for attack as the ‘bearers’ of their respective caste identities. Lower caste women, and Dalit women, in particular, are attacked and abused with impunity. Upper caste women may also be the objects of threats, intimidation and attack, but mere suspicion of attack on upper caste women by lower caste men can lead to brutal reprisals (see Box 2). Interventions to address violence against women, including sexual violence and harassment, need to be informed by this wider canvass of inter-caste and inter-communal tensions.

24

Box 2: Caste conflict and gender in Chunduru
In Chunduru village, Andhra Pradesh, a Dalit student called Ravi, who was attending a cinema show, is said to have accidentally touched a Reddy (forward caste) boy sitting in front of him with his foot. (The Reddy castes are one of the major dominant castes in rural Andhra.) Though he promptly apologised, the Reddy boys present attacked him, upon which Ravi fought back with the help of his friends. Sensing trouble, Ravi was sent away from the village by his parents. When he finally returned, he was caught by Reddys, severely beaten up, forced to swallow liquor and taken to the police station where he was accused of having sexually harassed an upper caste woman in a drunken state. In a parallel incident, another Dalit boy was attacked with a knife by a Reddy for having ostensibly brushed against two Reddy girls outside the cinema hall, which, however, he denied. When other Dalits showed signs of solidarity with the families of the Dalit boys, a social boycott was set in motion against the Dalit community for almost a month. This meant suffering from power and water shortages in their homes, and having to go outside the village in search of agricultural work or to buy provisions. The tension culminated in the gruesome murder of 19 Dalits, who were hacked to pieces in the fields by Reddy men, while the police did nothing. The importance of this incident lies in its prehistory, which is one of changing power relations between the Reddys and the Dalits in that area. The Dalits of Chunduru had, over the years, managed some upward mobility and perhaps even a degree of autonomy with better than average wages in a high productivity area, through education, by obtaining government jobs with the help of reservation quotas and via access to Christian institutions as Dalit converts. Thus, even though the Reddys controlled most of the land (with holdings ranging from 2.5 to 14 acres of land, with no Dalit owning more than 2.5 acres), the Dalit community boasted a number of graduates and postgraduates, whereas no Reddy had gone beyond high school. The incidents at the cinema hall symbolised a larger and longer sensibility amongst the Reddys that the Dalits no longer knew their place. Especially crucial is that Reddys fomented their anger by focussing on the real and fantasied access Dalit men now had to Reddy women. While none of the allegedly harassed women came forward to file cases, it is reported that, after the massacre, three hundred upper caste women took out a procession proclaiming that their ‘modesty had been outraged’ by Dalit men, thus retrospectively providing justification. (Kannabiran and Kannabiran, 1991).

As caste has emerged as a national political issue, feminists and the wider academic community have belatedly begun to realise that caste issues and the questions of power they encompass cannot be separated from an analysis of unequal gender relations. Until very recently, most of those attempting to improve women’s status, or tackle gender oppression ranging from state departments to women’s organisations - had been influenced by the dominant caste-less ideology. To date, the women’s movement has been dominated by middle-class and/ or upper caste women and tended to serve their interests, through strategies

25

of e.g. legal reform. Recently, within the women’s movement, Dalit (and to a lesser extent tribal) women are beginning to assert their own voice. There have been calls for the establishment of a federation of Dalit women, separate from both existing Dalit political organisations and the women’s movement as presently constituted. The priorities of tribal women are increasingly being taken up by women’s organisations and women in NGOs, such as lack of representation in local institutions, the effects of alcohol abuse, gambling and sexual exploitation of tribal women by non-tribals. In future, aid organisations aiming to work with women and women’s groups will be faced with a more diverse and fragmented, but more representative, women’s movement. 3.4 Tribals and gender Tribals make up 7.7 percent of India’s population, of whom 60 percent are concentrated in Central and Northern states (Madhya Pradesh, Orissa, Bihar, West Bengal, Maharashtra) and are overwhelming concentrated in rural areas (94 percent). There is a general lack of information on gender issues in the tribal context beyond the anthropological literature which focuses on kinship and marriage systems. Stereotypes of tribal women as powerful members of egalitarian societies, as loose women or witches, or as the most oppressed among women in Indian society, vie with each other in existing literature, but none are very illuminating about the nature of gender relations among tribal communities, nor about how these relations are being shaped by wider processes of social change. Processes of hierarchization are evident among some ‘tribal’ communities where societies are taking on some features of the caste system. This may imply less autonomy for women among those groups taking up caste practices such as dowry. However, little is known about intra-household relations in tribal communities, particularly about household decision-making patterns. Further research relating changing marriage patterns, household formation, decision making and related practices in tribal communities and their impact on gender relations would be informative in this context. Whilst often characterised as isolated, egalitarian communities, tribal communities are increasingly incorporated into the monetary economy, though migration, indebtedness, bonded labour etc. Women as well as men are caught in these relationships and often suffer further discrimination, e.g. in agricultural wage rates. Tribals are increasingly fragmented and thus collective identity or homogeneity among tribal communities cannot be assumed. Given this, and the trends noted above, in planning interventions in tribal contexts, it is important to take into account local categories and identities rather than externally imposed ones. The access of tribal communities (and particularly women) to the natural resource base is increasingly threatened with erosion of land rights, large scale development projects and, even, environmental initiatives. The identification of tribal women’s interests with environmental concerns is often romanticised and misplaced. The priorities of tribal
26

women in relation to the natural resource base must be considered in the planning of environmental interventions. The categorisation of tribals along with Scheduled Castes (Dalits) in official statistics and as a target group for interventions has led to distinct issues in relation to tribal communities being overlooked. There has been a tendency for non-tribals to capture benefits of programmes aimed at both Scheduled Castes and Scheduled Tribes. Tribals, at least in some states, appear to suffer even greater poverty than Dalit groups. In as much as reservations have benefited these groups, men appear to have benefited more. Tribals, and tribal women particularly, may require separate anti-poverty interventions rather than programmes targeted at Scheduled Castes/ Scheduled Tribes together. Access of tribal women to education and their retention within the system is particularly poor (see Chapter Five on education). Specific measures are required (e.g. campaigns, incentives to compensate for loss of household labour) to increase attendance of tribal girls in educational institutions; but it would also be important in this context to assess the perceived relevance of formal education to tribal communities and particularly young tribal women. 3.5 Gender and communalism Majority Hindus account for 86.5 percent of the Indian population, compared to 11.4 percent Muslims, 2.4 percent Christians, 2 percent Sikhs and 1 percent other religious groups. The primary current axis of communal tension is between Hindus and Muslims, although other axes of communal conflict also exist and have been more prominent in other periods. Muslims tend to be more concentrated in urban areas (over half in some states) than other groups and in particular states (e.g. West Bengal, Kerala). An understanding of the construction of gender in relation to communal identities is vital to inform analysis of the contemporary manifestations of communalism, as well as for practical development work to benefit women from the various communities. As with caste issues, women tend to be the bearers of communal identities, but they may also have a stake in advancing communal projects. Communalism has posed a particular problems for the women’s movement in India, where, in the name of progressive, particularly legal, reforms, there is often an implicit (and unconscious) communal agenda of Hindu cultural superiority being promoted in the name of women’s interests. The women’s movement itself, partly in reaction to Western feminism, has become imbued with Hindu symbolism. There is a tendency among middle-class (and/ or upper caste) Hindu women, who have some economic and social room for manoeuvre, to perceive the patriarchal systems governing Muslim women as more oppressive than those of Hindu society and, thus, to promote legal reforms on the basis of Hindu strictures. Whilst ‘patriarchal’ practices in the Muslim community have been the subject of much protest and of

27

legal reform by the state, Hindu practices associated with Hindutva (such as the re-emergence of sati in Rajasthan) have proved less easy for women’s organisations and others to combat.

Box 3: The Shah Bano Controversy
In 1985, the Supreme Court ruled in favour of Shah Bano, a divorced poor Muslim woman, who had filed for maintenance from her husband under the Criminal Procedure Code. The case created a major controversy in different parts of the country and represents a turning point in India’s history. Since this was not the first time such an order had been passed, the negative reactions took everyone, the Government included, by surprise. Statements about the specific backwardness of Muslim personal law were made at the time of passing judgment, along with references to the need to formulate a uniform civil code according to the aims of the Constitution. (Mann, 1994) This was perceived as a threat to the Muslim community at large, especially when they were feeling the brunt of a rising Hindu right-wing in other areas. Not only Muslim religious leaders, but also groups of Muslim women themselves, protested against the judgment, demanding that Shah Bano be bound by Islamic Sharia’t law rather than an intrusive and ‘external’ criminal code. Those who fought in favour of Shah Bano - other Muslim women and men, mainstream women’s organisations, liberals and so on - found themselves in direct alliance with Hindu fundamentalists, who spared no effort in their support of Shah Bano’s plight. Far from upholding women’s rights for their own sake, upper caste Hindu men were, in effect, setting themselves up as the protectors and saviours of Muslim women from Muslim men. (Pathak and Sunderrajan, 1989). Shah Bano herself came under increasing pressure to withdraw her case, especially from Muslim fundamentalist leaders. In the end she did, since the only alternative would have been to suffer social ostracism. The party in power at the centre, the Congress, was meanwhile in great fear of losing the Muslim vote as the result of adverse reaction to the judgment. When the Muslim Personal Law Board pushed for a new Bill in Parliament that would reaffirm the privileged place of Muslim Personal Law and set aside secular law in these matters, the Congress Party complied. It went so far as to amend the Constitution in order to pass the Bill, which became the Muslim Women’s Act of 1986 (Protection of Rights of Divorce). Now the husband is only legally obliged to return the divorced wife’s mehr or dower and maintain her for the three months of iddat, the period just following divorce. If she is subsequently unable to maintain herself, the responsibility falls on her children, parents or relatives, or, failing this, on the Muslim Welfare Board (the State Wakf Boards).

Events of the mid-late 1980s saw communal issues affecting women coming to the fore. The Shah Bano controversy (see Box 3) highlighted the way in which both state and religious groups could manipulate gender issues to serve other agendas. The explosion of communal politics onto the national scene in the form of unprecedented inter-communal violence, occurred following the destruction of the Ayodhya Mosque in late 1992. Muslims were the main victims of recent communal violence, and specifically Muslim women, although Hindu

28

women were also attacked. It is argued here that communalism (Hindu-Muslim) is to date, a largely urban phenomenon, since Muslims are concentrated in urban areas. Communalism acts as a unifying ideology, attempting to downplay caste differences among Hindus which have become heightened in recent years. It is also partly fuelled by economic uncertainty and the erosion of middle-class privilege. In the 1990s, communalism has provided ‘empowering’ opportunities for some women, whereby they have attained prominent public positions, for example in women’s wings of Hindu parties. Religion has become a public issue, allowing these women scope for public activity, although this is still bounded by patriarchal attitudes within Hindu political organisations. Far from being inherently conservative, some women leaders in Hindu parties have common positions with the secular women’s movement (e.g. on sati, dowry etc). In the 1990s, communal identities appear to be important to some middle-class and upper caste women as a means of dealing with economic and social uncertainties but also as a form of resistance against continued patriarchal practices (e.g. dowry). Contrary to popular belief that Muslims have benefited from undeserved privileges, data show that they are under-represented in government employment and over-represented among both the rural and, particularly, the urban poor. The upsurge of anti-Islamic sentiment in the 1990s has led to the strengthening of Islamic fundamentalism and the consequent suppression of debate and diversity of opinion within Muslim society. This has left Muslim women, in particular, with very little room for manoeuvre. It is increasingly difficult for Muslim women to take public action in their gender interests, without becoming caught in inter-communal conflicts. Minority religious groups, as well as lower castes and tribals, may be inclined to distrust existing political structures, parties and movements, given their history of manipulation by such organisations and thus be reluctant to engage with them. This may be equally true of the women’s movement, from the perspective of Muslim, Dalit and tribal women particularly.

29

4. GENDER BIASES IN SOCIAL SERVICES: Health and family planning, water and sanitation and housing 4.1 Social sector policy, funding and biases Social sector policies, institutions and delivery systems are influenced by historical legacies as well as by the interests of the various interest groups who influence state resource allocation. Current ideology about the nature and causes of poverty, about the role of the state in basic needs provision as well as prevailing views of gender roles and relations, also shape social services provision. Unless carefully conceived, social services provision can act to reinforce and reproduce gender biases. Moreover, not all women are equally served by social sector provision; biases also prevail according to region, location, class, caste and community. This chapter and the following one on education provide ample evidence of such biases in the Indian context. To date, social sector spending has been characterised by certain biases. Firstly, government subsidies have been concentrated in economic rather than social services, tending to benefit private entrepreneurs and better-off farmers. Secondly, the share of state allocations to health and education in Plan outlays have been declining steadily, except for a recent (since 1985) rise in the share of education expenditure (see Table 2). The share of health allocations continue to decline. Thirdly, social sector spending has been biased towards urban areas and towards higher level services. As a result of cumulative biases, social sector subsidies are mainly captured by relatively high income groups. For example, only 20 percent of health subsidies and about 35 percent of water, sanitation and housing subsidies reach the rural sector, where two thirds or more of the Indian population lives.
Table 2: Percentage share of health and education in Five Year Plan outlays (Plans I-VIII, 1951-1997)
PLAN AND PERIOD HEALTH* EDUCATION First Five Year Plan, 1951-6 3.32 7.86 Second Five Year Plan, 1956-61 3.01 5.83 Third Five Year Plan, 1961-66 2.63 6.87 Annual Plans, 1966-69 2.11 4.60 Fourth Five Year Plan, 1969-1974 2.12 4.90 Fifth Five Year Plan, 1974-1979 1.92 3,27 Sixth Five Year Plan, 1980-5 1.86 2.70 Seventh Five Year Plan, 1986-1991 1.88 3.70 Eighth Five Year Plan, 1992-7 1.75 4.52 * N.B. Figures for health outlays do not include allocations for family planning.

The gender implications of these biases in funding have not been fully analyzed. However, it is clear that biases towards higher level services are likely to favour men, who, for example, form a progressively higher proportion of students at higher levels of the education system. Moreover, the lack of provision in areas such as health, pre-school, and lower levels of education, and water and sanitation, particularly in rural areas, increases the work burden of

30

women directly, reducing their possibilities for productive employment and/ or taking a toll on their health in terms of reduced leisure and increased workload. There is increasing emphasis on NGO and community participation in social sector service delivery, and women are expected to play a key role within community based provision, often as unpaid volunteers. However, women’s participation will be constrained by their existing work burden, unless there are clear incentives and economic or social benefits to women’s participation built into community-based service provision. Both the Indian Government and the World Bank claim a shift in the nature of aid and state spending from a concentration on investment in infrastructure towards investment in human resources development. Increased allocation to basic social services provision is seen as one means to improve the productivity of the poor. Recent analyses suggest that, although a slight increase in social sector allocations has taken place at the centre, from 3.4 to 4 percent of GDP, at state level (where 80-90 percent of social sector spending comes from) there has been a decline in real terms. Overall, spending on social services has declined from 16.5 percent of centre and state levels expenditures combined in 1990/1 to 15.4 percent of these expenditures in 1993-4. Education may be benefiting from increased resources, but possibly at the expense of health. Even in education, government may be trading off quality against quantity. In recognition of the potentially negative social impact of adjustment policies (see Chapter Two), in 1992, the World Bank proposed a ‘Social Safety Net Sector Adjustment Programme’. In other contexts, compensatory programmes implemented following adjustment have been limited in impact both in general terms - being ad hoc, piecemeal, politically motivated and lacking integration with existing structures and programmes - and also in terms of their gender impact, since they have tended to favour newly redundant formal sector employees who are predominantly male. The relatively early proposal for a social safety net programme in India and its comparatively large scale (although still very small in relation to the scale of poverty in India) give some cause for optimism that the lessons from the limitations of earlier compensatory programmes in other countries will be taken on board in this case. On the other hand, the relegation of poverty alleviation to a social sector programme, the likelihood of strong overall fiscal constraints and the weak record to date of both social sector institutions and development schemes in serving the poor, particularly poor women, also give considerable grounds for caution about the potential impact of this programme. 4.2 Health institutions and health policy There are a range of health service providers in India, including government services, private practitioners and voluntary organisations. Whilst there has been considerable investment in government health infrastructure, this has been heavily biased in favour of urban curative services. At the same time, urban-rural differentials in hospital coverage have widened (21
31

percent of hospitals were in rural areas in 1986 compared to 39 percent in 1956) and intended coverage of primary health centres (PHCs) and sub-centres has not been achieved. In 1987, there was one primary health centre per 40,215 population on average, compared to a target of one per 30,000. In 1988, only 15 percent of PHCs fulfilled the prescribed norms in terms of population coverage and supplies. There was a particular shortage of lady health visitors (LHVs) suggesting that women’s health needs specifically were being neglected. One recent study found that 7.1 percent of rural births take place in government hospitals compared to 43.1 percent of urban births. Of home deliveries, 9.4 percent were attended by medical personnel in rural areas, compared to 23.1 in urban areas. Poor and worsening coverage and quality of care in government health services has led to a mushrooming of the private health sector. In 1974, 16 percent of hospital beds were in the private sector; by 1988 this figure had risen to 30 percent. Recent data suggest that 84 percent of expenditure on health is privately rather than state financed. All social groups use private health care; the proportion of household health expenditure on private care ranges from 72 percent in low income groups to 95 percent in high income groups. Voluntary agencies have also played a significant role in health provision and particularly in promoting alternative models of community health care. There are an estimated 7,000 voluntary health organisations in operation in India. Government is increasingly encouraging non-government health agencies to take on implementation of its programmes. Access to health care is affected by income level: one study found that low-income households treated less than half of illness episodes compared to 60 percent in high- income households; health spending formed only two percent of household expenditure in highincome households compared to 10 percent in low-income households. Status within the household, clearly affected by gender, also determines access to health care (see below). Health policy as stated in the Eighth Plan stresses the goals of Health for All by the year 2000 and health for the underprivileged. Prioritization of primary health care is another theme. Beyond this, there is increasing promotion of both NGOs and the private sector as alternative providers. The quality of care offered by non-government health care providers and their ability to provide services to social groups currently excluded from provision cannot be assumed but requires investigation. If to date, there has been a failure to reach poor women in rural and urban areas, Dalits, tribals and Muslims, for example, this cannot be assumed to be only a problem of state inefficiency, since private and nongovernment actors have also been playing a major role in health provision for some time. Family planning is one area which has had an explicit and almost exclusive focus on women. Growth in family planning expenditure, particularly since the mid-1960s, is such that it is now almost equal to all other health expenditure put together. Moreover, family planning is now the lead agency in public health provision. These funding and organisational priorities reflect a dominant approach to women’s health in which women are targeted primarily as child bearers. Recent widespread criticism of family programmes has led to some
32

modifications in the use of language relating to fertility control and in the delivery of family planning services. However, approaches to family planning based on women’s individual ‘right to choose’ need to address the limitations on women’s control over and choices about their sexuality and reproductive behaviour. 4.3 Gender and health There are now a host of indicators available on women’s health in India showing their poor health status compared to men. However, there is also a need to move beyond reproducing these data to examining the causes of differential mortality by gender; gender-specific health environments; and socio-cultural and other norms surrounding health, which create different perceptions of and responses to women’s and men’s health. India’s adverse sex ratio, for example, is well known and shows no signs of improvement, having declined from 972 females per 1000 males in 1981 to 927 in 1991. This imbalance in the sex ratio is widely associated with gender-based discrimination in the allocation of food and in access to health care within the household. However, a more careful analysis reveals considerable variations in the sex ratio according to state, age, caste, communal group, and expenditure class. Kerala is the only state with a positive sex ratio at 1036, compared to states like Haryana (865) and Punjab (882) with highly unfavourable sex ratios. Excess female mortality is concentrated among young women and girls; death rates among both boys and girls in the 0-4 age group are at their highest but are higher among girls in spite of boys’ biological disadvantage at infancy; maternal deaths also account for 13 percent of female deaths before the age of 24. Scheduled Tribes were found to have higher sex ratios than the average population, possibly indicating the relatively better position of tribal women within their own communities. The comparatively low sex ratio among Dalits is harder to explain and varies considerably by state, being lower than average in some relatively prosperous states and higher than average in less well-off states. Such variations in the sex ratio need to be investigated with care at micro-level. There is also some evidence, supported by data here, that the sex ratio is less unequal among the poorest classes. Data (see Table 3) show variations in the adult sex ratio for different expenditure classes for 1990-1. There is a steady decrease in the adult sex ratio with the rise in per capita monthly expenditure which holds for both rural and urban areas. This may be due in part to the lower or absent female earnings in poorer households where there tends to be a higher concentration of women. It may also indicate that women stand a better chance of survival in poorer households where they are economically most active and therefore most needed. By contrast, in better off households, women may be perceived as a greater burden and thus, whilst better fed, may be ‘allowed to die’. Hard and fast conclusions are difficult to

33

draw, but it is clear that, among better off sections of the population, gender discrimination within the household does not disappear and may even intensify.
Table 3: Gender demography across economic classes (Compiled from NSSO, 46th Round, 1990-1, Table 1)
Monthly Expenditure Class (Rs.) % of total population < 110 110-215 215-385 >385 ALL 15.8 52.2 25.3 6.7 100 male 1554 1541 1587 1409 1543 female 1574 1507 1449 1067 1461 RURAL URBAN

Adult Sex Ratio 1013 978 913 757 947

% of total population 4.1 32.8 38.7 24.4 100 male 1372 1703 1676 1429 1591 female 1624 1654 1526 1117 1426

Adult Sex Ratio 1184 971 911 782 896

Women both share in many of the diseases to which men are prone and are also especially or uniquely affected by others, such as anaemia, gynaecological infections, backpain and depression. Issues such as the impact of infertility on women’s health; or the effects of violence (see Chapter Three) on women’s health have as yet received relatively little attention. Also, women may be more susceptible than men to diseases which cause death. Comparisons of the morbidity of men and women in the same households usually show female morbidity to be higher, possibly due to lack of health care. Differential morbidity by gender is partly due to the different health hazards to which men and women are exposed. One study in Tamil Nadu showed a higher incidence of respiratory diseases among girls exposed to smoke-filled kitchens. There may also be differences in the duration and intensity of illness by gender, these being correlated with poor nutritional status and inadequate health care. Because of undernourishment, girls are likely to take longer to recover from illnesses. Nutritional status and growth are also affected by illnesses and their severity and duration, and the health care received during and after illness. Macro-level data show little improvement in the overall nutritional status of the poor over the past fifteen years, except for the nutritional status of pre-school children; no significant differences between the sexes were found in these overall trends. However, a number of micro-studies have documented girls’ disadvantage in access to both food (including breast milk) and health care within the household. This bias may specifically affect higher birth order girls. Undernourished girls are likely to grow into undernourished women who will perpetuate an inter-generational cycle of undernourishment. These factors contribute to the low female to male sex ratio in India. Maternal mortality accounts for 2.5 percent of all female deaths in India, but for 12.5 deaths of rural women between the ages of 15 and 45. Risk of maternal death is exacerbated by high average pregnancy rates, early marriage and pregnancy of teenage girls, closely-spaced births
34

and inadequate care during pregnancy and delivery. Toxaemia and septicaemia are common causes of maternal deaths; anaemia affects 60 percent of Indian women. Another cause of the deteriorating sex ratio is female foeticide or sex-selective abortion following amniocentesis. This has been in evidence since 1977 but has spread more widely since 1982. In 1984, one study found that 40,000 female foetuses were aborted in Bombay alone in one year; further data to verify this or the incidence of female foeticide in other areas was not found. Repeated abortion of healthy female foetuses could also have a negative impact on women’s fertility and health. Legislative measures to curb the practice have been taken - in Maharashtra in 1988 and in other states, including bans on the advertising of sexselection services. Other states have followed suit but as at late 1993, national level legislation was still in preparation. There is little information on the effectiveness of legislation outlawing sex selective abortion. Women’s health is also affected by their work. The relationship between work and health is quite complex. The assumption that economic participation improves women’s status and thus health is too simplistic; it may do so but only under specific circumstances where they have control over increased earnings and feel the right to greater consumption, for example. Whilst work can increase incomes and thus spending on food and health care, this will not necessarily benefit women themselves. At the same time, longer working hours and occupational hazards can impact negatively on women’s health. Women’s consumption is inadequate, particularly among poor women, but they work extremely long hours. Adolescent girls are often working 10 hours or more a day by the age of 15. Boys are more likely to be compensated for their work with additional food.

Box 4: More work means poorer health? Women in Hyderabad slums
One study on the impact of women’s work on both their own health and that of their children, undertaken in slum areas of Hyderabad, produced disturbing results. Women’s morbidity increased with the additional work load and the fact that the hours spent on housework and child care (as much as six hours a day) did not reduce. The small increase in their earnings were largely swallowed up in paying for private health care for their children (where payment was associated with ‘superior’ services), while the women themselves relied upon the Maternity and Child centres run by the Government. These centres, however, are not equipped to deal with work-related illnesses, and were therefore of little or no assistance. (National Institute of Nutrition, 1993.) Similar studies are currently being undertaken elsewhere - for example, in Baroda (Gujarat) and Haryana - on women’s productive work and child health. They emphasise that work can adversely affect women’s health.

35

Ill-health, poverty and poor working conditions form a vicious cycle with relatively greater impact on women, who carry the burden of sickness in the family and are thus often constrained to work in the unregulated sector. Women are exposed to a range of occupational health hazards, from their domestic work as well as paid work and may be especially prone to occupational hazards because of their concentration in unregulated sectors. Specific occupational health hazards suffered by women include: bad posture; damage to eyesight; respiratory problems and exposure to dust and toxic chemicals; and forms of mental stress, anxiety and depression. Insufficient attention is paid to women’s occupational health problems which are often dismissed as caused by other factors. With the implementation of structural adjustment and the resulting likely expansion of unregulated and unorganised forms of employment, the need for systems of monitoring and protection for occupational health problems is increasing. Such systems would need to address the gender-specific nature of occupational health problems. 4.4 Gender bias in access to health care National survey data from 1990 found evidence of lower medical contact rates of female than male children and that female children were the most disadvantaged group in the household in this respect. The states with the largest gender differentials in medical contact rates were Orissa, Haryana and Punjab. Other smaller-scale studies have shown that female children are only one third of total children attending outpatient facilities; and only 16.5 percent of children admitted to hospital. Of those who are admitted, girls are more likely than boys to die, suggesting that they are only brought to hospital at an advanced stage of illness. However, survival chances of girls vary by birth order: whilst first sons and first daughters had roughly equal chances, the second plus daughter was at much greater risk, indicating selective neglect as between female children. This indicates that analysis of gender differentials in mortality and morbidity needs to contextualize gender categories according to age, birth order, familial relationship etc. At least four sets of factors conditions gendered access to health care, i.e. need; permission; ability (including affordability); and availability. The extent of women’s relative health needs has already been indicated, but evidence from elsewhere suggests that the probability of girls’ illnesses being reported is much lower than boys; women are socialised into accepting pain and suffering. In terms of permission, i.e. social factors affecting access to health care by gender, women’s constrained mobility and literacy are a disadvantage. Ability to access health care is limited by direct and opportunity costs, and the lack of fit between the timings of clinics and hospitals compared to women’s schedules. Poor women cannot afford to wait for long periods at government facilities so they tend to use private health care for all but severe or chronic illnesses. The availability of health care is restricted in terms of coverage and quality (see above) especially in rural areas. Other issues here are the relevance of the care provided to women’s needs; and the culture and attitudes prevailing in health sector institutions. Many women, especially those from rural areas, find the health system quite alien. The ‘scolding’ attitude of medical staff also dissuades women from using health

36

sector institutions. Finally, medical staff often adopt communal or other biases, for example, purveying the idea that Muslim women have too many children. This can lead to poor communication between medical personnel and client, resulting in unnecessary health risks. Awareness raising about women’s and specifically girls’ health problems may be important to promote early recognition and treatment of illness among girls. Health delivery systems need to be designed to take account of limitations on women’s flexibility in timing and mobility in attending health facilities. For a combination of these reasons, most women in rural areas continue to use home remedies or local health systems. There has been some success in integrating allopathic and non-allopathic medicine at local levels, but not on a broad institutional scale. Meanwhile, ‘modernisation’ and environmental degradation may represent threats to indigenous knowledge bases on health and health resources. Measures are required to indigenous health resource and knowledge bases and to develop their potential, especially among women. 4.5 Family planning There has been a great deal of controversy over family planning programmes in India. In some sections of the bureaucracy, among the elite and in some international agencies, the rate of growth of the Indian population is perceived to be responsible for increasing poverty, overcrowding, unemployment and so on. Others perceive poverty to be a cause rather than a consequence of population growth, since poor families may have additional children for security or as an economic resource. Family planning investments are seen by some as a substitute for development and structural change; vertical family planning programmes are relatively easy to implement, compared to long term improvements in the economy, in health and so on. Due to their primacy in funding and organisational terms, family planning programmes have been diverting resources from other health care uses; for example, Auxiliary Nurse Midwives, who are burdened with a wide range of health tasks at sub-centre level, are liable to focus mainly on meeting family planning targets and neglect their wider primary health care role. The assessment here is that the impact of family planning programmes has been relatively poor. There has been some success in increasing contraceptive usage, but limited impact on the birth rate. Terminal methods have been widely adopted but tend only to be used after family completion. Non-terminal methods have as yet made little impact. A number of limitations of family planning programmes are presented, in relation to information, attitudes, the methods promoted and the lack of a gender perspective. Information tends to be limited to single methods, usually sterilization, rather than a range of methods being presented, together with their relative advantages and disadvantages. Here again, communal bias is prevalent: the widespread belief that Muslims do not use contraception and therefore need to be specifically targeted by family planning programmes is
37

countered with evidence that, apart from markedly lower use of sterilization and higher use of the oral pill, their pattern of utilization of contraception is similar and overall only slightly below that of other religious groups. As regards methods, breast-feeding requires more concerted promotion and support, for its contraceptive effects and also nutritional benefits. This would require interventions in the workplace to assist breastfeeding. Terminal methods have a controversial history. Since the mass sterilization of men in camps in the 1970s, terminal methods have subsequently focused almost exclusively on women; the proportion of tubectomies in total sterilizations may be as high as 93 percent. Given that vasectomy is a comparatively straightforward procedure, awareness raising and promotion to counter the negative image of vasectomy is required. Similarly, barrier methods are rarely supplied to women on the grounds that they are ignorant of how to use them; in reality the need for ongoing supplies and monitoring (and therefore ongoing resources) may be the major constraint to their promotion. Injectables have attracted a lot of criticism from women’s organisations. They are considered to be under-tested, difficult to reverse and to have a range of potentially dangerous side effects. The context of lack of care in which these, as well as other contraceptive methods, are supplied, is a major problem in monitoring problems with contraceptive use. Family planning programmes need to promote a wider range of contraceptive methods, including balanced information about them all. Breast-feeding, barrier methods and vasectomy, in particular, need to be further encouraged. Follow up care needs to be provided in conjunction with family planning services, to ensure proper use, avoid complications and monitor for side-effects. In spite of the current emphasis on women’s right to choose, there is a lack of recognition of the constraints on women’s rights to choose over questions of their fertility and sexuality. Women working in the formal sector are to be penalised under new government legislation which will deny them maternity leave after the second child. There is little recognition of the fact that women are rarely independent decision-makers. It is just as important to educate men regarding responsible family planning and reproductive decision making. 4.6 HIV/AIDS HIV/AIDS has only recently become a prominent health issue in India. Since the first AIDS case was identified in Madras in 1990, 522 cases (as at the end of 1993) of AIDS have been officially reported to the WHO. A surveillance programme run by the Indian Council for Medical Research, which has so far screened at least 1.73 million people, has identified over 12,000 cases of HIV infection, of whom 15 percent were blood donors and 43 percent the ‘heterosexually promiscuous’. Cases of HIV and AIDS are spread across the 19 states but over half are in Maharashtra and Tamil Nadu. Manipur is reported to have nearly 20 percent of India’s HIV positive population (compared to only 0.2 percent of the total population) due to the high concentration of injecting drug users in the area, who form 87 percent of HIV positive cases identified in a recent survey there. These figures only represent known cases. Estimates for the total number of AIDs cases across India (which however cannot be verified)
38

vary but reach as high as 100,000 AIDS cases and two million HIV cases. Localised studies of ‘high-risk’ groups and STD clinic patients also show rising rates of seroprevalence. None of the data available was disaggregated by gender. Unregulated blood banks and the widespread use of unsterilized needles in primary health care centres are thought to be major routes of HIV transmission in India, as well as unprotected sexual intercourse with multiple partners. In 1992, the National AIDS Control Programme was established under the Ministry of Health, with a budget of Rs.2800 million ($100 million) over the five year period of the Eighth Plan (1992-1997). Of this, the major component will be funded by a World Bank loan. The main activities of the AIDS Control Programme are to be: advertising; condom promotion; and provision of hospital equipment (presumably for blood screening). This represents 15 percent of the total health budget in the Eighth Plan period; only malaria control receives more at 19 percent. There is considerable concern in the Indian medical research and voluntary health sectors about the extent of resources being channelled into AIDS control, at the expense of other communicable diseases - now feared to be on the rise - or other broader health priorities. The way in which the resources of the AIDS control programme are being used is also widely contested; in particular there are objections to the predesigned, top down, and technical nature of the AIDS control programme and the use of foreign expertise. The importation of condoms has also proved controversial, particularly given the availability of local supplies. There seems to be little analysis of the HIV/AIDS from a gender perspective, or discussion of gender-sensitive prevention strategies in India. As elsewhere, commercial sex workers are being singled out as a major ‘high risk’ group, and there are moves towards greater policing and possible mandatory testing of this group. Gender stereotypes mean that ‘bad’ women tend to be blamed for the spread of the disease; whilst ‘good’ women are not even recognised as potential sufferers. Social as well as physiological factors mean that women are at greater risk of exposure to HIV infection than men and their position often makes it difficult to take preventive measures. Some NGOs in India are working with men with promiscuous sexual behaviour (e.g. truck drivers) to offer health education and STD treatment, for example the Southern India AIDS Action Programme (SIAAP) in Madras. In spite of the difficulties of bringing sexuality into public discourse, the National Council for Educational Research and Training (NCERT) has begun preparing a sex education programme for adolescents, as part of existing population education lessons, which will include discussion of reproduction, sex-related hygiene, teenage pregnancy, HIV infection and AIDS and drug abuse. There is no detailed information about the extent of awareness and knowledge of HIV/AIDS in India, whether and how this differs by gender or other factors, which would inform the design of AIDS education campaigns. Television, magazines and friends seem to be
39

important channels of information about HIV/AIDS, but accurate perceptions of how to prevent AIDS may not be widespread. Women (particularly rural women) are less likely to have direct access to information through mainstream media and may be more easily reached and influenced by information channels such as relatives, friends, women’s groups etc. There should be greater emphasis on education and awareness raising through discussion of HIV/AIDS issues in existing women’s organisations, literacy programmes etc, with resources to back this up, rather than a one-sided focus on public information through mainstream media. Equally, there is a need for parallel ‘peer education’ programmes with men. There is clearly considerable scope for wider NGO involvement in decentralised AIDS awareness campaigns with a gender perspective. Women’s health NGOs would seem particularly well placed here. A positive aspect of the current focus on AIDS which can be built upon is the opening up of discussion of STDs, issues of sexuality and reproductive health more generally, areas which were previously considered outside public discourse. This may enable improved treatment of STDs, particularly among women who may be less likely to report them; and enable more open discussions about sexual behaviour. In this context, it is important that AIDS education also looks at the attitudes and behaviour of young men and women towards sex and relationships, and provides women with training in how to be more assertive in refusing male sexual demands, as well as training boys and young men to be more responsible in their sexual behaviour. Measures to integrate the use of funds designated for AIDS control into more general health interventions and thus to support the improvement of the primary health care system may be important in the Indian context, particularly where lack of supplies and/ or poor practices in village-based health centres may be a major cause of HIV transmission. 4.7 Water, sanitation and housing Official estimates for 1993 were that 78.4 percent of the rural and 84.9 percent of the urban population in India had access to safe drinking water, although these are thought to exaggerate actual coverage, since many villages were only partially covered. Moreover, increasing scarcity may be affecting access and these figures do not account for facilities which are not functioning. In rural areas, women and children walk long distances to collect water; in urban slum areas, poor women rely on standpipes which are beset by problems of low water pressure, short durations of supply, large numbers of users and are located some distance from women’s homes. Reliance on hand pumps in poor urban areas particularly may be contributing to a lowering of the water table, ultimately affecting access for everybody. Water collection and management is now widely recognised as being predominantly women’s work under prevailing gender divisions of labour. Poor water supply facilities and increasingly scarce water resources thus create a considerable labour burden for women and also contribute to poor health conditions within households and communities. There is a
40

need to consult women - of all communities - in the design and implementation of new water supply projects. Failure to do so has, in many instances, led to project failure. Women should also be involved in technical and managerial aspects of communitybased water provision, with appropriate incentives, training and rewards; women should not be seen as cheap or voluntary labour for community-based service provision - this will lead to poor services and/ or incur high opportunity costs for the women concerned. It is also important, however, to overcome opposition and gain the support of men for a more public role for women in these areas. In new water supply provision, there is a need to consider the range of uses of water of women, which are not restricted to the domestic sphere, but also include irrigation for agriculture, livestock, small-scale enterprises etc. The capacity and management of new systems and tariffs need to take account of the range of uses of water as well as affordability for women as well as men. The opportunity cost of women’s labour needs to be included in any cost benefit analyses of new water supply facilities, as well as consideration of the need for flexible systems so that women can fit water collection and management around other activities. Access to water in India is critically mediated through caste relations, such that Dalit women may not have direct access to village water supplies. These issues are rarely discussed in the literature or in project design. Discrimination against Dalit women in access to water indicates that this may be a priority area of intervention to support Dalit women in particular. Sanitation provision is much less widespread than water supply in India. There is a need for a more co-ordinated approach in water supply and sanitation since the impact improved water supply may be limited where lack of sanitation facilities means that new supplies are easily contaminated. In 1988-9, only 3.15 percent of the rural population had access to government-assisted sanitation facilities, with eleven percent of rural households having private facilities. Men in general may attach much lower priority to sanitation facilities than women, which may explain the lack of provision in this area; it has also been suggested that sanitation provision may be a higher priority for better-off rural women, living in the centre of villages, than poorer women on the edge of settlements, though evidence for this is limited. In 1987-8, in urban areas of India, one third of the population had no access at all to lavatory facilities; 60 percent of those who did shared toilet facilities. In urban areas, lack of privacy in sanitation is a priority for all women, not just the well-off. Since they cannot be seen relieving themselves in public, they are forced to do so under cover of darkness, leading to considerable discomfort and possible health problems. Consideration should be given to the need for increased public toilet facilities in urban areas around places where women work and live. In one sanitation project in an urban slum area, women emerged with different priorities from those proposed under a government housing scheme. Whereas the original proposals
41

favoured private toilet facilities in dwellings, women themselves preferred shared public facilities, which they perceived as more hygienic and less work (i.e. someone else would maintain them). Interestingly they also felt that provision of private toilet facilities was more likely to result in their homes being appropriated by higher income groups. Again, it is important to consult women about the design of new facilities and implementation of sanitation projects and where possible to involve them in skilled labour, technical and managerial aspects of such programmes (e.g. in latrine construction). To date, gender-specific aspects of housing development have received limited attention. Central government allocations to the housing sector fell from 34 percent of plan outlays in the First Plan (early 1950s) to nine percent in the Seventh (1985-90). Housing needs by the year 2000 are estimated at 32.6 million units in rural areas and 31.2 million units in urban areas. The Eighth Plan does, however, give priority in housing to single women and femaleheaded households, as well as Scheduled Castes and Tribes. However, no concrete measures are proposed and it is expected to take a ‘reasonable amount of time’ to meet the housing needs of these groups. In order to extend housing provision to poor urban groups and particularly to female-headed households, enabling mechanisms for the provision of housing finance are needed. This is particularly the case for those outside formal employment, of whom women are a large proportion. 5. GENDER BIAS IN SOCIAL SERVICES: Education 5.1 Women in education policy Gender bias in education dates back to the colonial period, when only a minority of upper caste and middle-class women had access to formal education, and even then, they took separate curricula, often focused on domestic skills and moral and religious education. Language formed the central divide in educational policy; the elite were educated in English, whilst the masses were instructed in the vernacular. Education for women was also largely in the vernacular of the particular state. A series of committees and commissions on education since Independence have frequently referred to the issue of women’s education. In the 1970s, it became apparent that the formal education system had not touched the lives of the majority of the population, and particularly women. Within the Sixth Plan (1980-5), the eradication of illiteracy, income-generation schemes and non-formal education for poor women were prioritised. By the Seventh Plan period (1986-91), women were beginning to be viewed as a critical human resource, who required inputs by way of training and skills development. Two recent policy documents, the Programme of Action (1992) (based on the National Policy of Education, 1986) and Education for All, updated in 1993, have come out with strong, if somewhat different, statements and guidelines regarding women’s education.

42

The opening chapter of the Programme for Action is entitled ‘Education for Women’s Equality and refers to the need for interventions ‘for women’s empowerment’. It claims to represent an important shift from a standpoint of equal opportunity, to one of pro-actively making education a vehicle for equality, and contains and impressive list of objectives for women’s education. However, there is little detail about how empowerment through education is linked to women’s living and working condition and to relations of power at the local level. In contrast, a more instrumentalist approach to women’s education is apparent in Education for All, whereby women’s education is explicitly linked to fertility reduction goals. The present emphasis on women in educational policy coincides with a change in orientation of the education system as a whole. A greater emphasis is to be placed on elementary education and on minimum levels of attainment at this level, rather than seeing primary education merely as a preparatory phase in the school system. In this context, attention is required to what benefits the attainment of an elementary school degree can specifically bring to girls, especially those in rural areas. This shift in policy finds expression in changes in funding allocations within the Eighth Plan (1992-7). The share of elementary education in overall education expenditure is to increase from 35 percent in the 1970s and 1980s to nearly 50 percent in the 1990s. Social safety net funding via the International Development Association (IDA) arm of the World Bank will also be made available through the District Primary Education Programme. The share of secondary education in total education expenditure will diminish to 18 percent in the Eighth Plan (compared to 24 percent in the Seventh Plan) and of higher education, more drastically, from 16 percent to eight percent. Higher education is increasingly seen as a luxury which should therefore be subject to marketization and privatization. Support to technical education remains the same at 14 percent. Adult education, first introduced with the Sixth Plan, currently receives nine percent of education expenditure. Whilst it is too soon to give any detailed sense of what these allocative shifts will mean on the ground, the increased proportion of funding to elementary education, if accompanied by appropriate policy initiatives, holds the potential for improved investment in female education. Conversely, the reduction in the proportion of funding at secondary and particularly higher education levels implies increased reliance on private funding, and/ or possibly a reduction in coverage or quality of provision in these sectors (depending on the trends in overall funding, demographic shifts and on the extent to which complementary state-level funding is invested in these areas). Given demand-side constraints to female education (see below), this may reinforce existing biases against young women at higher levels, unless countervailing measures are taken, as for example in Andhra Pradesh, where schooling is now free for girls up to college level. As well as these shifts in broad allocations, recent policy documents stress community participation in education and the need for monitoring of educational provision, through panchayat committees with ‘due representation of women’. The emphasis on community participation and monitoring allows some scope for greater accountability in service
43

provision. At the same time, it implicitly assumes that local-level institutions can adequately represent the range of social groups within communities and that community members, including women, are willing and able to take on these tasks. 5.2 Trends in enrolment, retention and literacy Although growing numbers of children from all backgrounds are attending school in India, there are significant numbers who never enrol and even more who drop out soon after enrolment. Aggregate data conceal considerable regional, rural-urban and other biases in enrolment, retention and literacy. There are a number of biases in provision and access, towards urban areas, towards upper caste groups, towards Hindus rather than Muslims, towards boys rather than girls, which are mutually reinforcing. Those least well served by the education system are thus tribal girls from remote rural communities in ‘educationally backward’ states.

Table 4: Gross enrolment ratios, primary and middle school, by gender and caste, 1992-3
PRIMARY(Classes I-V) BOYS GIRLS OVERALL 118.1 92,7 SCHEDULED CASTES 128.0 92.2 SCHEDULED TRIBES 126.7 88.6 Source: Education for All - A Graphic Presentation, NIEPA, 1993 MIDDLE (Classes VI-VIII) BOYS 80.5 75.1 58.5 GIRLS 53,8 44.0 32.0

The percentage of girls in overall enrolments at each level of education has been growing over time. However, a gender gap in enrolment remains at all levels and although girls are enroling in increasing numbers, a much smaller percentage are retained in the system. Official data on education should be assessed with caution, since gross enrolment ratios (GERs) are susceptible to inflation, particularly where enrolment campaigns are underway. In 1991-2, the gross enrolment ratio (GER) at primary level was 88.1 for girls compared to 116.6 for boys. At upper primary level the corresponding figures were 47.4 percent for girls and 74.2 percent for boys. However, there is a great deal of variation between states: in Rajasthan, the primary and upper primary GERs for girls are 50 and 23 percent respectively; the corresponding figures for Kerala are 98 and 104 percent. GERs at primary level for Dalit girls are only slightly lower than the overall female average; for tribal girls the difference is slightly more marked. In contrast, the GERs for Dalit and tribal boys are higher than the overall average at this level (at least according to official data). However, the enrolment ratios of Dalits and tribals drop sharply at higher levels.

44

The number of children between 6 and 14 who have never enrolled in school may be as high as 28 million, which accounts for 18 percent of that age group. All evidence suggests that this number would include a higher proportion of girls than boys. Other than lack of any access for these children, high drop out rates are the most serious issue affecting basic education. Drop out rates are higher for girls than for boys across all social groups, which is why disparities between girls and boys increase at higher levels of education. In 1988-9, the drop out rate for girls between Class I and Class VIII was 68 percent, compared to 59 percent for boys. Drop out rates for Dalits and tribals are higher for both sexes but the gender gap remains. Drop out rates are also significantly higher in rural than in urban areas, with only 25 percent of rural students overall reaching Class VIII. There are particular constraints on the educational progress of first generation learners, where the tutoring role of parents, and particularly mothers, is absent. Whilst such problems affect both sexes, the impact on girls is likely to be greater, since they are less mobile may thus have less access to other sources of support. The spread of private tutorial classes to help pupils cope with the burden of the school curriculum and the competitive examinations system may also be favouring boys. There has been some narrowing of the gender gap in enrolment at secondary level but a large gender differential remains. GERs at this level were 54 percent for boys in 1990 compared to 42 percent for girls, having increased from 39 and 20 percent respectively in 1980. A worrying trend is that the rate of increase in the ratio of girls to boys in secondary education slowed during the 1980s. Drop out continues at secondary level, again with a gender differential across all social groups and by Class X drop out is highest for tribal and then Dalit girls. The rural urban divide also continues at this level, with 12 percent of urban girls able to complete a secondary education compared to only two percent of rural girls. Almost half the graduates of higher education institutions are urban males (who form 12 percent of the total population); by contrast rural women make up only 4.8 percent of graduates (although they form 38 percent of the population). Overall, the proportion of women’s enrolment in total enrolment in higher education increased from around 11 percent in the early 1950s to 30 percent in 1987. Interestingly, the gender gap is shrinking faster at post-graduate level. Only a tiny proportion of women from among Dalits, tribals or Muslims have access to higher education; entry for women from these groups may only be possible when they have attained middle-class status. The reservations policy for Scheduled Castes/Scheduled Tribes, to the extent which it is been effective, appears to have largely benefited men from these groups. Moreover, of those women who do enter higher education, 55 percent are concentrated in the arts or humanities. There are some shifts in gender representation across subject areas, however, with increasing numbers of women taking commerce and science courses as the number of undergraduate colleges has proliferated. The recent decision to cut back on public expenditure in higher education and moves towards privatization in the sub-sector are likely to impact negatively on women. Cutbacks appear to
45

be concentrated in arts and humanities subjects, with a high representation of women. Families may be increasingly unable or unwilling to bear the private costs of higher education for daughters as well as sons. Cutbacks in university funding could also affect women’s studies programmes, through which much of the research into gender inequality has been carried out. In 1991, official literacy rates for men and women were 64 and 39 percent respectively. Although proportionally, the illiterate female population has declined from 78 percent in 1981 to 61 percent in 1991, in absolute terms the same population has increased in numbers by 59 million. Over two thirds of the literate female population live in rural areas. Variations in literacy rates between states and regions are considerable. But not even the ‘best’ state has equal literacy rates for women and men. There is a strong correlation between primary enrolment rates and female literacy. In terms of literacy differentials between different social groups, rural tribal women fare worst at 13.4 percent. The Total Literacy Campaigns (TLCs) which began in Kerala in 1990 and swept through many parts of the country have changed the face of adult literacy programmes, which had been making relatively slow progress prior to this. What has not yet been adequately analyzed is the overwhelming presence of women in these campaigns both as learners and as volunteer teachers. Within the TLCs, women have also organised to demand better facilities or income-generation schemes. However, this movement now faces a serious challenge in terms of sustaining the gains made. More experienced teachers are required to sustain the involvement of neo-literates What is necessary at this stage is a careful analysis of the successes and failures of the Total Literacy Campaigns to date. Especially important would be a better understanding of the motivations behind women’s high level of involvement and how this can be built upon. Non-formal education (NFE), designed to meet the needs of children in poor areas who have considerable work responsibilities, is more flexible than the formal system. It is the most rapidly expanding sector in education in India. The number of NFE centres in India has grown from 126,000 in 1986 to 238,000 in 1993, with all-girls centres increasing from 20,500 to 79,000 over the same period. However, even in these centres, girls form only a third of total enrolment. One of the major concerns regarding NFE is that it will turn out to be a second rate form of education for poor girls and boys by allowing the formal system to ignore these pupils. Similarly, the recruitment of voluntary teachers to cover teacher absenteeism avoids tackling the real issue; an alternative would be to create mechanisms to support improved professionalism and commitment among existing teaching personnel. There is concern that girls will in the near future become the main recipients of NFE and thus benefit from qualitatively poorer education than boys. Whilst the flexibility of non-formal education allows girls with heavy workloads to get some access to education; measures are also needed to reduce their workload, rather than work around it. NFE
46

also requires well conceived bridges into formal education provision to prevent ghettoization of disadvantaged groups in this sector. The Early Childhood Education programme began in 1986, apparently as an extension to the Integrated Child Development Scheme for pre-primary child care. It aims to be both a preparation for primary education and a support service to school-going girls and working mothers, to relieve them of childcare responsibilities. However, providing only a limited creche service will not effectively enable girls to go to school or mothers to work. Increased convergence and co-ordination with primary school locations and timings is required to make pre-school schemes more effective. 5.3 Differential access to education by religion and caste Data show that Muslim women are constrained in their access to education, but little is known of the precise factors behind this, since it tends to be attributed to ‘culture’ or ‘tradition’ rather than to problems in provision. Educational provision in Muslim areas is differentiated along both class and gender lines; elite families will tend to send their children to secular schools as a route to upward mobility, whilst the poor mainly use religious schools or madrassas. More women than men also attend religious schools. The growing communalization of Indian society may also be leading to increasing reluctance in the Muslim community to send women to non-Muslim schools. Other factors connected with marriage and purdah restrict educational opportunity for Muslim women, but not necessarily more than for other social groups. Provision of single sex facilities or schools, particularly at postprimary level, is important in terms of increasing the access of Muslim girls to education, especially in more conservative regions. Gender differences have not been made an explicit part of the government reservations policy to increase access of Scheduled Castes and Scheduled Tribes to education. For this reason, reservations have probably benefited males more than females and this is reflected in enrolment data. As noted above, tribal women are the worst served under the present system. Box 5 provides some detailed data on differentials in educational enrolment and drop out by gender, caste etc. Pressures on government to improve implementation of reservations may increase in future; it would be important to incorporate a gender-aware analysis and monitoring of the impact of reservations to feed into policy and implementation mechanisms.

47

Box 5: Gender, caste and tribal differentials in educational enrolment and drop out
* In Classes I to V Scheduled Caste children account for 15.79 percent of total enrolment; of this only 39.46 percent are girls. * In Classes VI to VIII, Scheduled Caste children account for 13.61 percent of the total enrolment and only 33.16 percent are girls. * In Classes I to V Scheduled Tribe children form 7.91 percent of total enrolment; of this only 38.37 percent are girls. * In Classes VI to VIII, Scheduled Tribe children account for 5.19 percent of total enrolment and only 32.33 percent are girls. * At 68.73 percent, girls belonging to Scheduled Tribes have the highest drop out rates at primary level. * The corresponding figure for Scheduled Caste girls is 53.39 percent. (Grover, forthcoming).

Various factors may lie behind the high drop out rates of low caste pupils, particularly girls. Location of schools in the upper caste part of village is one factor; this may particularly affect young girls who are vulnerable to harassment. The grinding pressure of facing prejudices at school everyday should also not be underestimated as a deterrent to attendance (see Box 6). Attempts to increase the access and retention of girl students need to consider specific constraints to tribal, Dalit, and Muslim girls

48

Box 6: A Dalit woman in Maharashtra
When we were transferred to Uttar Pradesh (UP) we decided to say ‘don’t know’ if someone asked our caste. We would say we are Indian from Maharashtra. While we were living here, I never thought that we belonged to a low caste. In UP we found out that if we told our caste we wouldn’t get water to drink. I always wondered if I should take advantage of our caste in order to get reserved seats or concessions in college or school, but I have never done it...I felt I should hide my caste, my religion [Buddhism], because people won’t think much of me or respect me if they knew about them. So I never talked about all these things. In school or college I had to have a ‘proof of caste’ for scholarships or for concessions in fees. I would not tell about my caste. Once while we were in college, our group went on a picnic. The girls had asked me to bring puris and chutney. So I took that. But only one girl bothered to eat it and praise it, the others never came near the food and I had to bring it all back. I felt very hurt and after that I never went to a picnic. I went through graduate and then post-graduate courses avoiding people! Though I had scholarships and I was recognised as an English scholar there were still those who didn’t want to have anything to do with me. So I was always a loner. - Ashoka, cited in Bhave,1988

5.4 Constraints and opportunities for female education There is considerable consensus in the literature about reasons for the lag in girls’ participation in education compared to boys and for the higher drop out rate of girls. Various factors have been extensively reviewed, often classed as supply- and demand-side constraints. However, in reality they are inter-related and the relative importance of different factors is not always easy to assess. Whilst constraints to female education have been extensively documented, less is known about factors enabling girls to stay in school. In view of the current emphasis on increasing access and retention of girl students, more insights in this area would be of value. On the ‘demand’ side, family considerations, poverty and work demands, and labour market factors are major constraints. These effect boys as well but are gender-specific in their impact. Concerns about girls’ education in relation to their marriageability but also lack of incentive to invest in daughters who marry out affects both middle and lower class families, though in different ways. The extent of exogamy and thus the extent of ongoing connection and support between daughter’s family and natal family may also effect incentives of parents to invest in

49

female education. There are contradictory pressures, to educate daughters in order to marry them into more educated families, and/ or provide them with good career prospects (where second incomes are increasingly vital to households), but not to the extent of making it difficult for them to find a partner, who would be expected to have a higher level of education. In some instances education may been seen as a substitute for dowry. Poverty is also experienced differently according to gender. Biases in intra-household allocation of resources mean that boys tend to be favoured in terms of food allocations, and spending on health care and also education, although there is less evidence here. These biases also interact, so that, for example, girls may perform less well at school because their health and nutritional status is poorer. It is also argued that the opportunity cost of girls’ domestic labour is higher, particularly where they are substituting for working mothers’ labour in the home. Policies to promote female employment need to address the possibility that girl’s education may be affected in the short term when their mothers go to work. One major policy for addressing demand-side constraints has been the provision of subsidies and incentives to encourage girls’ attendance, to reduce the direct costs of girls’ education and to offset the opportunity costs of girls’ education. Education is free for all girls up to secondary level and in some cases beyond, e.g. in Andhra Pradesh. Incentive schemes include free midday meals, uniforms and text books as well as scholarships. Scheduled Castes and Scheduled Tribes are specifically targeted with such schemes; the proportion of school children benefiting varies between states from 13 to 27 percent, mainly focused on rural areas. Girls are not specifically targeted. Studies of the impact of such schemes in Tamil Nadu have shown that enrolment and retention rates have increased. Research into the impact of measures to increase female enrolment has shown that packages of measures work better than single measures in isolation and that such packages need to include measures to offset the costs to parents of female education. However, the relative contribution of different interventions to increasing female enrolment in such programmes is poorly understood. This points to the need for project-related research to identify the relative impact of different factors. Incentives and subsidies are one such approach, but can be counterproductive if they are too narrowly targeted, due to high costs of selection and the possibility that those who fall outside the targeted category might be withdrawn from school. Labour market discrimination is another gender-specific constraint which indirectly affects female education. Whilst education increases the potential earnings of both sexes and the gender gap in earnings narrows at higher levels of education, men still earn more than women at the same level of education. For example, a literate man may earn the same as a female secondary school graduate. Further, the unemployment rate of female college graduates is much higher than that for males at the same level, particularly in rural areas, which may be partly a reflection of gender bias in subject areas discussed above. Incentives for parents and female students themselves to continue investing in female education are limited where boys with the same education can earn more and there is a higher risk of unemployment for females. Economic as well as narrowly educational policies are needed to address
50

gender bias in education. In particular, more attention is required to improving employment opportunities for educated females at different levels of attainment, perhaps by relaxing recruitment criteria for certain types of job. Supply-side constraints include ‘school factors’, i.e. the quantity type and quality of school provision, the distance of schools from residences, facilities provided, the nature of the curriculum, teacher: pupil ratios, and the characteristics and quality of teaching staff. Whilst coverage of primary schools in India is quite high overall - 94.5 percent of the rural population have a primary school within one kilometre - there are still insufficient schools in remote tribal communities. In these cases, distance will be a particular deterrent to girls’ attendance. For Dalits, social exclusion rather than physical distance may be the problem, i.e. location of the school in upper caste parts of villages. Facilities in rural primary schools are poor in terms of access to drinking water and sanitation facilities and classroom provision. In rural areas, separate facilities for girls are almost nonexistent in primary schools and single sex schools few, even at the upper primary level. Fear for girls’ safety and moral reputation is a gender-specific deterrent to school attendance, in travelling to school, in mixing with fellow students, in coming into contact with male teachers and in living in residential facilities away from home. Given the very real risks of harassment and attack, the question of girls’ safety in travelling to school, and of vulnerability to abuse or harassment by male students and teachers has not received sufficient attention. Accommodation for female students is often inadequate. Government hostels for Dalit and tribal women are overcrowded and have poor facilities. There have been reports of rape and sexual abuse of Dalit women in such hostels. Adequate security provisions need to be made for female boarders. There has been considerable work on revising the curriculum in India, including from a gender perspective, e.g. by making textbooks more girl-friendly. The revision of text books needs to go beyond the addition of more positive images of women and girls to incorporating discussions of gender discrimination. Other kinds of bias which interact with gender biases also need to be addressed in educational materials - e.g. caste and communal biases. There is also a need to address the ‘hidden curriculum’ which contributes significantly to reproducing gender disparities, e.g. teacher-pupil interaction, staff hierarchies, extra-curricula options and so on. The number of pupils per teacher has been rising at both primary and upper primary levels (from 39 to 45 in the case of primary schools and from 34 to 43 at the upper primary level between 1981 and 1991). Moreover these aggregate figures mask urban bias and regional differences. At present, female teachers form only 21 percent of primary and 23 percent of upper primary teachers in rural areas

51

Middle-class urban women who are qualified teachers face constraints in relocating to rural areas. There is a need to identify women candidates for teacher training (perhaps with associated incentives and revised recruitment requirements) from rural areas. The relatively high unemployment rates of the tiny minority of rural female graduates suggests that there may already be potential recruits here. However, increased recruitment of female teachers should not become a rigid policy in the short-term where candidates are in short supply; there is also a need to change the stereotype that male teachers are a threat to girls and to deal with offending male teachers in actual cases of sexual harassment or attack. 6. THE STATE, NGOs AND THE MAINSTREAMING OF GENDER ISSUES 6.1 The women’s movement in India The women’s movement in India has a long history, dating back to the nineteenth century, and has played a key role in influencing both the state and wider civil society, especially in its most recent phase since the early 1970s. In the 1990s, however, a wide range of actors have taken up women’s, or gender, issues as part of their agenda, including the state, political parties, international agencies and a growing number of non-government organisations. Considerable gains have been made by the women’s movement and there is a wealth of experience - of grassroots organising, of the design and implementation of services and development projects, of lobbying the state and of working within state structures - to draw on. The women’s movement has also built up a substantial body of research on women and development and gender and development issues as well as an institutional research capacity. Particular issues taken up by the women’s movement in the last two decades include: violence against women; the law and legal reform; health and family planning issues; and issues around women’s work, employment opportunities and employment conditions. Many women’s organisations work in several of these areas at the same time. A number of strategies have been adopted in relation to these issues. For example, work around violence against women has included public awareness campaigns and demonstrations, lobbying the government for legal reform, taking up individual cases and extending support to women victims, confronting perpetrators of violence against women and organising social boycotts of offenders. Demands for changes in police structures and practices have also featured strongly, resulting in the setting up of special women’s units and special all women police stations in some cities. However, these strategies in relation to the police are now seen as having further marginalised women victims of violence within police structures. There is now a shift (as in Maharashtra) towards composite policing, where training and equal opportunities policies are being integrated. The new shift away from

52

women-specific interventions to the integrated gender-aware policy of composite policing needs to be encouraged in future strategies.

Box 7: The Anti-arrack Agitation in Andhra Pradesh
Protests against government-backed sales of country liquor or arrack have been taking place over the last decade in different parts of Andhra Pradesh with support from different left parties. However, it is in certain coastal Andhra districts, especially in Nellore, that protests gained momentum during 1992 in conjunction with the spread of adult literacy classes under the Total Literacy Campaign. Rural women acted collectively within their respective villages to stop the sale of arrack by destroying arrack shops and burning arrack packets, attacking excise officials and the excise police, and publicly shaming or confronting men who drank. The movement spread rapidly to different parts of the state, including interior parts where greater state repression had initially prevented the movement from gathering momentum. Women began to demand the banning of arrack in their villages. The strongest catalysing force behind the movement was the literacy classes. As in other states, these classes were primarily attended by women and girls; at least half the teachers were women. In Nellore, the Jana Vignana Vedika, a left-oriented voluntary organisation, aided by administrators and government officials, has been co-ordinating the literacy drive. At the beginning of the movement, an inaugural function for the literacy programme in a village, attended by a State minister and the District Collector, was disrupted by some drunken men. The women of the village demanded the closure of the arrack shop so that they could have their classes in peace. Willing to promise anything to ensure the success of the programme, the officials complied. Stories featured in the literacy primers also provided inspiration to the movement, although mostly it was personal experiences and tragedies that led women to join the struggle. The government has been heavily dependent upon revenues from the sale of arrack, which have been rising steadily over the years to Rs.6300 million in 1991-92. The Congress Party in power was, therefore, caught in a quandary when the major opposition party, the Telugu Desam, first attempted to take over the movement. The opposition leader, N.T. Rama Rao, toured the state, holding massive rallies in which he proclaimed his support for prohibition. The irony was not lost on everyone, since it was during his term as Chief Minister in the early 1980s that mass production and sale of arrack was first launched. The government could not attack or curtail the movement because there was no central leadership; women were taking up local agitations and initiatives in their particular villages. Moreover, prohibition is enshrined in the Constitution as one of the directive principles of state policy. It also has a strongly moral tinge and invokes images of Gandhi. As the result of all these conflicting pressures, the state government had to weigh the loss of revenue and illegal pay-offs against the fear of losing in the next elections. In 1993, it decided to enforce prohibition. At the time of writing in October 1994, there is a further move by different women’s organisations to extend prohibition to all forms of liquor. The government has announced that it will honour the demand if it comes from women. (Anveshi, 1993).

53

A recent example of a popular women’s movement is the anti-arrack campaign in Andhra Pradesh (see Box 7) which came to a head in 1992, in protest against government-backed sales of arrack (or country liquor) in poor rural communities. This mobilisation has different features to the ‘typical’ campaigns of the women’s movement, being rurally based and mainly taken up by uneducated, poor and low caste women. The movement had no identifiable leadership and spread from village to village assisted by the Total Literacy Campaign being promoted at the time. Women organised at village level to prevent the entry or sale of arrack in their villages and by shaming men who drank. Arrack became a focal point, enabling the women to comprehend their daily problems related to work, the family economy, health, education and their personal lives. Women felt the acute injustice of a situation where water had to be brought from a distance of four kilometres in some villages that were regularly supplied with arrack. Women wanted better health services and schools for their children. They referred to the change in their lives and the peace of mind they enjoyed in villages where arrack was no longer sold. What is surprising about this movement is that range of interest groups, including the state government which felt compelled to lend their support to it. Responding to pressures from the mounting campaign, in 1993, the state government prohibited arrack sales. This shows a very concrete way in which women at grassroots level have been able to influence state policy. It would be important to gain some sense of what this unique movement has meant to the rural women themselves, where their support has come from, what forms of opposition they have encountered and in what ways this success can be built upon. Since 1993, an all-India women’s organisation, SAMATA, has come into being to form a support network by helping to maintain the momentum around the literacy campaign. Women’s organisations have focused a lot of energy on influencing state policy; they have also had widely varying attitudes towards the state, ranging from support and collaboration, to mediation and direct confrontation. In the 1990s, there are some key shifts in the relationship of women’s organisations to the state which may determine future priorities and strategies: - Attention has shifted from of focus on policy changes to questions of implementation; - Women’s organisations which have taken a stance of holding the state accountable for its failures in services provision are now facing a dilemma, as women’s organisations themselves, along with NGOs more generally, are being actively encouraged to take up service provider roles in areas previously served by the state; - Divergence and tensions between non-government development organisations and women’s movement organisations, the latter being mainly urban-based and advocacyoriented, the former being primarily rural and involved in development programmes, may be subsiding in the face of shifts in the broader political and economic context;

54

- The shift in locus of control and planning away from the central state with liberalisation and decentralisation implies a need for women’s organisations to look elsewhere - perhaps to international economic fora, for example - to promote their struggles; - There are also spaces for women’s organising and lobbying opening up at the state level, where concrete policy shifts and resource allocations may be obtained, and also at district and lower levels, with the decentralisation process. 6.2 Government policy on women and the national machinery Government policy towards women in the immediate post-Independence era was basically welfarist in orientation. Mahila Mandals or women’s centres set up in the context of Community Development Programmes featured child nutrition programmes and income generating activities, but drew in mainly relatively well-to-do women. As noted above, the 1970s was a decade of intense political opposition to the government, including feminist activity. In 1974, the Committee on the Status of Women, comprising leading feminists as well as technical experts, drafted the report Towards Equality under the aegis of government, as part of its obligation to report on its progress towards women’s equality for the 1975 UN Women’s Conference. This report is a founding text of the Indian feminist movement. It concluded that, barring the gains made by a minority of middle-class women who had found employment in the expanding public services sector, the situation of women had worsened in many respects, in spite of two or three decades of planned development. Government policy underwent a shift to a typically ‘WID’ approach with the Sixth Five Year Plan (1980-85), which recognised women’s central role in the economy and incorporated a number of the demands of women’s organisations, for example, the need to give joint titles to husband and wife where transfers of assets occurred. The Sixth Plan also contained contradictions, however, presenting the family as the basic unit of development whilst supporting the need for women to increase their economic independence. The Seventh Five Year Plan (1986-91) went further, presenting individual women as potential beneficiaries of development schemes and emphasising the need to raise awareness about women’s oppression and to build women’s self confidence. It also envisaged an integrated and multipronged approach to women’s development, incorporating employment, childcare, education, nutrition, health etc. The National Perspective Plan on Women, 1988-2000 - the government’s second major policy document on women - was also published in this period, combining a review of existing approaches and programmes with recommendations for future action, including a strong emphasis on training as a strategy, the separation of welfare and development activities between the Social Welfare Boards and Women’s Development Corporations, respectively, as well as an increased role for the voluntary sector.

55

With the Eighth Plan (1992-1997), a decisive shift in overall development strategy has occurred away from centralised planning at a national level. There is now strong emphasis on promoting community participation and people’s initiative, and a range of institutional options for the delivery of services, particularly through the increased role of the voluntary sector. However, there is no special emphasis in the Eighth Plan on the participation of women, nor other marginalised groups. The range of existing policy documents on women do not provide a coherent strategy or approach. The language of empowerment and the demands of many women’s organisations have been incorporated in places but instrumental approaches are also apparent; in others, a welfare approach appears to persist, or there is little mention of women at all. Welfarist approaches persist particularly in relation to Scheduled Caste of Scheduled Tribe women, or those from educationally backward or minority districts. Elsewhere, ‘poor women’ are referred to as an undifferentiated group, with no reference to other axes of discrimination, along caste or communal lines. Nor do the various government policy documents collectively provide an overall policy framework. They tend to serve as plans of action rather than policies disseminated through the various ministries and to which the government can be held accountable. An exception to this, at state level, is the recent Maharashtra State Policy on Women. This is ground breaking because it attempts to link interventions in different sectors and commits the government to specific resource allocations. Also, for the first time in a government document, there is recognition of the need to address cultural or traditional factors which keep women in subordinate positions, identified particularly as family issues. Another important facet of this policy document is that there was widespread consultation with women’s groups and minority bodies in its preparation. The National Machinery on women comprises a variety of cells, Departments, corporations and boards, as well as specialised units within government departments. In 1993, the National Human Rights Commission was set up with a mandate to investigate violations of women’s rights. Besides the lack of overall policy framework highlighted above, there are a number of other reasons for the collective failure of these bodies to influence government policy making processes. Firstly, there is no regular system for incorporating findings of reports in different sectors, nor for consultation and discussion between different Departments to integrate their efforts. This means that the Department of Women and Child Development, which should play a key role in advocacy within the government, in practice does all the WID work and as well as implementing its own programmes with limited resources and staff. There is also little networking or collaboration between the various components of the National Machinery and both the National Machinery and state-level bodies lack resources and institutional powers. There is a need for greater attention to transforming existing departments from within by working with key personnel over a period of time: this is the challenge of mainstreaming.

56

The ratification of the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW) provides a further tool to make state policy accountable and to hold specific departments responsible for their functions in relation to women and development. 6.3 Women in development schemes One manifestation of the increasing government focus on women is the increase in the number of development schemes aimed specifically at women, currently numbering over 200, although most are concentrated in a limited range of departments and some are externally funded. Most are narrowly woman-focused with little emphasis on gender relations. The record of government in the implementation of development programmes is weak; this is now being addressed mainly through greater involvement of NGOs. Mechanisms are in place to allocate a proportion of funding to voluntary sector activity in every social sector ministry. Further, the government plans to set up an all-India monitoring agency to oversee the implementation of rural development schemes, consisting of senior ministry officials and the heads of ten NGOs. It is not clear whether this agency embodies any gender expertise nor whether gender-sensitive criteria for monitoring and evaluation will be employed. Poverty alleviation schemes can be broadly divided into those which are aimed at creating self employment (e.g. IDRP, DWCRA); those which provide temporary wage employment on public works schemes (EGS in Maharashtra, JRY, NRY); and minimum needs programmes in nutrition, health, sanitation, housing etc. Some such schemes specifically target women; others are intended to include a percentage of women beneficiaries; others make no specific provision for women’s participation. There are both overall and gender-specific problems with these schemes. In self-employment schemes, provision of credit in itself is insufficient to stimulate long-term sustainable selfemployment without other supports, market demand and market access. Targets for women’s participation in mixed schemes may not be met or may be subverted. Women’s participation can be hampered by head of household approaches, particularly where male defaulters are barred from schemes, sometimes leaving women with little support ineligible for access to credit. Experience from elsewhere has shown that even where women are targeted with loans, they may not control their use or the returns from investment. In women-specific programmes, there are often considerable pressures on women field workers, who may be expected to work collectively at grassroots level, in the context of vertically organised programmes and who are expected to perform a wide range of tasks, with insufficient support and training. For those poor women who do participate in self-employment schemes, there are high expectations, to act as entrepreneurs, understand the market and take risks, with limited support. Areas of activity which are encouraged may be in markets with other larger players. It may be necessary to consider more interventionist measures which would offer a

57

niche in the market, or minimal guarantee of market access to those who seek selfemployment in such areas. Wage employment schemes also have drawbacks. They tend only to provide temporary or seasonal employment at very low wage rates and thus whilst they may perform a limited safety net function, they are unable to pull beneficiaries out of hard-core poverty. More attention is needed to asset creation at village level in wage employment programmes. The extent to which women participate in such schemes is highly varied (ranging, for example, from 6.8 percent in JRY to 40 percent in EGS). Further, wage employment schemes have often been discriminatory in wages offered, with no justification. The extent to which poverty alleviation programmes have allowed women specifically to build up assets and over what period of time, needs to be examined. More attention needs to be paid to factors which would facilitate women’s participation (e.g. type of work expected, recruitment mechanisms and attitudes of those recruiting workers, childcare provision), drawing on the lessons from more successful schemes in this respect. To date, poverty alleviation programmes have been concentrated in rural areas, since poverty has been assumed to be a largely rural phenomenon. Since 1989, NRY has offered support for micro-enterprises and wage employment in urban areas, but coverage is limited and the proportion of women benefiting is not known. However, recent data suggest that urban poverty may be more widespread than previously thought and structural adjustment may intensify urban poverty, with gender-specific impacts. In this context, existing anti-poverty programmes in urban areas need to be extended and improved upon, with particular attention to their gender impact. Small towns particularly may be being by-passed by current anti-poverty efforts. Schemes for the educated unemployed do not have special provisions for women, in spite of the fact that the unemployment rate for educated women is higher than that for men and is particularly high in rural areas: 23 percent of female high school graduates reported unemployed in rural areas compared to only seven percent of men with the same level of education; at college graduate level the figures were 30 percent and 10 percent respectively. The possibility of rising retrenchments under structural adjustment may add to the problems of the educated unemployed. Particular action is needed to assist educated women, especially those in rural areas, to enter or re-enter the employment market, where they face different constraints to men. Alternative development programmes for women have also been sponsored by the Indian government: in particular the Women’s Development Programme in Rajasthan (WDP) (begun in 1984 in Rajasthan) and the more recent Mahila Samakhya programme in Uttar Pradesh, Gujarat, Karnataka and Andhra Pradesh. Rather than a focus on service provision or poverty alleviation per se, these programmes have adopted an awareness raising and ‘empowerment’ approach. They have a number of particular features: interaction and cooperation between government and NGOs; key role of women field level workers; collective participation of women; emphasis on women themselves providing direction to the programme; and relatively low levels of funding.
58

In WDP, tensions arose between the NGO and government sides of the programme and a number of problems emerged regarding the role of sathins (village level animators). Sathins were under pressure to take up a number of functions related to government programmes. Some of them aspired to become regularised workers, with commensurate rewards, rather than acting in a largely voluntary capacity as grassroots animators. Although programmes such as WDP and Mahila Samakhya are seen by some as models, more systematic evaluation of their benefits is needed before such approaches are widely adopted in other programmes. There is evidence of qualitative benefits, but there is also a need to assess, over large samples, to what extent changes in service delivery and women’s access to resources have occurred and what has happened to livelihoods and poverty alleviation as a result of these programmes. This raises the broader question of whether it is possible to measure the impact of empowerment on more tangible economic and social goals. 6.4 NGO approaches to gender issues As many as 18,000 voluntary associations are registered with the government in India, including several thousand non-government development organisations (this does not include grassroots local groups, of which there may be over 100,000, of whom half are probably women’s groups). However, of the thousands of NGOs, only a small proportion - probably less than 20 percent - focus activities explicitly on women and even fewer are ‘women’s’ NGOs. Moreover, there is a concentration of NGOs generally and women’s NGOs specifically in the West and the South. In the North, where gender disparities are the worst, there is a low density of NGOs. The resources, infrastructure and skills base of existing women-focused and women’s NGOs and networks needs to be drawn on in order to support the local development of NGOs in areas of the country where NGOs currently have little presence. Key areas of NGO activity with women include: economic support and services; infrastructure provision; health and family planning; social, political and legal issues; labour conditions; and support or resource functions (e.g. training, research, documentation, networking etc.). NGOs working with women, or with a gender perspective, in India range in their activities across a number of sectors and issues, and many may be involved in several areas at once. The multi-functionality of many women’s NGOs may limit their ability to access resources where these are channelled in a sectoral framework. A number of different types of NGO or voluntary agencies are involved in women-specific programmes or work from a gender perspective. In the past, there has been something of a divide between women’s movement related organisations, who have been largely urban based and advocacy oriented, and non-government development organisations, who are mainly rurally based and have been involved in social programmes and poverty alleviation for a considerable period. Some women’s organisations have become increasingly
59

professionalised and taken on a more NGO character, whilst others have attempted to maintain a movement character. There have been shifts in thinking and approach in a number of areas of NGO and women’s organisation activity over recent years. Income-generation schemes promoting traditional feminine skills such as pickle making, sewing etc. are now widely seen as limited in impact: in fact they may increase women’s workload for little return. Often such programmes have failed to assess market demand or provide adequate market access and thus lack economic viability. In health, women’s NGOs particularly have been highly critical of the primacy given to family planning programmes by government and donor agencies, of some of the methods adopted in such programmes, and of the concomitant lack of provision for women’s general health needs, about which considerable evidence now exists. There is increasing interest, also, in non-allopathic medicine, where women (especially from Dalit or tribal backgrounds) are thought to have some knowledge and skills, particularly in rural contexts where access to government services is negligible. In infrastructure development, women have organised to demand better services from the state; there is increasing emphasis on areas such as water and sanitation provision, perceived as women’s priorities and on increased control of women over the management of these resources. Socio-political activities have also shifted ground: there is increasing recognition of the diversity of the women’s movement, especially as Dalit women have become more visible; uniform interests or strategies can no longer be assumed. There has been a loss of confidence in legal reform or legal awareness as a strategy for working with women, given problems of implementation and communal, class and other differences which have emerged in this area. There has been a move towards supporting and mobilising popular campaigns (e.g. the antiarrack movement described above) and other forms of social boycott. Due to the earlier efforts of the women’s movement, work in areas such as domestic violence is now considered legitimate development activity. The current focus on gender training as a strategy (see next section) has led to a proliferation of organisations offering training of different kinds to NGOs and government but there is some scepticism about the impact of such training as it is currently constituted. Some NGOs with a long record of work in rural communities are now seeking ways of increasing poor rural and particularly Dalit women’s access to land, forest resources and secure livelihoods. These programmes have emerged out of long-term interactions and discussions with women in particular rural localities. Their importance lies in the attempt to increase the asset base of poor rural women. This in an important area of work which requires further development and support. Overall, NGOs are increasingly seen, by both government and donor agencies, as alternative service deliverers to government as part of a wider agenda of decentralisation and liberalisation. Government departments in the social sectors are now directed to allocate a certain percentage of funds to NGOs. Given existing gender bias in access to social services, particularly, this raises questions about the implications of increased NGO involvement in service delivery. The underlying assumption that NGOs will be more efficient,
60

participatory and equitable service deliverers than state agencies needs to be questioned, particularly from a gender perspective. Long-standing concerns have been heightened among women’s organisations and in the NGO sector more generally about the implications of their greater role in service delivery in the contemporary Indian context. This implies a need for scaling up of activities and pressures to be more accountable to funders and possibly to meet particular targets to demonstrate efficiency, or, in this context, gender sensitivity. These pressures are perceived as potentially limiting the flexibility, participatory nature, capacity for innovation and critical role of NGOs generally, but specifically those working with women, where considerable investment in awareness raising and organisational work may be required. There is a question as to whether women’s NGOs, specifically, face particular constraints in scaling up activities. Alongside investment in enhancing the capacity of NGOs for expanded service delivery where appropriate, there is a need to provide mechanisms for continued support to smaller NGOs, particularly those working with women. One such approach may be the provision of decentralised innovation funds. Whilst NGOs, as well as government programmes (see above) are increasingly targeting women as beneficiaries, many still lack a gender perspective: there is a tendency by focusing on women as a broad category to homogenise women’s interests across communal, class, caste and other differences. There are relatively few NGOs working with men, alongside women, from an explicitly gender perspective. Crude targets such as increasing the numerical representation of women are insufficient indicators of gender-awareness. Mechanisms are also needed to ensure that poor, lower caste or minority women are adequately represented among beneficiaries. Encouragement is needed for the development of gender-aware approaches to working with men. Gender sensitivity is increasingly a requirement for donor support of NGOs. This raises questions about the ways in which funding agencies can assess, monitor and evaluate the gender awareness of NGOs and support the development of their capacity in this respect. Support for the development of gender-aware monitoring and evaluation methodologies and of institutional capacity of NGOs is important, so that NGOs do not perceive criteria for gender-sensitivity as externally imposed demands with little relevance to local conditions. Such methodologies need also to incorporate attention to the differences between women. With the exception of women’s organisations, women are under-represented in NGO work, and tend to be concentrated in field rather than leadership positions. Women in NGOs face a number of particular constraints due to the internal structures and institutional culture of NGOs and conflicts between their expected behaviour at work and in the domestic sphere. Efforts to increase the gender-awareness of NGOs need also to address these internal organisational issues and look towards the introduction of equal opportunities policies. There has been considerable growth in networks of NGOs and women working in NGOs in recent years. Latterly, this has been given a new impetus by networking efforts in preparation
61

for the UN Conference on Women in 1995, with NGO consultations occurring at local, state, regional and national levels. This momentum needs to be sustained beyond the immediate goal of working towards the Beijing conference. There is a need to build on existing networks but also to extend them across institutional boundaries, so that women’s NGOs or women working in NGOs also develop stronger links with women outside the NGO sector (e.g. researchers, women in local government, women in the administration). Links across the rural-urban divide and, specifically, between urbanbased women’s movement organisations and rural-based development organisations need to be encouraged. Dialogue between women’s movement organisations, gender specialists and technical experts in particular sectors, particularly in areas such as natural resources management and irrigated agriculture, could facilitate the development of a more clearly articulated gender perspective in these areas, where it is currently lacking. 6.5 Gender training As a methodology for making members of development agencies and government officials aware of gender issues, gender training emerged in India during the late 1980s and is currently in the process of being widely institutionalised. The goals and therefore the content and expected as well as actual outcomes of gender training vary widely, depending on the context. Gender training is carried out by a range of institutions and individuals and workshops or courses vary considerably in length, the most common being two to three day workshops. There is a broad distinction between approaches from a feminist perspective which emphasise issues of women’s oppression and self-empowerment and those with a more ‘professional’ focus which see gender training mainly as a means to improve project and programme design or service delivery. Whilst there is considerable use of western models of gender training, Indian models are also being developed. The recipients of gender training also vary widely, including members of women’s groups, grassroots organisers, government administrators, policy makers public servants and field workers, NGO personnel and, most recently, elected representatives under panchayati raj (see below). There is a broad distinction between training amongst women’s groups or for those working directly with them and training amongst development functionaries. It is particularly important that men are a significant proportion of those who receive gender training, although this may call for additional energies and skills on the part of trainers. Involving men in discussing their own needs and need for change makes it easier for them to address gender issues. A positive development in the gender training field is the growth in training of trainers, who can subsequently carry out ongoing training in their own organisations (see Box 8). Whilst, at an individual level, gender training is often considered beneficial, there are few systematic assessments of the impact of such training. A number of constraints limit the impact of gender training:
62

- There are few incentives for people to attend gender training courses and, particularly, to implement change subsequently; - Gender training is often limited in impact because of the curriculum adopted. In particular, ‘gender’ often means simply ‘women’; there is not enough attention to the issues which men have to address in order to change their behaviour. Also, caste, communal and other divisions are glossed over or not raised. The lack of a systematic analytical framework in many gender training courses undermines the value of the training; - Gender training may become an end in itself. Many training modules do not address the larger institutional or organisational context and either fail to consider or underestimate the institutional constraints for change.

Box 8: The India Gender Planning Training Project
The Gender Planning Training Project (GPTP) is a joint project of the Overseas Development Administration (ODA), U.K. and the Government of India (Department of Personnel and Training - DOPT) aimed at institutionalising gender training into administrative training institutions (ATIs). The project is based on a ‘training of trainers’ approach. The project covers five states: Uttar Pradesh, Maharashtra, Rajasthan, Karnataka and Madhya Pradesh and is based on a three-way linkage between the state ATI, an NGO and a government department linked to a theme selected by each state for each year of the project. The training methodology used was developed at the Institute of Development Studies, through the ‘Women, Men and Development’ course. This methodology was identified as appropriate for institutionalising training for planners and implementors in the state administrative system. The project aims to adapt the analytical framework of this methodology to selected institutions and trainers. The GPTP is now in its second year and its outputs to date have been: 40 trainers from state training institutions, national sectoral institutions, NGOs, academic institutions and government departments. Ten training modules have also been developed on gender issues in relation to: violence, literacy, cooperatives, health, anti-poverty programmes, girls’ education, forestry, entrepreneurship development, Panchayati Raj as well as a general module on gender and development. (Ramya Subramanian, personal communication).

To strengthen gender training programmes, gender training needs to be mainstreamed within managerial and administrative training. Gender training programmes need to build in institutional incentives (not necessarily financial) to encourage institutions to invest more seriously in training and follow-up. Gender training needs to be situated in a broader context and connected to other axes of inequality.

63

There is a need for a better treatment of the institutional constraints under which development functionaries operate, which will vary between different sectors, levels and organisations, and for follow-up. Gender training must be systematic and analytically coherent. It must include follow up and be sustainable. It must support the project of transforming practice. In order to mainstream gender issues, the stake-holders in development agencies must be addressed as key agents. People at the lower end of government hierarchy, in particular, need to be encouraged to ‘subvert’ the system. Finally, gender training must address male needs but without losing sight of the politics of gender. 6.6 Women’s participation in Panchayati Raj The system of panchayati raj, or local self-government at the village level, was introduced prior to Independence and incorporated into the Independence constitution. However, for forty years, the system had been fairly ineffectual, due to stalling of elections and lack of resources. In 1989, attempts to revive and reform the system began and a number of proposals were made to Parliament which, for various reasons, were not carried through. At this time, the issue of reservations for women within panchayati raj was also widely debated in a number of nationwide conventions. There was considerable controversy about the need for reservations for women; however, existing women members of panchayati raj - both elected and nominated within the prevailing system - emphasized the need for an improved reservation policy. In 1993, after a long process, the Constitution 73rd Amendment Act was passed at both central and state levels, coming into force in April of that year. The basic features of the Act include: the constitution of local government (panchayat) bodies in a three tier system in most states; extension of the existing reservations system to include one third minimum representation of women, including among chairpersons; direct election of panchayat members; and increased resources and powers for local government councils. The attention currently being given to panchayati raj is linked to the need for decentralisation within the overall policy shift under structural adjustment, where the role of the central state is being reduced. It is hoped that panchayats will prove be more cost-effective and efficient in their delivery of services. However, the revival of a panchayati system, especially one that grants greater powers to women, hinges crucially on the devolution of power from existing vested groups, whether it be the state bureaucracy or local lobbies. Thus, the actual outcome of the decentralisation process remains uncertain. There is likely to be considerable variation in the extent to which a genuine process of decentralisation occurs, depending on the extent to which particular state governments are willing to devolve power and resources, the nature of local level political interests and processes and the general state of civil society in particular regions. In some states, there has already been considerable stalling of the election process.

64

These will be constraining factors on the extent to which women are able to represent their interests in local self-government. More than 100,000 women will be elected onto panchayati raj institutions throughout the country, including several thousand chairpersons. In the past, very few women stood for elections; the nomination system for women operated on the basis of patronage by the dominant political or social group and led to more or less tokenistic representation. Some states, like Andhra Pradesh and Karnataka, introduced a reservation policy of 22-25 percent in the mid-eighties but this was seen to be inadequate. Even though these reservations were based on an electoral process, those elected were largely women relatives of the sarpanches (panchayat leaders) or other influential members. There is now a growing literature on women’s existing and potential participation in the panchayat system. A special information and resource centre called UMA (Utsahi Mahila Abhyudaya) has been set up in Bangalore, dealing exclusively with women and panchayati raj, which is a branch of the Institute of Social Studies Trust (ISST) in New Delhi. UMA has conducted workshops and seminars and now brings out a regular newsletter on the subject of women and panchayati raj. A number of NGOs are also actively involving themselves in promoting panchayati raj. Government journals such as Yojana or the NIRD journal are also bringing out regular articles promoting women’s participation in local self-government. The promotion of women in panchayati raj is seen as a means to address women’s subordinate position in rural society and their lack of political power. But there is an equal if not greater emphasis on the gains that are expected to accrue to society by having more women as political representatives in local bodies. In particular, women’s participation in panchayati raj is expected to bring about: - improved implementation of development schemes for women and other sectoral programmes in e.g. education, health and sanitation, horticulture etc.; - a more active role of women in the mediation of disputes, particularly at familial level; - a fresh and incorrupt input into local government. However, most reviews on the position of women in the panchayats to date emphasize the difficulties women encounter in standing for elections and in carrying out their functions once elected, because of prevailing social norms. Unrealistic expectations may be being raised of women members, particularly when considered in the light of past experience and of the constraints mentioned above. Moreover, there is perhaps insufficient attention to the personal economic and social costs of women’s participation. Some are more sceptical about the reservations policy. While reservations may be important in the short-term for boosting women’s participation in local government they cannot be a substitute for long-term affirmative action or equal opportunities policies which are

65

transformatory in terms of enhancing women’s access to positions of power and resources across the board. In spite of remaining controversy about reservations policy, and constraints to a genuine decentralisation process, there is unanimity that women, both as candidates and representatives, require support and training. It will be important to ensure that such training is provided for women from all communities, particularly Dalit, tribal or Muslim women. Moreover, continued efforts are required to increase women’s participation through the general pool of candidates as well as in reserved seats. One area of possible intervention is in strengthening links between NGOs and women in local government. Such alliances have proved effective in other countries, e.g. Bangladesh. NGOs have traditionally by-passed local government, seeing it as corrupt or ineffectual. Alliances between women within local government institutions and NGOs working in those localities could provide women in panchayati raj with a counterweight to vested interests dominating local politics and a means to co-ordinate more effectively efforts for change as well as service provision at local level.

66

ANNEX 1: TERMS OF REFERENCE PROPOSED INDIA - GENDER ISSUES BACKGROUND PAPER COMMISSIONED FROM BRIDGE BY ODA

1.

WORKPLAN

Following a preliminary meeting with Michael Schultz of ODA on 23 December 1993, BRIDGE (IDS) agreed to present a workplan and budget for the proposed India - gender background paper to feed into the development of ODA’s WID strategy for India. At this meeting, it was broadly agreed that: a) BRIDGE would be the overall manager of production of the report, but BRIDGE/IDS staff would work in collaboration with an in-country researcher/institution. Some of the work will be done in India and some at IDS. the Indian collaborator would spend time at IDS during the period of finalisation of the report;

b) (i) Learning from the experience of the Ghana report, and considering the volume of material already in existence on India, the paper on India will not give a comprehensive descriptive analysis of women’s situation in India. The report will summarise conventional wisdom and aggregate data only where appropriate and concentrate instead on highlighting new issues and ideas drawing on the findings of recent gender research. The report will be strongly forward looking. (ii) As far as possible, the content of the report should relate to areas of current ODA activity (both geographically and sectorally) but this will not be to the exclusion of other possible areas of intervention. (iii) The report will be structured around key issues, with scope for cross country comparisons, drawing on the wider gender and development literature and developing innovative ideas. Some possible themes/questions were discussed at the meeting, for example: - the current and, in India, likely future social/gender impact of structural adjustment; - gender biases in access to social services (especially in education and health sectors); - developments in the NGO sector (particularly e.g. in relation to legal literacy, violence against women etc.) - the extent to which gender training/planning have led to ‘mainstreaming’ of gender concerns in development interventions;

the degree of influence of women’s organisations over the policy process. Ideas will be further elaborated by BRIDGE and submitted to ODA in the form of a draft outline of contents of the report. c) There will be close and ongoing consultation between ODA and IDS and the Indian researcher during the preparation of the report. Interim outputs will be circulated (e.g. drafts of particular sections of the report). Concretely, this means that meetings (at least three in the UK) will be held during the course of preparing the report. (Appropriate points at which meetings might be held between ODA and BRIDGE representatives are: detailed outline; first draft; revised draft. Consultative meeting(s) of the in-country collaborator with ODA representatives in India during the in-country research phase would also be valuable.)

d) Given the scope of the exercise the whole process would take at least three months, to start as soon as possible. Much of the actual research and initial drafting will be done by the Indian collaborator, with BRIDGE/IDS providing mainly advisory, supervisory and editorial inputs. However, the precise division of work between UK/India and between IDS personnel/Indian researcher will depend on: the thematic structure of the report (see (b) above); the particular areas of expertise/knowledge of the collaborator relative to those of staff at IDS working on the assignment; access to literature in country relative to that at IDS.