Hands Not Land Book

Hands Not Land:
How Livelihoods are Changing in Rural Bangladesh
Kazi Ali Toufique Cate Turton
The Bangladesh Institute of Development Studies (BIDS) is a multi-disciplinary organisation which conducts policy-oriented research on development issues mainly in the context of Bangladesh.
The Department for International Development (DFID) is the British government department responsible for promoting development and the reduction of poverty.The government, first elected in May 1997, increased its commitment to development by strengthening the department and increasing its budget.
Hands Not Land:
Kazi Ali Toufique Cate Turton
Kazi Ali Toufique, Research Fellow, Bangladesh Institute of Development Studies, Dhaka, Bangladesh and Cate Turton, Senior Consultant, the IDL group, PO Box 20, Crewkerne, Somerset, UK. ISBN 1 86192 492 5 This publication was funded by the United Kingdoms Department for International Development. However, the Department can accept no responsibility for the information provided herein. Cover photograph: Gabriele Till Editing: Omar Sattaur Design and typesetting: Paul Samat Printing:
Contents
Acknowledgements Contributors Preface Summary Part 1: Poverty in a changing rural landscape Chapter 1 Chapter 2 Chapter 3 Part 2: Original research papers Livelihoods in contemporary rural Bangladesh: an overview of research papers Rural development trends: what the statistics say Bimal Kumar Saha Agricultural and non-agricultural livelihoods in rural Bangladesh: a relationship in flux Kazi Ali Toufique Frontiers of change in rural Bangladesh: natural resources and sustainable livelihoods Mohammed Asaduzzaman Rural development policy in Bangladesh Sattar Mandal Rural poverty: patterns, processes and policies Rushidan Islam Rahman Migration and rural livelihoods Rita Afsar The informal institutional framework in rural Bangladesh S.Aminul Islam The formal institutional framework of rural livelihoods in Bangladesh Paul Thornton Re-thinking local governance: towards a livelihoods focus Hossain Zillur Rahman Gender dimensions of rural change Santi Rozario The causes of vulnerability in rural livelihoods S.Aminul Islam 7 9 11 13 15 21 33
39 47
57
65 71
79
89
97
105
113
121
131
Acknowledgements
The analysis and thinking contained in this book represents the sum of the wisdom and experience of a large number of people who contributed their time to help us with the field work, prepare papers and share their experiences and perspectives on the key livelihood issues and challenges in rural Bangladesh. First and foremost we would like to thank our team members in the Department for International Development (DFID) livelihoods study: Paul Thornton and Aminul Islam with whom we spent many long and productive hours building a picture of the key livelihoods issues. Stimulating and challenging discussions with Binayak Sen, Hossain Zillur Rahman and Sattar Mandal further developed our understanding of the political, economic and institutional context, which shapes the livelihoods of the poor.The hard work and efforts of the field teams are much appreciated.Their efforts ensured that this book remains grounded in the reality of poor peoples struggles to sustain a livelihood. Lastly, we would like to acknowledge the support of DFID Bangladesh Advisers who actively participated at different points in the process helping us to develop a holistic picture of livelihoods issues. Donal Brown and Tim Robertson deserve special thanks for their contributions and support.
Contributors
Rita Afsar, Research Fellow, Bangladesh Institute of Development Studies, Dhaka, Bangladesh. Mohammed Asaduzzaman, Research Director and Director General (a.i.), Bangladesh Institute of Development Studies, Dhaka, Bangladesh. S.Aminul Islam, Professor, Department of Sociology, University of Dhaka, Dhaka, Bangladesh. M.A. Sattar Mandal, Professor, Department of Agricultural Economics, Bangladesh Agricultural University, Mymensingh, Bangladesh. Hossain Zillur Rahman, Senior Research Fellow, Bangladesh Institute of Development Studies and Executive Director, Power and Participation Research Centre, Dhaka, Bangladesh Rushidan Islam Rahman, Research Director, Bangladesh Institute of Development Studies, Dhaka, Bangladesh. Santi Rozario, Senior Lecturer, Department of Sociology and Anthropology, University of Newcastle, New South Wales,Australia. Bimal Kumar Saha, Research Fellow, Bangladesh Institute of Development Studies, Dhaka, Bangladesh. Paul Thornton, Director,Verulam Associates Ltd., South Asia Office, Dhaka, Bangladesh. Kazi Ali Toufique, Research Fellow, Bangladesh Institute of Development Studies, Dhaka, Bangladesh.
Preface
Over the last 50 years, there has been no dearth of models of development: industrialisation; central planning; basic needs; area development; gender dimensions; structural reform; sustainable development; sustainable human development, etc. They first came as new fads and then like old rugs were thrown into a corner. A flaw in most of them has been that they overlooked one basic fact. Even if development action is well-meaning and intended to benefit a particular group of people, unless the poor themselves are involved in the process of influencing its content and form, they will not take much interest in it. Poor people will take an interest in the development process when they find that it responds to their concrete situation, which only they experience and, therefore, which they can articulate best. Unless their voices are heard and the context in which they struggle to make a livelihood is correctly understood, any development action will, at best, be only partially successful. One thing is clear. We need to understand better how the poor live, the opportunities they may have and constraints they face. Otherwise, progress in eradicating poverty will continue to be slow. The Sustainable Livelihoods approach is one approach that can help us to better understand the opportunities and constraints of the poor taking in to account the economic, social and political context. This book provides some new ideas for development practitioners on how to approach the challenge of the eradication of poverty in Bangladesh. Its origins lie in a study of rural livelihoods commissioned in 2000-2001 by DFID UK. The study comprised included fieldwork in several rural areas of the country and the preparation of several background papers on various aspects of rural development and the links to various formal and informal processes within the country. A large number of researchers from the Bangladesh Institute of Development Studies (BIDS) participated actively in the process contributing a number of papers and participating in the ensuing debates and discussions on rural development issues.The focus of the debate at all times was on the realities of the poor and how they relate to the dynamic economic, social and political framework within which they struggle to sustain a livelihood. It was not an academic debate of the Sustainable Livelihoods approach. For nearly four decades, BIDS has been a place to think, write and debate on Bangladeshs development policies and issues. In this process, BIDSs has actively cooperated with many development agencies both local and international.This book is the fruit of one such collaboration with DFID. I thank all my colleagues who have been involved in the process and most of all Kazi Ali Toufique and Cate Turton who have laboured hard to make this book a reality. I thank DFID Bangladesh and in particular, Donal Brown, for their support and help in ensuring that findings of this study can be shared as widely as possible. Mohammed Asaduzzaman Research Director and Director General (a.i.), Bangladesh Institute of Development Studies, Dhaka, May 19, 2002
11
Summary
The phrase rural Bangladesh no longer means what it once did.We believe it to be out of date as the distinction between urban and rural life is no longer clear cut. Powerful external economic forces, including those of globalisation and the expansion of physical infrastructure especially roads and bridges, rural electrification and the growth of marketing outlets are creating a rural landscape that is increasingly urban in character and have radically transformed village life. New livelihood opportunities are emerging often in the non-farm sector. The numbers of small shops, tailoring and other craft enterprises, rickshaw pullers, petty traders in villages and local bazaar centres have grown substantially. Remittances now form a critical part of the rural economy. However, change is happening faster in some places than in others and for some people more than for others.We see a continuum rather than a divide from areas where traditional views and images still hold true to areas where a more modern picture is taking hold. Some people, too, have been unable to embrace change and the new opportunities it brings. For many of the poor, who have little or no access to land, their primary asset remains their labour a healthy pair of hands is critical to their livelihoods. But whether they are engaged in agricultural labouring or in the non-farm sector they continue to be marginalised from the development process. This assessment of livelihoods is the result of applying a range of tools and methodologies. More important than examining these in detail, and precisely how they were used, is to appreciate the spirit of open inquiry in which this assessment was conducted. We wished to answer the question: what are life and livelihoods like today for men and women in rural Bangladesh? We feel that our findings offer an up-to-date and realistic picture of life in a country that is in the throes of tremendous change and have implications for anyone concerned with livelihoods and the poor in modern Bangladesh.
Hands Not Land:
Part 1: Poverty in a changing rural landscape
15
Chapter 1
POVERTY IN BANGLADESH
Approximately half of the population of Bangladesh lives below the poverty line1. This amounts to 63 million people engaged in a daily, never ending struggle to meet their basic subsistence needs. For half of these 63 million people, life is even more difficult: they live in extreme poverty often without land or a homestead, without a source of regular income and in households that are disadvantaged by being headed by a female or that have disabled or ill members (World Bank 2002). The story does not end here. For one-fifth of the population approximately another 25 million of Bangladeshs people poverty may be just around the corner. Known as tomorrows poor, their tenuous grip on a decent livelihood is threatened by shocks and stresses which they are unable to manage; the illness of a family breadwinner, the loss of land due to erosion, or an increase in food prices can be enough to push families below the poverty line. Bangladesh experienced a modest decline in poverty indicators during the 1990s. The most favourable estimates put that decline at around one percent per year (World Bank 2002). Despite the undeniable achievements, however, the improvement in percentage indicators should not obscure the fact that due to population growth rates the number of people living below the poverty line continues to increase. At the current net rates of poverty reduction, it will be another 50 years before poverty is eliminated in Bangladesh. The challenge Bangladesh faces today is to scale up the rate of poverty reduction.
The improvement in poverty indicators is something to build on and understanding the dimensions of these positive trends is important.They can tell us where progress is being made. They can also help us to identify which groups of people are managing to improve their livelihoods and which groups continue to be marginalised. The trends in poverty indicators tell us that:
q
greater progress has been made in reducing human poverty2 than in reducing income poverty3. There has been significant improvement in some social indicators such as primary education enrolment and fertility rates; certain manifestations of extreme poverty have declined significantly. For example, more people have access to basic clothing and housing and the proportion of the population having no meals, or fewer than one meal a day in certain seasons, has fallen; inequality in income distribution is increasing; the extreme poor have less access to the benefits of economic growth; there has been little change in the gender dimensions of poverty. Femaleheaded households are still more likely to live in poverty and females within households are still more likely to be less well-educated, more likely to be malnourished and more likely to fall ill.
Poverty trends also tell us that improvements in the livelihoods of the rural poor have lagged behind those of their urban counterparts; poverty is not uniform throughout the country:
16 Hands not Land: How Livelihoods are Changing in Rural Bangladesh
80 percent of the poor live in rural areas.The rate of extreme poverty there remains twice as high as in urban areas; poverty is geographically stratified; the most deprived districts are found in northern Bangladesh.
THE STORIES BEHIND THE STATISTICS

It is established practice to disaggregate poverty statistics into rural and urban categories. And Bangladesh has been no exception to this. There is no shortage of development narratives or prescriptions to accompany this picture. The point of departure then becomes the rural poor and the established views and perceptions on the way the rural poor, as a definable entity, live. Traditional stories Mention of the rural poor in any development context conjures up images of village communities, largely self-contained and directly or indirectly dependent on natural resources for their livelihoods.The picture is of farmers and artisans eking out a living from less and less land, unable to reach markets easily because of poor roads and communications, little transport, poor education and few facilities either for recreation or business. For Bangladesh it is no different. Issues such as access to land, agricultural growth and the achievement of food self-sufficiency have been central to the rural development debate since the countrys independence. At around 800 people per square kilometre, Bangladeshs population density is the highest among the worlds non-industrialised nation states and access to land is a major limiting factor.There has been longstanding concern over the inequities of land ownership: 56 percent of rural households are classified as functionally landless that is they own less than 0.2 ha of land. The fact that the proportion of landless people continues to increase is seen by many commentators as cause for concern. Agricultural development, mostly focused on boosting rice production, has been the cornerstone of government rural development policy. Food-grain production has
Bangladesh achieved good economic growth during the 1990s. Between 19912000, real gross domestic product (GDP) increased by 60 percent, averaging a growth rate of five percent per year (World Bank 2002).The fact that this has failed to translate into accompanying rates of poverty reduction and that large sections of the population notably the rural poor have not shared in the fruits of the development process is of serious concern. It is clear that relying on trickledown mechanisms alone will not be enough to bring about the reductions in poverty that are needed: the challenge facing Bangladesh today is not so much the redistribution of wealth but the redistribution of the new livelihood opportunities that are emerging. Understanding livelihoods Livelihoods approaches are based on holistic multi-disciplinary analyses of poor peoples livelihoods.They explore: the vulnerability context in which people live, the assets they have access to, the strategies they adopt, the policies and institutions that shape their decisions and the outcomes they aspire to. Ensuring that the poor are able to take advantage of new opportunities means we need a better understanding of who the rural poor are, where they live and what barriers and constraints they face in taking up new livelihood opportunities. This kind of understanding relies, in turn, on an approach that addresses the multidimensional nature of poverty and necessitates investigation and analysis of the major issues and trends affecting rural livelihoods.
Poverty in Bangladesh 17
nearly doubled since 1970 and food-grain self-sufficiency is now well within reach. Conventional wisdom holds that dynamic agricultural growth is a necessary, but not sufficient, condition for development. The understanding is that growth in nonfarm opportunities is strongly linked to and arises from, strong agricultural growth. The question that this book seeks to address is whether the traditional images, concerns and issues that have underpinned approaches to rural development in Bangladesh truly reflect the reality of poor peoples livelihoods today. Have conventional approaches to reducing poverty addressed the concerns that are important to the poor, and do they recognise the opportunities they have to build a sustainable livelihood? This book explores the present-day rural reality and looks at how todays rural population makes a livelihood. Is there a need for a new narrative? A quick look around a rural Bangladeshi village is enough to tell a story of how rural life is transforming. Physical infrastructure
is expanding, more and more villages enjoy rural electricity and marketing and communication channels are continuing to improve. The changes have brought new livelihood opportunities for some: increasingly, rural villagers are just as likely to be employed as rickshaw pullers as they are agricultural labourers, and will travel further afield in search of a livelihood. Todays rural landscape is a far cry from yesterdays picture of rural communities in remote country villages. In short, contemporary rural reality is increasingly at odds with traditional views on rural livelihoods. National statistics reflect this transformation. The share of agriculture in GDP declined from 32 percent in 1981 to 25 percent in 2000 (Shahabuddin and Quasem 2002). And looking more closely at the statistics we find evidence of some interesting stories to be told. Table 1 shows that, nationwide, the number of landless households (those without land for a homestead) is decreasing at roughly three percent per annum.At the same time, functional landlessness
Table 1 Trends in landlessness, 1983-96

Category of households/landless 1983-84 Number of households % 1996 Number of households % Growth (%) per annum
Landless: category I 276,977 Landless: category II 2,713,969 Landless: category III 3,898,181 Near landless: category IV (marginal) 1,702,652 Remaining households: category V 5,225,867 All 13,817,646 Source: BBS (1986, 1999)
2.00 19.64 28.21 12.32 37.82 100.00
162,229 5,003,042 5,191,979 2,494,606 4,976,326 17,828,182
0.91 28.06 29.12 13.99 27.91 100.00
-2.93 5.23 2.42 3.23 -0.39 2.15
Note: Category I: households without homestead land. Category II: households with homestead, but without cultivated land. Category III: households with homestead and cultivated land (up to 0.2 ha). Category IV: households with homestead and cultivated land (0.2-0.4 ha). Category V: rest of the households with homestead and cultivated land (0.4 ha and above).
the number of households with homestead land but without cultivable land and/or with cultivable land up to 0.2 ha is increasing sharply, at the rate of 5.23 and 2.42 percent per annum respectively. On the one hand, these figures give some cause for hope.They suggest that extreme landlessness (poverty) is decreasing. But, they also leave us with some questions. Whilst for some the increasing rate of landless is a cause for concern a more pertinent question to ask may be how landless households are sustaining their livelihoods? This book aims to explore the livelihood stories behind these statistics. It aims to reveal how the rural poor make a living in the context of the far reaching changes suggested by national statistics.
The idea was not to provide a deep academic analysis of rural livelihoods in Bangladesh. Rather than building a comprehensive and representational picture of livelihoods through a structured exploration of different components (for example the vulnerability context, peoples assets, their strategies), the assessment focused on trends, broad directions of change and new opportunities and threats that were emerging as a result of the changing rural context. Above all, the assessment attempted to build a picture of the realities of livelihoods from the bottom up, with field realities dictating the direction of the assessment at all times (see Key features of the approach, below). Most importantly, we purposely set aside our preconceptions and began the assessment with a blank sheet of paper. The team asked of rural communities questions like: What has
A NEW RURAL DEVELOPMENT NARRATIVE

The approach The conclusions drawn in this book, concerning modern rural livelihoods in Bangladesh, are based on an assessment of livelihoods carried out by a multidisciplinary team for DFID UK.The picture it paints is the result of using different tools and methodologies. More important than the tools and their application, however, is to appreciate the spirit of open inquiry in which the assessment was conducted.We wished to answer the question: what are life and livelihoods like today for men and women in rural Bangladesh? The fundamental questions always explored the change people saw happening in their own lives, the changes they expected and what they perceived as the livelihood implications of those changes.
Key features of the approach

The focus was on rural livelihoods not natural resources what is happening to livelihoods in rural Bangladesh not natural resources and the role they play in livelihoods. It focused on the dynamics of change not on compiling a detailed audit of livelihoods. It did not focus narrowly on rich and poor but looked at who wins and loses in the emerging livelihood scenario. It focused not on individual livelihood gains and losses but on aggregate livelihood gains for rural communities and how they were distributed. It integrated the institutional perspective into the process it began at the local level and worked upward and outwards seeking to understand how policy and institutional processes manifest at the local level.
Poverty in Bangladesh 19
changed in the last 10 years? How do you see the future evolving? What challenges/opportunities do you have today? And so on. From here, we developed some ideas on how rural livelihoods were changing. There was a particular focus on understanding who is winning and who is losing in the fast-changing rural scenario. Two approaches were used to substantiate and validate our ideas. Firstly, more detailed fieldwork was carried out in four locations selected to represent a cross section of Bangladeshs rural landscapes.The fieldwork aimed to provide more detail on the ways in which rural livelihoods are changing. In order to validate findings emerging from the fieldwork, we commissioned national experts to write papers on cross-cutting themes and rural issues in order to see if the stories emerging from our rapid field assessment were consistent with the current status of knowledge, and supported by broader statistical trends. Of central importance was that the assessment attempted to link micro-evidence with macro-evidence and strived to make the connections between the lessons emerging from the fieldwork, the macrolevel data and the current state of knowledge derived from the key issues papers.
the livelihoods of the rural poor. The summary draws on individual insights provided by the papers and the fieldwork. The first part concludes with a postscript, which looks at questions posed by the new rural reality and assesses the implications of this books key messages for the way in which rural development is approached. It also gives an honest assessment of how far we think we have reached in generating a new understanding and identifies some of the missing links in the picture. The second part of this book provides an introductory overview to the papers and is followed by a complete version of each of the research papers.
NOTES
1 A minimum requirement of welfare, usually defined in relation to income or expenditure, used to identify the poor. Individuals or households with incomes or expenditure below the poverty line are poor. Those with incomes or expenditure equal to or above the line are not poor. It is common practice to draw more than one poverty line to distinguish different categories of poor, for example, the extreme poor. From (DFID 2001). 2 Poverty defined with respect to factors of human capital such as access to education, health outcomes, sense of well-being; empowerment; and so on. 3 Poverty defined with respect to a money-based poverty line for income or expenditure.The distinction is made between this and other concepts that emphasise the many dimensions of poverty. From (DFID 2001).
STRUCTURE OF THIS BOOK

Drawing on the different elements of the assessment, this book attempts to paint a picture of what life is like for the rural poor in a country that is rapidly changing. This book is split into two parts.The first part summarises the main livelihood themes and trends emerging. It identifies some key livelihood headlines and speculates on the implications of these for the future rural scenario and in particular for
21
Chapter 2 What is life like for the poor in rural Bangladesh?

The landscape of Bangladesh has changed dramatically over the last decade. Infrastructural development has significantly changed the physical landscape in rural areas. In response to improvements in markets, communications and transport, people have forged new types of rural livelihood and, in doing so, have changed the human landscape too. This chapter looks first at how the livelihoods of rural people have changed. It then goes on to explore which groups of people are able to take advantage of the new opportunities and which are not. The chapter is based on insights from the authors own fieldwork, the findings of the commissioned papers and the outputs of the fieldwork case studies. rural areas are increasingly urban in appearance. Livelihoods are adapting to take advantage of the new opportunities afforded by improved infrastructure and communication. Many village households are now sufficiently connected to district headquarters, towns or bigger cities, to have created rural livelihoods that are dependent on incomes derived in urban areas. Villages can therefore no longer be viewed as physically isolated or economically discrete communities. The rural economy is in transition Appearances are not deceptive! Villages are not only more urban in appearance, but are more urban in character as well. The role that rural areas play in the national economy is changing. Rural areas no longer just serve a food-production function but are also important sources of labour for urban areas. Agriculture has traditionally formed the heart of rural livelihoods. However, this is changing fast. Nationally, agriculture was the slowest-growing sector during the 1990s and, overall, declined in importance from 29 percent to 25 percent of GDP. In contrast, over the same period, the industrial sector increased in importance from 21 percent to 26 percent of GDP, with real GDP in this sector increasing by an impressive 86 percent (World Bank 2002). Despite its poor performance relative to other sectors, the agriculture sector did still continue to grow, if somewhat erratically. However, the type of growth has been different from the rice-led growth of earlier times. Crop diversification, farm
LIVELIHOOD HEADLINES
If an observer were faced with the challenge of writing an article on the changing face of rural Bangladesh, what would the headlines be? A new rural-urban continuum A clear cut rural-urban divide no longer exists in Bangladesh. A traveller from town to village would find it much harder today to say precisely where the landscape changes from urban to rural. What does this really mean? The most prominent change is in physical infrastructure.There are more roads and bridges, concrete is replacing the traditional building materials of wood, earth and straw. Electrification has penetrated deeper into rural areas and new businesses and marketing outlets for consumer goods and services are proliferating. The overall result is that
mechanisation (notably the expansion in the use of power tillers for land preparation) and the exploitation of new ecological, technical and economic niches (such as vegetable production and integrated fish-rice production) contributed most to economic growth in the crop sector. The livestock and fisheries sub-sectors have been particularly vibrant, despite the fact that there was a decline both in access to, and status of, common property resources, particularly aquatic and fisheries resources. Toufique (see Agricultural and non-agricultural livelihoods in rural Bangladesh: a relationship in flux, pg57) argues, however, that the nature of agricultural growth has not been labour-generating. This raises questions which we return to later in the chapter over who has benefited from agricultural growth. The classic argument in rural development is that growth in the non-agricultural sector is closely linked to growth in the agricultural sector. Firstly, incremental income from agricultural growth is spent on goods produced by the non-agricultural sector (the consumption link). Secondly, agricultural growth requires inputs and farm implements produced and maintained by the non-agricultural sector (the production link). Thirdly, agricultural produce is processed by the non-agricultural sector (the backward link). But here lies the paradox that Toufique (op. cit.) claims has been hitherto under-emphasised altogether in the context of rural development in Bangladesh: Bangladesh has witnessed spectacular growth in the rural non-agricultural sector during a time when the agricultural sector grew slowly. The missing link comes in the form of migration, urbanisation, infrastructural growth, and the impact of globalisation. These developments and processes are creating new ways of earning a living.Whereas all rural livelihoods were
at one time directly or indirectly related to agriculture, the changes that have swept through rural Bangladesh have spawned new livelihoods that are completely independent of agricultural activities.Today, rural people are migrating to towns and cities for work connected with the creation and maintenance of new infrastructure, or with trading that is linked to national and international markets, even though they may live in villages. Divergence in the relationship between the two sectors at the macro-level, however, should not obscure the fact that important relationships remain at the micro-level. The livelihood strategies of many of the rural poor continue to straddle both agricultural and non-agricultural activities. The critical conclusion to be drawn is thus not a sectoral one per se, but of the need to adopt a holistic view of the local economy and its changing mix of livelihood opportunities (see Re-thinking local governance: towards a livelihoods focus, pg113). Diversification is the name of the game The traditional image of the peasant farmer sitting at the centre of the rural economy has long disappeared from much of rural Bangladesh.The reality is that rural households Multiple livelihoods are as likely to be involved In Lalmai village, only in non-agricultural liveliabout 16 percent of hoods as they are in farmhouseholds depend ing and, increasingly, they entirely on agriculture. derive incomes from mulTen percent of households tiple sources. The greatest rely on agriculture and expansion has been in the service; eight percent on services sector. The numagriculture and business; ber of small shops in and 66 percent on villages has increased subsharecropping, wage stantially, as have tailoring labour, shopkeeping and other craft enterprises, and other activities rickshaw pulling and petty (Islam 2001).
What is life like for the poor in rural Bangladesh? 23
trading in villages and local bazaars.We are seeing the growth of rural-urban centres1, the growth of rural industries and the rural transport and services sector. People wishing to take advantage of these new livelihood opportunities must draw heavily on a range of assets: human, social and financial assets are fast becoming as important as natural assets (such as access to land and water) once were. Access to these assets enhances the capacity of households to shift from one livelihood to another or to combine livelihood strategies. Macro-level data add further to our understanding of the relative importance of these new livelihood opportunities. Saha (see Rural development trends: what the statistics say, pg47) showed that the share of rural household incomes derived from agriculture decreased from about 63 percent in 1987 to 54 percent in 1994, while the share derived from non-agricultural activities increased from 37 percent to about 48 percent during this period2. The percentage of population economically active in agriculture also decreased from about 66 percent to 55 percent and increased from 35 percent to 45 percent in non-agricultural activities over the same period. These figures should not disguise the fact that, for many people today, livelihood strategies comprise both agricultural and non-agricultural contributions. A crucial issue is the quality of livelihood diversification and not just livelihood diversification per se. Both Toufique (op. cit.) and Saha (op. cit.) recognise that the growth of opportunities in the nonagricultural sector occurred primarily at the lower end of the occupational scale. Zillur Rahman (see Re-thinking local governance: towards a livelihoods focus, pg113) notes the trend towards an occupational hierarchy for the poor in which casual daily labour is the least preferred employ-
ment. The competition is for piece-rate labour contracts and fixed-rent tenancies in the farm sector and for non-farm employment in rural construction activities, transport operations and at the lower end of trade and service activities. For many of the rural poor, livelihood diversification has been distress-driven3. Diversification improves poor peoples ability to manage risk and spread income and consumption more smoothly across seasons. Saha (op. cit.), for instance, notes that the landless are involved in non-agricultural work not only to increase their income, but also to reduce sharp fluctuations in income over the whole year. He claims that the sectoral shifts of the labour force out of agriculture are largely [a result] of the adoption of survival strategies by the landless and land-poor for getting out of the poverty trap, thus offering them little scope to improve their well-being.We return to this point later in the chapter. Rural populations on the move Migration and mobility are of critical importance in constructing livelihood strategies in modern rural Bangladesh. Patterns of mobility are diverse and complex. They range widely, from commuting at one end of the spectrum, to temporary absence from the home for a couple of days to several years, to regular seasonal migration or permanent relocation (see Migration and rural livelihoods, pg89).
q
Approximately three million Bangladeshi nationals were employed abroad between 1976-99, all of them temporary migrants. Migrating in order to secure seasonal agricultural work is an important livelihood strategy for many poor men and women in the more deprived areas of Bangladesh.
Daily commuting from village to urban centres, upazilla and district headquarters for work is a growing phenomenon.
low sex ratios. From the mid-1980s, there has been a dramatic decline in the sex ratios of all metropolitan cities. A major factor in the decline of the sex ratio in urban areas has been the demand for cheap usually female labour generated by the RMG industry. Existing evidence suggests that the RMG industry alone absorbed 1.5 million workers during the last decade of which more than 90 percent of them were migrants from rural areas (see Migration and rural livelihoods, pg89). This has had a tremendous impact on livelihood opportunities for poor rural women. Large numbers of young women, usually from poor rural families are independently migrating to the capital city to secure a livelihood a livelihood option that was hitherto considered strictly male. Although most would agree that increased migration opportunities have had a positive impact on livelihoods, there are some caveats to this. Firstly, the option of adopting migration as a livelihood strategy is not open to everybody. Rushidan Rahman (see Rural poverty: patterns, processes and policies, pg79) notes that, in the case of international migration:
q
Quite how important the livelihood strategy of migration has become in rural Bangladesh can be illustrated by the fieldwork findings. In many places, income and remittances from migration have replaced agriculture as the major source of household income. Our fieldwork showed that, for some villages, more than 80 percent of income was derived from outside the village. The phenomenon of multi-locational households in which household members may temporarily reside away from the village in order to secure desirable work is becoming common. Rural populations are mobile, but who in the household is actually moving, and what are the implications for rural livelihoods? Afsar (see Migration and rural livelihoods, pg89) has found that:
q q
new migrants are usually young adults; migrant labour in the ready-made garment (RMG) sector is even younger, since more than 90 percent of those employed are under 30; overseas migrants are predominantly male and above the age of 30.Twentyfive percent of these workers are over 40.
The traditional picture of young males leaving their villages to find work to support their families back home is changing as more and more women join their ranks.The scale of this phenomenon is illustrated by the changes in the sex ratios of sending and receiving areas. Areas with more men than women are conventionally considered as receiving areas, and have a high male to female sex ratio and those with more women than men are considered sending areas, with
poor households cannot afford to make the necessary initial investment; the process of migration takes place through middlemen, which exposes poor households to additional risks of various types of exploitation; the migration process may be linked with the problems of trafficking of girls from poorer families.
Secondly, the impact of remittances on the sending communities may be less than expected. Rushidan Rahman (op. cit.) reports that the potential contribution of remittances to poverty alleviation is not
being fully realised. Remittances may not be invested in productive rural sectors or labour-intensive industrialisation that could generate large-scale employment for the rural poor. In the fieldwork village of Lalmai for example, remittances were used firstly in repaying loans, house construction Impact of remittances costs, educational expenses Remittances have and then agriculture. This reshaped class structure reinforces the de-linking and introduced argument (the separation elements of urban of agricultural from non and cosmopolitan or agricultural growth) that londoni life (Islam, 2001). we explored earlier in this chapter. A second factor is that remittances create demand for imported luxury items and are therefore not spent on locally produced goods. Both the constraints on migration for work and the fact that remittances may not fully benefit the wider rural economy partly explain the apparent anomaly between increasing livelihood opportunities on the one hand, and the growing inequity seen in rural areas on the other. The question of who can gain from the new opportunities is explored in more depth later in the chapter. Gender dimensions of rural change Asking people in rural Bangladesh what has changed in your village? elicits different responses from men and women. Women report that they now have access to new kinds of livelihoods and are increasingly involved in paid employment outside the home. Their participation in public space is also increasingly common. For example, every union parishad (UP) now has to have at least three women members, out of a total of 13. Many women have found new roles as members of non-governmental organisations (NGOs) or participants in their meetings and activities.
The improved linkage to the global economy has provided a range of new livelihood possibilities for village women. The RMG industry is one example, but there are others, such as the manufacture of jute handicrafts for overseas markets and a host of other manufacturing opportunities such as tobacco processing, paper-bag making and coir rope-making, to name but a few. Afsar (op.cit.),for instance,reports Globalisation and that the RMG industry diversification absorbed 1.5 million Since the capitalist workers during the past system has little regard decade; more than 90 perfor traditional gender cent of these were women roles and endemic rural migrants from rural areas. poverty provides a strong NGOs are a new source of incentive to work, female employment in vilglobalisation has opened lages. Women are finding up many opportunities work as health and family for women enabling planning visitors, family them (or forcing them) welfare visitors, teachers, to move out of workers on income-genpreviously restricted erating projects, etc.Whilst proscribed roles (see women in most parts of Gender dimensions of Bangladesh do not work rural change, pg121). directly in the fields, women in some areas have begun to do various kinds of field work, both as agricultural wage labourers and on family land or sharecropped land. This is viewed by many, says Rozario, as a temporary expedient to be ended as soon as the family can afford to do without it (see Gender dimensions of rural change, pg121). Another significant impact on womens livelihoods in the past decade was the growth of micro-credit schemes, which lend primarily to poor village women. It is common knowledge, however, that while loans are taken by women, they are mostly used by men. What has been the impact of these changes on womens livelihoods and on

the structure of gender relations?4 Has womens entry into new livelihood activities contributed to their empowerment their ability to exert influence within and beyond the household? Womens logic Women may have little reason to use credit to contest norms such as parda (for example by undertaking incomegenerating projects involving outside work). They are as likely to use credit to withdraw from outside work so as to conform better to gender norms. Such women are nevertheless exercising choice and, improving their livelihood options, and raising their status (see Gender dimensions of rural change, pg121). In practice, women who migrate to the cities for factory work, or who take on other kinds of work involving substantial mobility outside the household, are often obliged to take on more responsibility for decisions in their lives. Their families appreciate the resources but are almost always concerned at the threat to the womens reputation. They pose a threat to the idealised role of the male as breadwinner; the danger of male violence, both from unrelated men and from their own men, is real for these women.
remittances. At the same time, we can expect that the presence of women in outside work, and their role as significant providers of cash income both via work and via access to credit, will gradually affect village gender values. Gender relations are unlikely to change rapidly, but the boundaries of acceptable female behaviour are likely to extend significantly. The institutional rule book is changing The livelihood choices that people make are governed by institutions: the rules, norms and conventions that shape human behaviour.The institutional framework or rule book is a result of an interaction of formal5 and informal influences6. Formal influences are the policies and organisations governed by statute, law or other transparent, accepted and legitimised systems. The informal influences are the covert social norms and deep structures (e.g. class, caste and communal relationships, gender and other power relations; shalish, samaj7 and other traditional decision-making systems) that inform personal and institutional behaviours of actors at all levels (see The formal institutional framework of rural livelihoods in Bangladesh, pg105). The respective power and impact of both, and how the two mesh together, is really what determines what the rule book looks like. Have these rules changed in the past decade? The answer is an over-riding yes! So, what does the rule book look like today? The institutional players are much more diverse, and there are many more. Two decades ago, formal structures were limited to a school, a UP and a tea stall.The teacher doubled as UP secretary and a NGO worker might have visited the village regularly. Today, growth and improved linkages to the outside world have generated a richer and more dynamic institutional life. Many villages
Opinions differ on the question of whether micro-credit schemes have empowered women. Some have argued that the funds accessed, even if withdrawn by women, are mainly controlled and used by men. Rozario (see Gender dimensions of rural change, pg121) argues that women use micro-credit for purposes they value and in terms of their culturally determined logic. Given the current trends, how might womens livelihoods change in the future? Women from poorer families are likely to face further pressure to enter into outside work, either locally or through migration to urban centres. Dowries are likely to rise further, driven up in part by the increased cash availability in rural areas as a result of
now host small business enterprises and traders as well as traditional shops and tea stalls. Education may be delivered by madrassah8, State school, NGO-run school, private or community schools. A similar range of providers serves health needs.Agricultural extension is most often provided via advice from the private-sector supplier, followed by the NGO and much less frequently the block supervisor. From a situation where there was an interface between government administration as an occasional visitor and informal institutions, there is now a complex set of institutional forms and relationships. Secondly, market forces are emerging as more significant dictators of the rules of the game. The increased integration of rural Bangladesh into the global economy is challenging existing norms. Until recently, the private sector at village level comprised local shops with a limited range of goods, artisans and cottage industry. Infrastructure development, notably roads, bridges and electrification, has significantly changed the physical and socioeconomic landscapes. Multi-national and national suppliers and wholesalers are active even in remote locations. This has led to agencies and networks for supplies and marketing that link the village through the district (or bypassing it) to national and international levels. A third development is the growth of local government. UP members are playing stronger roles as chairs of formal and unofficial committees and bodies. However, role confusion and a lack of authority and accountability between local politicians (union parishad) and local bureaucrats (at upazilla/district level) are limiting the potential role of local government. Finally, informal institutions, social structures and class formation in rural com-
Local governance
In Maniknagar the samaj is strong and reinforced by Islamic orthodoxy; the UP chairman is benign, and local government is now powerful enough to challenge the traditional power bases. In Haorpur the samaj is being encapsulated by patrimonial authority the domination of a British citizen-cum-UP chairman resulting in personalisation of political power. He does not attend the UP office and has a retinue of middlemen. In Ratanpur governance reflects traditional pluralism a weak samaj, a religious complex devoted to a saint and patron-clientism. Macro-politics overlays this. In Lalmai emergence of a civic community centred on a Community Based Organisation (CBO), which has partly taken on the functions of the samaj. It plays a key role in development activities in the village. Political awareness is high but united in the common interest of the village. Source: Islam 2001
munities are also in rapid flux. Social networks are breaking down and there are signs of change in social structures, with women Change in social increasingly visible in structures institutional space and There is evidence of elected representatives gradual disintegration in being drawn from outside family networks especially established elites. The among the poor as a result changes are neither uniof economic pressures, form nor comprehensive, dispersal of families but two themes emerge. through river erosion, Firstly, growing complexurban migration, etc. ity and sophistication. (see Gender dimensions Secondly, traditional of rural change, pg121). power structures based on the historic and deep social structures are giving ground to new power relations mediated increasingly by market forces,
although older elites may still control petty and grand corruption (see The informal institutional framework in rural Bangladesh, pg97). This change is evidenced by the dramatic changes in landtenure arrangements that have come about in the past two decades. Sharecropping arrangements are giving way to fixed-rent tenancy and medium-term leasing arrangements. Saha (op. cit.) notes that the area under share tenancy has declined from about 74 percent in 198384 to about 62 percent in 1996, whereas
the area under fixed-rent and other tenancy arrangements has increased from about 26 percent to 38 percent. The growing range of different institutions and forces gives rise to increased variability across rural Bangladesh in the way power is played out among the competing institutions. However, Zillur Rahman (op. cit.) reveals that, despite such complexity, rural people have clear ideas about the functions and capacities of both formal and informal institutions. Table 1 illustrates the findings of a survey to explore these issues, which he claims demonstrate a clear degree of popularly legitimised role-specialisation. The good news is that this institutional diversity provides opportunities for new livelihoods to emerge and thrive. Zillur Rahman (op. cit.) claims that Table 1 should be read as an opportunity map which makes explicit the cast of actual and potential actors relevant to local governance and the potential for synergies. An important caveat is, however, that some of the actors may also be associated with negative governance attributes such as corruption and violence. The bad news is that policy formulation remains centralised in Dhaka and therefore disconnected from local institutional realities. The failure of centrally determined policy either to influence or respond appropriately to changing livelihood needs at the micro-level thwarts the attempts of the poor to construct new livelihood strategies and pursue better livelihood outcomes. Whilst this is the case, the business of most formal institutions will proceed with little change and will continue to have little relevance to the poor. The extreme poor will remain out of the reach of NGOs; policies and legislation aimed at reducing poverty will
Table 1 Institutional actors and functions

Function Actor perceived to be most relevant
Dispute resolution Traditional elders Educated class Local elites Chairman/member (UP) Law and order Traditional elders Educated class Youths/clubs Community action Disaster-coping NGOs Local elites Government agencies Youths/clubs Community action Political leader/workers Protection Traditional elders from harassment Educated class Political leader/worker Chairman/member Local elites Government agencies Development NGOs Government agencies Business samities Social welfare Youths/clubs See Re-thinking local governance: towards a livelihoods focus, pg113.
remain ineffectual, as will institutions intended to address vulnerability. We return to this issue in the second part of this chapter.
landlessness is increasing and alternative employment in the village remains limited; q extended family networks, traditionally an important safety net for the poorest families, are gradually being eroded.
q
ONE-OFF HEADLINES OR THE EMERGENCE OF A NEW NARRATIVE?

Are our headlines backed up by others experiences? Rozario (op. cit.) provides good evidence that they are: she notes the following changes in Bangladeshi society in recent years which have impacted significantly on women, but also have broader application. According to Rozario, whilst the new livelihood opportunities are often more accessible to men (given the prevalent bias in favour of men), other governmental and NGO interventions have redressed some of that imbalance. Positive changes include: increased school attendance among females and the poor; q increased access to finance, which has increased womens financial independence; q increased pressure, through NGOs, to modify aspects of gender relations; q increased access to social safety-net provisions such as Vulnerable Group Development (VGD), involving work for food or cash.
q
The net result seems to be the emergence of a new set of winners the middle poor and, in some cases, women. The losers remain the poor and extreme poor.
CHANGE BRINGS NEW WINNERS AND LOSERS

Change is rapid and far-reaching in rural Bangladesh.The new rural scenario is one in which urbanisation is an important demographic feature; where globalisation and regional development bring both threats and opportunities; and where agriculture is one activity amongst many other actual and potential rural and nonrural activities. Livelihood wins in the agricultural sector are less significant than in the past. This is not to underplay the considerable impact that high agricultural growth rates had in reducing rural poverty, particularly during the 1970s and 1980s, or its continuing role in the future, ensuring The new rural reality: food security and absorbwinners and losers ing large parts of the q Rural-rural differences labour force. Rather that in poverty incidence time has moved on, and the and rates are as present wave of livelihood important as opportunities is emerging rural/urban differences. in the non-farm sector and q There is increasing often from outside the vilpolarisation between lage. Human capital is rural areas in terms becoming as important as of the nature of natural capital when it livelihoods comes down to taking adopportunities. vantage of new livelihood q There is increasing opportunities.Hands, not inequity both within land, is the critical asset in and between rural the new world. communities.
There is a new mix of winners and losers, says Rozario, and the poorest are possibly worse off than ever. For example: patron-client relationships within village communities are disintegrating, with consequent loss of security for poorer members of the community; q local resources once available to the poorest, such as fuel wood, common grazing lands and fishing grounds, are dwindling;
q
Change is happening faster in some places than in others The emerging rural landscape defies simple description. Today we are seeing an increasing diversity of rural environments; a continuum within rural areas, from those where traditional views still hold true to others where a more modern picture change is taking hold. The implications for livelihoods are not yet clear, but the disparity in the level of poverty at either end of this continuum is increasing. Paradoxically, it seems that the fastest growth rates in non-farm livelihood opportunities are in areas where agricultural growth rates were highest in the past. People from these areas are generally better-educated, have access to more assets and a superior infrastructure, enabling them to take advantage of the new opportunities. Not surprisingly, the highest rates of poverty decline are found in these areas. In contrast, poverty is endemic in areas that have lower agricultural potential. People here begin with fewer natural, human, physical and financial assets. The widening regional poverty gap is highlighted by Afsar (op. cit.), who notes that, with the exception of Faridpur district, all other districts experiencing high levels of emigration have Human Poverty Index9 (HPI) values of up to 35 percent. Those districts mainly in the north that perform worst (with HPIs of over 45 percent) experience less emigration. The traditional disaggregation of poverty statistics into rural and urban categories can obscure these very important regional differences differences in poverty levels between rural areas can be as great as the differences between urban and rural areas.
Who are the winners and losers? To what extent are the poor connected to the new growth and the livelihood opportunities it creates? The increasing inequalities within villages underline the fact that the poor, and notably the extreme poor, remain isolated from the new developments. Those endowed with or having access to resources have thrived, whilst those lacking them have remained locked to agriculture or moved to non-agricultural livelihoods that are less productive and bring lower returns. There are some positive signs that women also benefit from emerging livelihood opportunities, enjoying greater mobility and economic power but Growth but not there are also some poverty reducing? counter-indications such Rural income and as the rising incidence and expenditure inequality cost of dowry and the increased during the early continued discrimination 1990s. Expenditure data in wage levels. shows that between 1991-92 and 1995-96, Both in the farm and nonthe Gini coefficient in farm sectors, the poor rural areas increased from struggle to access the new 26 percent to 29 percent. opportunities. They are In 2000, it declined to unable to benefit from 28 percent. Inequality agricultural livelihood opin income is likely to portunities as they have be even higher (see Rural little or no land. AgriPoverty: Patterns, Processes cultural growth is driven and Policies, pg79). by mechanisation which, in some cases, reduces the need for agricultural labour and so removes an important source of livelihood for poor households.Although mechanisation generates some new jobs in technical services and the trade in equipment and spares, such opportunities are inaccessible to the poor.As well as being excluded from new opportunities, there is also evidence to suggest that poor people are losing access to traditional livelihoods. For example, access to common property resources such as water bodies, grazing land, and
forests is becoming more difficult, as better-off households seek de jure or de facto long-term leases to them. The capacity to shift livelihoods is increasingly becoming a new dividing line within the ranks of the poor, with those unable to negotiate such shifts emerging as new categories of poor. Examples here are rural artisan groups and nature-dependent ethnic minorities whose traditional occupations are disappearing but for whom compensatory entry into new occupations is uncertain at best. A contrasting example is that of women labourers who lost their traditional employment in dheki (home-based manual rice milling) but found compensatory entry into milling work in the chatals (mechanised rice mills) (see Rural Poverty: Patterns, Processes and Policies, pg79). Many rural households have successfully shifted out of agriculture but, again, a disproportionate share of benefits seems to accrue to households from middle and upper income groups. Social, institutional and economic barriers limit poor peoples abilities to take advantage of the new opportunities. Labour is their major asset but even this is compromised by their lack of education and skills, poor health and the fact that there are more female workers than male. Some barriers have been overcome; would-be migrant workers have been able to secure loans as a result of better credit markets, for example. Without appropriate skills at a sufficiently high level, education and knowledge, poor people quickly reach a glass ceiling that limits further progress. This has created polarisation.When poorer households do find employment in the non-agricultural sector, they may find that wages are no higher than agricultural wages and the conditions are worse. The emerging
Barriers to entry impeding poor peoples access to new opportunities

Even if some non-farm activities do not require special skills, organisational and entrepreneurial ability is not common among the poor. Such activities require numeracy, literacy and management ability.They also require knowledge about the outside world: when and where to buy the raw materials, capital equipment and how to market the output, etc. These are the ingredients of psychological capital in which the poor are usually deficient (see Rural Poverty: Patterns, Processes and Policies, pg79).
diversity and complexity in the institutional framework with new rules and systems can make the benefits inaccessible to some, especially the poor and vulnerable who lack knowledge or skills to operate in a new institutional context. The result is that the poor remain, as always, institutionally marginalised. The final chapter explores some of the questions that the new rural reality raises.
NOTES
I Rural-urban centres refer to the rural towns (often the district and sub-district headquarters) which have exhibited fast growth rates over the last decade. 2 Agricultural or farm income includes income from crop and non-crop sectors, while non-agricultural or non-farm income refers to income from nonagricultural activities such as trade, transport, construction, services, etc. 3 Start (2001) describes the U-curve of diversification, which sees non-farm activity concentrated among the poor as a survival or coping mechanism and among the rich as an accumulation strategy.Toufique (2001) explores in detail some of the elements of distress-driven migration in Bangladesh. 4 See Gender dimensions of rural change, pg121, for an introduction to the structure of gender relationships.
5 See The formal institutional framework of rural livelihoods in Bangladesh, pg105, for an overview of the formal institutional framework. 6 See The informal institutional framework in rural Bangladesh, pg97, for an overview of the informal institutionals. 7 Samaj refers to the indigenous institution centred on the hamlet (or para) or in some cases a larger area of the village. Headed by ritual elders it represents the normative order of the collectivity of the community which it reproduces through ritual and ceremonial performances.The samaj maintains social control, settles disputes and deals with the outside world on behalf of its members. Shalish is the instrument through which the samaj undertakes conflict resolution. It is a sitting of village elders and other leaders to settle disputes and punish offenders. 8 Madrassah is a traditional Islamic religious school. 9 The HPI is composed of three indicators: health, knowledge and overall provisioning. Deprivation in health is indicated by vulnerability to death at a relatively early age. Deprivation in knowledge is captured by the percentage of adults who are illiterate. Deprivation in overall economic provisioning is operationalised by the percentage of people (i) without access to safe drinking water; (ii) without access to health services; (iii) percentage of children under five who are moderately and severely underweight (Anand and Sen 1997).
33
Chapter 3 Postscript
This book has developed a new story of how rural Bangladesh is evolving. What are the implications of this story for the way development is approached? And is the story complete, or are there more chapters to be added? prehensive PRAs and suggest a need for open-ended discussions around how livelihoods are changing to complement these methods. Taking a step back and asking bigger questions about what has changed, how people see the future and who the winners and losers are is a valuable way to start (see Understanding livelihood dynamics, pg 22).
IMPLICATIONS FOR EXPLORING LIVELIHOODS

The picture of livelihoods presented in this book emerged from a rapid field survey aimed at investigating rural peoples perspectives on livelihood change. The fact that the majority of our observations were backed up by the commissioned papers, which brought together macrolevel thinking on key rural issues, suggests that such an approach holds much value, particularly in a context such as rural Bangladesh where change is so rapid. Effective development strategies need to be forward-looking and to anticipate how poor peoples livelihoods will evolve the new opportunities that are likely to emerge and how they can equip the poor better to ensure they are able to take advantage of them. At the strategic level, the management of information is critical we need enough information for an understanding of the broad directions of change: where there is consensus and where there is not. It does not mean we are not interested in the livelihood detail only that we must understand the scale of variation and ensure that strategies accommodate and respond to it. Real insights into dynamics and change are often difficult to come by through detailed questionnaire surveys and com-
QUESTIONS THE NEW RURAL REALITY RAISES

The picture of rural livelihoods that this book paints raises many questions for the way development is approached. To what extent is it about rural development? Rural livelihoods are not just about what is happening in rural areas. The ruralurban divide has been replaced by a ruralurban continuum. Interventions that support rural livelihoods are no longer located only in rural areas; they must include urban, national and international arenas and actors. Is there a need to target the poor directly? Many of the poor are missing out on the new opportunities and there is a need for a more explicit focus on the constraints and barriers they face. In areas of widespread deprivation and poverty, a singlesector response, whether that is in education or natural resources development, is unlikely to achieve much poverty reduction.The poor need to build up their assets on a range of fronts if they are to be able to compete in the new world.
Does the new analysis necessarily mean new entry points? Rural livelihoods are now about much more than agriculture and natural resources. Poverty-reducing livelihood prospects are greater in the non-agricultural arena. The poor lack the skills, education, health and assets to optimise the emerging non-agricultural opportunities. There is a need to focus on the barriers that prevent poor households taking advantage of new opportunities. Should we abandon agriculture? Maybe not it is still important to livelihoods. However, it needs to be approached in a different way. We need to look at agriculture for agricultural livelihoods sake and not as a rural growth engine. The traditional assumption that investments in agriculture which often miss the poor will indirectly reach them through the linkage effect and the creation of livelihood opportunities in the non-agricultural sector, is risky. We need to be much more focused if we are to reach the poor through agricultural-based interventions targeting them directly,
addressing market failures; creating level playing fields for small farmers; reducing transaction costs; looking carefully at low potential areas; supporting farmers in the new technology market place. Understanding differentiation of rural households in terms of access either directly or indirectly will be critical to assessing the potential that agricultural growth holds for generating livelihood opportunities for the poor.Where important productivity niches (ecological or location-related such as peri-urban areas) coincide with the presence of vulnerable and poor populations, there is scope for poverty-reducing agricultural growth. How can we best support diversification livelihood? Diversification away from agriculture is now increasingly a reality. The growth areas in the non-farm sector migration, transport, construction, etc. are not the traditional non-farm spheres where donors operate. Supporting diversification brings with it the same challenges as supporting agriculture. It is not necessarily
Understanding livelihood dynamics

q What
are the comings and goings in the community? q Is this a good year or a bad year, and why? q What is the most difficult time of the year for you, and why? q How do you manage during those months? q What are things that people are talking about in the community at the moment? What is the gossip? q Where do the young people hang out? What do they do/talk about? q How are relationships with outside institutions these days? q What are the bright spots? q What has changed in the last few years? q Who is getting richer? Can you think of a household that is better (worse) off now than a few years before? Why are they richer (poorer) now what happened? q What are your aspirations for your children? q If you compare this community to community x how is this one different is it the same, richer, poorer, and why?
Postscript 35
about public expenditure and it demands a multi-sectoral response (not just about skills but also about infrastructure, finance, etc.). Is weak institutional capacity the issue or the context? Policy formation and implementation is a chaotic and complex process operating at all levels. However, the prevailing approach in Bangladesh in the public, NGO, and even the larger elements of the private sector, is of a centralised Dhaka-based model of policy-making and planning. Such macro-level policy emanating from Dhaka is disconnected from institutional practice at the local level. The prevailing donor paradigm over the last decade has been to tackle the centre head on on the assumption that that is the way to sustainably scale up the rate of poverty decline. New sets of institutional questions are, however, emerging.There is a need to be clear about the timeframe we are working to, about the most effective way to change institutional practice at the local level (not necessarily from the centre) and who our key institutional partners are (private- as well as public-sector actors).
sufficient for strategy development, a new agenda has emerged at the next level down in terms of developing responses to support livelihoods in the new context. Some missing links We need to understand better: the macro-economic context how globalisation, trade agreements, investments, communications interface with poor peoples livelihoods; q how to make markets work better for the poor; q the political and social networks which influence constraints and opportunities for the poor; q systems of local governance the diversity of institutions at the local level and how they work; q the impact of the changing quality and quantity of common property resources on the poor; q the changing dimensions of migration and how the poor can access and benefit from them.
q
HOW COMPLETE IS OUR PICTURE?

The livelihoods picture in this book emerged from a rapid appraisal of livelihoods in a limited number of places in Bangladesh. Whilst it paints a picture of broader trends, it leaves unanswered many questions, particularly at the micro-level, of how individual households are managing and the intra-household implications of change. There are also outstanding questions as to where the majority of rural Bangladesh lies along the spectrum of change that was outlined in Chapter 2. Whilst broader understandings may be
And finally, above all, we need a better understanding of what it means to be poor and disenfranchised in the rural Bangladesh of today and why certain groups are excluded from the new opportunities. To achieve this we must move away from traditional categorisations of poverty such as rural and urban or landed and landless and seek to understand poverty in the context of the emerging livelihoods picture, where shortcomings in terms of access to social and human assets may be as limiting as access to land was in the past.
REFERENCES
ANAND, S. AND SEN, A. (1997): Concepts of Human Development and Poverty: A Multidimensional Perspective. Human Development Papers, 1997. United Nations Development Programme (UNDP). New York. BBS (1986): Census of Agriculture and Livestock, 1983-84. BBS, Dhaka. BBS (1999): Census of Agriculture 1996, National Series,Vols. 1 and 2. BBS, Dhaka. DFID (2001): Poverty: Bridging the Gap, DFID Guidance Notes, 2001. DFID, London, UK. ISLAM, S.A. (2001): Livelihoods and Institutional Assessment for DFID Bangladesh. Phase 2: Field Work Report. Unpublished. SHAHABUDDIN, Q. AND QUASEM, M.A. (2002):Agricultural Growth: Performance and Prospects. Seminar on Performance of the Bangladesh Economy. BIDS, April 2002. Dhaka. START, D. (2001):The Rise and Fall of the Rural Nonfarm Economy: Poverty Impacts and Policy Options. Development Policy Review,Vol 19 (4), 2001, Special issue on Re-thinking Rural Development, Overseas Development Institute, London. TOUFIQUE, K. (2001): Sustainable Rural Livelihoods in Bangladesh, IDS Research Report 45, Institute of Development Studies, University of Sussex, UK. WORLD BANK (2002): Bangladesh: Progress in Poverty Reduction. Background Paper, Bangladesh Development Forum, Paris, March 13-15, 2002.
Hands Not Land:

Part 2: Original research papers
39
Livelihoods in contemporary rural Bangladesh: An overview of research papers

In this part of the book we include eleven papers on various aspects of rural livelihoods in Bangladesh.These papers provided the knowledge and information that is the foundation for Part 1 of this book. All of the papers are based on existing knowledge: they relied on secondary sources of information. They make a bold attempt to explain the complex issues pertinent to livelihoods in contemporary rural Bangladesh.This overview draws out some of the interesting themes of the papers. One of the main themes of this book is land and access to land, and several of the papers examine land dynamics. Saha assembles a mass of secondary data in an attempt to analyse agrarian structures and livelihoods in rural Bangladesh, looking in particular at trends in land ownership, land tenure, income and employment structures. Data from the two agriculture censuses (1984 and 1996) show that there has been an increase in functional landlessness. However, extreme landlessness defined as households with no land did decline during this time span. There was some consolidation of land by medium landowners and a decline in land holdings amongst large landowners. Land fragmentation increased. The number of non-agricultural households has increased and consequently the contribution of non-agricultural income to the total household income. Explanations for these trends remain in the domain of conjecture. For example, in her paper on poverty, Rushidan Rahman opined that land fragmentation is due to demographic factors. Another important conjecture would be that land no longer brings sufficient income and the rich are selling land and investing profits in other non-agricultural activities, in education and in migration. It is also critical to view changing patterns of land ownership in the changed context of rural livelihoods in Bangladesh. As land becomes less and less important as a determinant of livelihood success and Saha provides ample information to support that it is becoming so one has to consider the importance of non-land-based assets in sustaining a livelihood. Islams recognition of how informal institutions, political resources and patronage are also shaping rural livelihoods also emphasises the fact that land ownership per se is not the major factor determining whether a household thrives. The papers also note changes in land tenure. Access to land by the poor is increasing through tenurial rights rather than ownership rights.The market for sale and purchase of land continues to remain thin. Within the tenancy market there has been a significant shift away from sharecropping arrangements to fixed-rent contracts. The literature has sometimes regarded sharecropping as a semi-feudal means of production in which tenants were exploited and land was eventually grabbed by landlords (Bhaduri 1973). However, Sahas data give no credence to this assertion. Sharecropping is not only declining in rural Bangladesh, but the share to landlords is ever diminishing. When we consider consolidation of medium-sized farms it seems that the contractual relationship in agriculture is moving more towards the benefit of the poor.
A second major theme of the papers is the changing structure of the rural economy and the implications for rural livelihoods. As Zillur Rahman writes,diversification is a defining characteristic of rural Bangladesh today and this is evident both at a household level (where people engage in more than one activity) and for the rural economy as a whole (with an ever-increasing range of activities present). Sahas data show that labour has shifted from agricultural to non-agricultural activities; consequently, more and more income is now generated off-farm. He argues that this reallocation of labour is the result of push factors agricultural livelihoods are not available for a large mass of the rural population and the poor are pushed into poorly paid, low-potential, non-agricultural activities. Rozario examines the gender dimensions of these shifts and picks out three key processes globalisation, migration and monetisation of the local economy which have changed womens livelihoods as well as womens perceptions of their role. Zillur Rahman comments on the ease with which rural people are able to shift between different livelihoods or attain and maintain multiple livelihoods. He concludes that land is no longer a key resource for economic and social mobility in rural Bangladesh. People have non-land-based livelihoods as well as multiple livelihoods. Although livelihood diversification is often assumed to be a positive process, the papers contest that this is always so. Toufique observes a dualism within the non-agricultural sector, where high-potential, peri-urban-based activities contrast with low potential, village-based activities.The former require more capital and skill and are more productive, whereas the latter are low-productivity activities that use less capital and more primitive technologies. In this context, the crucial issue is how people diversify their livelihoods. And for many especially the poor the diversification opportunities have come at the lower end of this spectrum. Zillur Rahman has termed this the capacity to shift: those with better capacity to shift gain while those with no capacity to shift lose. It is very difficult to switch from low-end diversification to high-end diversification because of constraints such as skill differentials, access to networks, etc.The issue is not therefore one of more diversification versus less diversification. It is the type and nature of diversification that is important. Migration is a key element of the rural diversification process and Afsar deals with the issue of migration and rural livelihoods with the utmost clarity. Amongst other issues, she also focuses on the who? question; which households have the capacity to migrate? She points to the fact that migrants are heterogeneous; the poorest may find their way to squatter settlements, while the better educated and skilled end up in better living conditions. She notes that the pattern of migration is complex. People migrate from agriculturally-developed regions as well as from vulnerable and agriculturally-backward regions. Policy makers often ignore this heterogeneity, and migrants are often unaccounted for in official statistics. Overall, the papers tell a story of a rural economy in transition.Toufique looks in detail at the new economy and analyses the changing links between agricultural and non-agricultural activities. He re-examines the traditional linkage assumption: that agricultural growth in turn promotes non-agricultural growth. He claims that empirical observations show that, during the period 1983-95, the non-agricultural sector grew impressively and that this cannot be explained by the meagre growth of the agricultural sector in the same
An overview of research papers 41
period. He concludes that this implies that the traditional links between the two sectors are undergoing both qualitative and quantitative change.Although there are insufficient data to conclusively prove Toufiques hypothesis, he identifies important changes, such as infrastructural development, migration and urbanisation, which may explain the impressive growth observed in the non-agricultural sector during this period. Afsars observations on remittances support this they have played an important role in smoothing consumption, increasing investment in human capital and providing for missing markets such as those that provide security for the aged, disabled, sick and homeless. At the same time, demand for outputs from the non-agricultural sector was fuelled by transactions within this sector and by the urban population. Toufique does not deny the traditional link between the agricultural and non-agricultural activities.The consumption link is created by the incremental income gained from agricultural growth that is spent on goods produced by the non-agricultural sector.The production link results from the demand created by agricultural growth for inputs and farm implements produced by the non-agricultural sector. Finally, the backward link refers to the processing of agricultural produce by the non-agricultural sector. These links are strong and agriculture will continue to remain an important source of livelihoods for a large number of people and for some time.What is important, however, is to recognise the limited role of agriculture in scaling up poverty reduction in Bangladesh. If there is to be a rapid decrease in poverty, one cannot afford to ignore non-agricultural livelihoods; both the traditional and new links between the agricultural and non-agricultural sectors must be strengthened. The papers present evidence of tremendous change in rural Bangladesh. However, a note of caution is necessary; the changes the papers observe should in no way imply that agriculture should be ignored. It is still by far the most important sector for rural livelihoods and especially so for the poorest households most of the poorest people continue to earn a living as agricultural labourers.What is needed instead is a greater understanding of the potential growth of agricultural opportunities. Since independence, the key objective of Bangladeshs Rural Development Policy (RDP) has been to attain self-sufficiency in food-grain production.The route for achieving this has also been very similar with a strong emphasis on the technological matrix. However, Asaduzzaman points to the untapped potential of agriculture-based livelihoods; in particular, at growth based on the new agriculture or the growth of high-value, non-cereal crops, livestock and poultry sectors. Mandal chooses to look at more indirect routes opportunities in the spin-offs from non-crop products and high-value crops. Low levels of income and a large population ensure that demand for food-grain will remain high. But, as income increases, people will start to consume more and more non-crop products such as poultry and livestock. These products need different technology for production and marketing. This is the route through which non-agricultural livelihoods are likely to develop. A key issue that Mandal examines is the extent to which the new rural landscape is reflected in the RDP. Such an assessment is perhaps critical at this point given that food self-sufficiency in Bangladesh has almost been achieved.This is a significant achievement and might provide the space and opportunity to look at the rural economy in a new light.A quick reading of Mandals paper would say that the RDP is now more or less as
it should be. However, a few problems remain.The main one is the failure of the RDP to reflect the changing realities of rural livelihoods in Bangladesh. Afsars paper, for example, questions the differentiation of the rural from the urban or the modern from the traditional, but the RDP does not reflect any uncertainty about these terms.This leads to the question as to whether Bangladesh needs a rural development policy or a development policy? By now the reader will have picked up enough messages to question whether the term rural livelihoods bears any relation to the reality of how the rural population of Bangladesh makes a living.Thus, although they may be considered rural in terms of the locations of their households, the papers highlight how they draw on a wide range of activities across the ruralurban spectrum to make a living.This blurring of the ruralurban divide is a recurrent theme in the papers, with authors commenting on the relationship from their different perspectives. For instance, Zillur Rahman states that improved infrastructure and population mobility have changed the rural-urban divide to a rural-urban continuum. He also introduces the concept of multi-locational households in which members may reside in a village while deriving his or her income from urban sources, or commuting between the two areas. Afsars paper outlines how rural people move in time and space to pursue a complex set of objectives:it therefore blurs the dividing line between rural and urban, or stretches it further between the modern and traditional. Rozario comments on the gender dimensions of this continuum, noting that women increasingly cross this rural-urban divide. The ready-made garment (RMG) industry has been a key driving force behind this process. This movement of labour and products narrows down the dichotomy between the rural and urban. Therefore, Afsar looks for a space in between. She frequently brings in the issue of growth of small and medium-sized towns and the role they can play in helping people to pursue better livelihoods. She repeatedly asks for the revival of the upazilla system the system that relocated legal and administrative systems to lower levels. It was expected that people would move to intermediate towns and create demand for goods and services which, in turn, would be translated into improvement in infrastructure. Islam attempts to understand more deeply the vulnerability of rural livelihoods. Vulnerability is defined in terms of exposure to risk and insecurity. For Islam, natural disasters are the main sources of vulnerability. Households that are better endowed with diverse assets are less likely to succumb to the downward spiral of asset depletion, insecurity, indebtedness and destitution that can sometimes result from the impact of these vulnerability factors. Many of the papers try to get to grip with the institutional framework that has underpinned the changes outlined above.The strongest message to come out of the papers is that the institutional framework is of paramount importance in determining whose livelihoods thrive. Islam examines the complexity of the informal institutional framework specifically samaj and shalish and also the traditional structure of patron-client relations. He notes that these institutions have frequently served to exclude the poor and women.This is a point examined in more detail by Rozario who notes that the extent to which women may benefit from new opportunities is partly governed by the degree
of flexibility shown by cultural institutions.While women in some areas have been able to gain from the new opportunities, other areas have been more resistant to change and institutions such as parda and the associated values of purity, honour and shame continue to constrain the economic, social and political contributions of women. In a nutshell, livelihood opportunities have increased womens ability to participate in outside employment, but this does not necessarily mean that women today are any more empowered. Islam notes that changes are occurring in the informal framework and it is critical to ask whether these changes are in a positive direction for the poor. He notes particularly the change in the nature of patron-client relationships on which the poor often depend in order to access resources and livelihood opportunities.These relationships are increasingly based on non-land resources, such as political capital, and are formed in an environment characterised by extreme lawlessness and violence. Rushidan Rahman refers to the rent-seeking activities fostered by these new sources of patronage, which distort incentives and generate income inequality (as there are severe entry barriers to new livelihood opportunities).This patronage can take many forms. One important form is political patronage, which often leads to resources being allocated according to political loyalty rather than according to the needs of the poor. Formal institutions are the main mechanism through which the State can impact on local livelihoods.They are also a route through which demand for new livelihoods and institutions can change State policies and attitudes. This two-way interaction can be complex; not only may State policies be implemented through this process, but also problems associated with policy formulation and implementation can be identified locally, communicated to the centre and addressed in future policy changes. National-level policies are implemented by intermediate institutions at the meso- and micro-levels.There is a significant information gap between these levels.The State can frequently fail to read the right signals, or read the wrong signals as right ones, because of political patronage and its negative influence on decision-making and implementation processes. A large number of actors and informal institutions intervene in the transmission process. Thornton emphasises three tensions in the formal institutional framework that contribute to its ineffectiveness.The first relates to the institutional dislocation of the microlevel (village/union), where a long tradition of self-governing informal institutions confronts more formal structures emanating from outside. The representatives from outside (public servants) are often absent and the representatives of self-government, even when elected to local bodies, have yet to gain effective power or voice. This disconnection is evident in the way that centrally determined policy fails to relate to the local level and, in turn, livelihood realities and dynamics fail to inform policy. A second tension exists between politicians and the bureaucracy.The bureaucracy dominates the agenda at the centre and this is mirrored at other levels. Power and dysfunctional behaviour from the bureaucracy conflicts with self-interest and lack of vision from politicians. The result is a weak and ineffective macro-governance environment. In response to this governance crisis, there is still only limited popular voice demanding change at any level. Rather, this third tension is inactive and replaced by disaffection and fatalism such that rural livelihoods are pursued with no expectation or attention to public systems.
This depressing picture should not obscure, however, the institutional dynamism evidenced at the local level which Thornton, Islam and Zillur Rahman all note. Market forces are setting new rules of the game. Changes in livelihood links and the growth of non-agricultural livelihoods, claims Toufique, have increased transactions many times over, thereby enhancing the role of the market. Popular voice has also started to surface. On the other hand, new actors are appearing in the institutional space. Increasing urbanisation, migration, etc., have started to change formal institutions, which are growing more diverse and complex. While the policy-making process remains centralised and bureaucratically controlled, institutional change at the local level can be impressive. Recognising this local level diversity, Zillur Rahman draws attention to the need for a livelihood-driven local governance agenda for Bangladesh.The new aim is to regenerate the local economy, and local governance is an important vehicle for achieving this. The regeneration of the local economy is not possible if local governance is weak and constrained. Traditional local government cannot serve the needs of a new cadre of actors who move beyond sectoral or rural-urban confines and frequently take up multiple livelihoods. As Zillur Rahman clearly points out, there are actors other than the local government who play a major role in pursuing livelihoods. He develops a multi-agent framework for local governance where different actors play different roles. Overall, the papers explore a range of different dimensions of the rural scenario. A key question, however, is whether the papers provide answers to one of the major paradoxes concerning the development of Bangladesh the slow rate of decline in poverty despite a relatively high rate of economic growth.Why should this be? What has happened to the expected impact of growth in micro-finance on poverty? Why has agricultural growth not translated into poverty reduction? What about the various safety net programmes? While we have recognised that income distribution has worsened, we are at a loss to explain it. There are some answers in the papers. We know, for instance, that graduation from poverty requires access to a range of assets and not just to land.This is evident from the fact that, as Saha reports, there has been a positive trend in land ownership.Yet this has been accompanied by increasing income inequality in rural areas. This suggests that, while land may be the crucial factor governing agricultural production, it may no longer be a major determinant of rural income.Access to a range of assets including land are also critical for insuring and mitigating against risk, as Islam and Rushidan Rahman highlight. The papers also show that increasing inequality is all about the inequitable access to the new livelihood opportunities arising from economic growth. It is important, however, to look at the issue of winners and losers carefully. For instance, whilst Rushidan Rahman claims the increasing inequality in rural incomes is a result of non-poor households receiving remittances, Toufique reports that recent findings show that the poor also migrate and send remittances (2001).The answer lies in the nature of the new opportunities that the poor can access always at the lower end of the income spectrum. Agricultural livelihoods are likely to remain important for the poor but are unlikely to provide the poor with a route out of poverty.
Finally, the papers recognise that it is the institutional framework that will determine the extent to which poverty reduction can be scaled up in rural areas.They note that macrogovernance reform is a precondition to achieving a sustained reduction in poverty. But this relies on intermediate institutions to emerge and function at meso- and micro-level to ensure that policy reforms are implemented at the level that really counts for the poor. For the foreseeable future, however, it is the informal institutions that will continue to shape every element of the livelihoods of the poor, limiting their choices and barring their entry to new opportunities. As Rushidan Rahman points out, the rural power structure is based on new activities and new institutions that are intertwined with the political, administrative and trading systems in a complex way. The fact that the rural poor can neither join this nexus nor ignore it has far-reaching implications. An effective attack on poverty should include improving access of the rural poor to these new institutions.This makes the task of poverty reduction difficult in Bangladesh.
REFERENCES
BHADURI,A. (1973):Agricultural Backwardness under Semi-Feudalism. Economic Journal 83, no. 329 (March 1973) TOUFIQUE, K. (2001): Sustainable Rural Livelihoods in Bangladesh, IDS Research Report 45, Institute of Development Studies, University of Sussex, UK, September, 2001, pp. 1-132.
47
Rural development trends: what the statistics say Bimal Kumar Saha
This paper reviews trends in land distribution and tenurial changes, employment, income and related socioeconomic factors that influence rural development. It draws on available statistical information and other data from surveys to uncover patterns and processes of change in livelihoods in rural Bangladesh. NON-FARM OCCUPATIONS ON THE RISE Table 1 highlights the shift from farming to non-farming occupations, drawing on recent national agricultural census data1. From 1984-96, the percentage of households in the farm sector decreased from 73 to 66 percent, whilst in the non-farm sector it increased from 27 to 34 percent. Over this period, the proportion of non-farm households grew at a rate of about four percent per annum, almost triple the rate for farm households. The number of agricultural households normally dependent on providing agricultural labour in the farm sector also declined at a rate of 0.26 percent per annum, whilst in the non-farm sector, labourer households increased at 2.89 percent per annum. In the same line, the proportion of labourer households in the non-farm sector increased from 43 percent in 1983-84 to 53 percent in 1996. The expansion in the participation of rural households in non-agricultural sectors suggests that employment opportunities in agriculture are decreasing, and that nonagricultural activities are more lucrative than agricultural work. Furthermore, the expansion of rural roads, the
TRENDS IN FARM AND NON-FARM LIVELIHOODS

SOURCES OF RURAL INCOME Rural people in Bangladesh construct their livelihoods from diverse sources, both agricultural and non-agricultural. Recent surveys suggest that a significant proportion of rural households derive employment and income from the non-farm sector (Quasem 1996). In 1987, the contribution of agricultural activities to rural income was 63 percent, declining to 54 percent in 1994. During the same period, the proportion of non-agricultural income increased from 37 percent to around 48 percent (Hossain et al. 1996).
Table 1 Distribution and annual growth of households by farm, non-farm and agricultural labourers
Sector *% of households 1983-84 % of agricultural labourer households % of households 1996 % of agricultural labourer households Annual growth rate ( %) Households Agricultural labour households
Farm Non-farm All
72.70 27.30 100.00 (13,817,646)
57.08 43.47 100.00 (5,495,300)
66.18 33.82 100.00 (17,828,187)
47.18 52.82 100.00 (6,401,453)
1.35 3.98 2.15
-0.26 2.89 1.28
Source: BBS (1986, 1999) Note: Figures in parentheses indicate number of households. *Estimates are based on basic data of farm and non-farm households in agricultural censuses of 1983-84 and 1996.
increased marketed surplus of agricultural produce and improved literacy rates have generated new employment opportunities in the rural trade and transport sectors, which in turn have encouraged greater occupational mobility of rural households (Hossain 2000). Various BIDS surveys2 (1987, 1991) show that the land-poor are increasingly engaging in non-agricultural activities. This shift not only increases their income and employment security but also reduces their vulnerability to fluctuations in income flows over the whole year. As discussed below, changes in access to cultivable land and increasing land poverty may be important driving factors behind these livelihood shifts.
whilst the number of large farms diminished. Between 1983-96, there was a significant reduction in average farm size (i.e. cultivated land per farm household) from 0.8 ha to 0.6 ha.There is also evidence of increased fragmentation and subdivision of agricultural land.The average estimated number of fragments per hectare increased from 3.03 in 1983 to 3.57 in 1996. LAND DISTRIBUTION Information on land ownership in Bangladesh is scanty. Table 3 shows the distribution of operational land holdings based on the agricultural census data for 1960, 198384 and 1996. It suggests that agriculture in Bangladesh is carried out predominantly on small farms. From Table 3, we can observe that over the span of 36 years (1960-1996), there has been a dramatic proliferation in the proportion of small and marginal farms, a relative consolidation of middle farms and a relative decline in the proportion of large farms. Importantly, the average sizes of all categories of farms have declined significantly over this period (Table 3).
CHANGES IN LAND DISTRIBUTION AND TENURE

The agricultural census data of 1983-84 and 1996 reveal significant structural changes to agriculture in Bangladesh (Table 2). Overall, the amount of land available for cultivation has decreased (Saha 1997). The area of cultivated land declined from 8.16 million ha in 1983 to 7.19 million ha in 1996 a loss of nearly one million ha of cultivable land since the early 1980s.Table 2 also shows that the number of small farms grew to more than 9.4 million,
Table 2 Changes in number of farm holdings and farm area in Bangladesh agriculture
Parameters Holdings: 1983-84 1996 Change (+/)
a. b. c. d. e. f. g. h.
Total holdings (000 ha) Non-farm holdings (000 ha) Farm holdings (000 ha) Small farm (000 ha) Medium farm (000 ha) Large farm (000 ha) Absolute landless (000 ha) Ag. labour holdings (000 ha)
13,818 3,772 10,045 7,066 2,483 496 1,198 5,495
17,828 6,030 11,798 9,423 2,078 297 1,815 6,401
+4,010 +2,258 +1,753 +2,357 -405 -199 +617 +906
(29.0) (59.8) (17.5) (33.3) (-16.3) (-40.1) (51.5) (16.5)
Farm area:
a. Cultivated area (000 ha) b. Av. farm size (ha)
8,158 0.8
7,192 0.6
-966 -0.2
Source: BBS, Censuses of Agriculture of 1983-84 and 1996. Note: Figures in parentheses indicate percentage change.
Rural development trends: what the statistics say 49
Table 3 Land holdings in Bangladesh in 1960, 1983-84 and 1996

Farm size (hectare) % of farms % of area % of farms % of area % of farms % of area 1960 1960 1983-84 1996 Average farm size (hectare) 1983-84 1996
Less than 0.02 0.21-1.01 1.02-3.03 3.03 and above All farms
13.0 38.0 38.0 11.0 100.00 (6,139)
1.0 15.0 45.0 38.0 100.00 (8,792)
24.06 2.74 46.28 26.24 24.72 45.09 4.94 25.92 100.00 100.00 (10,045) (9,178)
28.45 4.48 51.42 36.70 17.61 41.50 2.53 17.32 100.00 100.00 (11,798) (8,077)
0.1 0.57 1.74 5.1 1.42
0.1 0.52 1.66 4.8 0.91
0.16 0.49 1.61 4.69 0.68
Source: BBS, Censuses of Agriculture of 1960,1983-84 and 1996. Note: Figures in parentheses indicate households in thousands and land in thousand hectares.
LANDLESS HOUSEHOLDS A strong relationship between landlessness and poverty exists in Bangladesh (Hossain and Sen 1992). Landless and land-poor households typically rely on the sale of their labour in farm and non-farm activities.Table 4 shows that the proportion of absolutely landless households (without homesteads) decreased at a rate of three percent per
annum from 1983-1996. However, functional landlessness (without cultivable land and/or with some cultivable land up to 0.2 ha) increased sharply at the rate of 5.23 and 2.42 percent per annum respectively, indicating an alarming acceleration in the rates of landlessness; in 1984, functionally landless households constituted about 50 percent of households. By 1996, this had increased to 60
Table 4 Distribution of landless households in relation to the rest of the households and their growth rates
Category of households/landless 1983-84 Number of households % 1996 Number of households % Growth (%) per annum
Landless: category I Landless: category II Landless: category III Near landless: category IV (marginal) Rest of the households: category V All Source: BBS (1986, 1999).
276,977 2,713,969 3,898,181 1,702,652 5,225,867 13,817,646
2.00 19.64 28.21 12.32 37.82 100.00
162,229 5,003,042 5,191,979 2,494,606 4,976,326 17,828,182
0.91 28.06 29.12 13.99 27.91 100.00
-2.93 5.23 2.42 3.23 -0.39 2.15
Note: Category I: households without homestead land. Category II: households with homestead, but without cultivated land. Category III: households with homestead and cultivated land (up to 0.2 ha). Category IV: households with homestead and cultivated land (0.21 ha-0.4 ha). Category V: rest of the households with homestead and cultivated land (0.41 ha and above).
Table 5 Farm area by type of tenure in Bangladesh: Censuses of Agricultural of 1960, 1983-84 and 1996
Type of tenure % of farm 1960 % of area Average size of area (hectare) % of farm 1983-84 % of area Average size of area (hectare) % of farm 1996 % of area Average size of area (hectare)
Owner Owner-cum-tenant Tenants All
61.77 38.00 2.33 100.00 (6,139)
53.10 45.00 1.00 100.00 (8,792)
1.25 1.70 0.61 1.43
62.78 58.76 35.83 40.69 1.39 0.55 100.00 100.00 (10,196) (9,377)
0.86 1.04 0.36 0.92
61.50 57.80 37.50 41.61 1.00 0.50 100.00 100.00 (11,797) (8,178)
0.65 0.77 0.36 0.69
Source: BBS, Censuses of Agriculture of 1960,1983-84 and 1996. Note: Figures in parentheses indicate households in thousands and land in thousand hectares
percent (Table 4). If we consider marginal farms (up to 0.4 ha) along with landlessness of all types, we find that landless and marginal households constituted about 72 percent of all rural households in 1996, compared to 63 percent in 1984.
LAND TENURE
From 1960-1996, the land distribution pattern by type of tenure remained more or less stable. From Table 5, it can be seen that about 60 percent of farms are owner-farmers. Less than 40 percent of farms belong to the owner-
cum-tenant category. The majority of tenants are ownercum-tenants who already own some land, but rent additional land for improved utilisation of their farm assets (e.g. family labour, draught animals, farm equipment, etc.).Table 5 also shows that the average size of farms has fallen for all types of tenure arrangement. However, the average size of tenant holdings (owner-cum-tenant) is higher than for owner holdings in all the census years. Between 1960-1996, there has been little change in the proportion of farmers involved in tenancy. However, the proportion of area under tenancy has increased from
Table 6 Tenurial changes in 1996 over 1983-84

Rented in/operated area 1983-84 (000 acres) % 1996 (000 acres) % Annual Growth rate (%) of rented in/ operated land
Area rented in under sharecropping Area rented in under fixed-rent and other arrangements Total area under tenancy Operated area Percentage of operated area under tenancy Source: BBS (1986, 1999)
1,137 401 1,538 9,377
73.95 26.05 100.00 16.40
1,092 673 1,765 8,178
61.89 38.11 100.00 21.58
-0.32 4.41 1.15 -0.95
approximately 17 percent of operated area in 1983-84 to about 22 percent in 1996 (Table 6). In addition, there has been a significant change in the structure of tenurial arrangements. Sharecropping arrangements are giving way to fixed-rent tenancy and mediumterm leasing arrangements. Table 6 shows that the area under share tenancy declined from about 74 percent in 1983-84 to 62 percent in 1996, whereas the area under fixed-rent and other tenancy arrangements increased from around 26 percent to 38 percent. According to the 1983-84 census, 90 percent of the sharecropped area was under the 50:50 share category, decreasing to 80 percent in 1996. In contrast, the proportion of the two-thirds sharecropping arrangement increased during the intercensal period, suggesting a reduction in the exploitative nature of sharecropping. The different kinds of sharecropping arrangements (e.g. half, one-third, two-thirds, etc.) do not appear to vary across farm size categories. The main types of non-sharecropping rental arrangements, are mortgage, fixed-rent and khai khalashi (a type of fixed-rent contract). Significant variations across farm size categories do not appear to exist either in these types of non-sharecropping arrangements. Notably, a significant proportion of land-poor households rent land though mortgage and khai khalashi contracts. Therefore, these emerging types of rental markets may not be considered as a step towards acquisition of land by linking them with credit transactions. Zohir (1999) makes a similar observation.This point, however, warrants further investigation. Tenurial arrangements may also be analysed by village and district-level survey data. Survey results show considerable variation in tenurial arrangements. For example, the Differential Impact Study (DIS) conducted by BIDS in 62 villages in 1987 found that the adoption of modern crop varieties was positively associated with the incidence of tenancy. The percentage of tenant cultivators and cultivated area under tenancy were higher in the high adopter villages as compared to low adopter villages (Hossain et al. 1991). Typically, there are considerable variations in types of share payments, whether in cash, in kind or certain pro-
portions of both. These variations may be due to differences in land quality, crop variety, availability of technology, the relative strength of tenants vis--vis the landlords, and a range of other factors related to the tenancy market. However, at the national level, there is limited information on the extent of these variations and the relative importance of different types of tenancy arrangements. Village and district-level studies suggest that around three-quarters of rented land is under sharecropping. However, DIS data show that, with the adoption of high-yielding varieties of rice, the sharecropping system gives way to fixed-rent tenancy. Information on security of tenure in rural Bangladesh is sparse. In the absence of written agreements, tenants have no security of tenure, and landlords can evict any tenant at any time. According to the 1977 Land Occupancy Survey of Rural Bangladesh (BBS 1977), approximately 70 percent of all tenant households reported operating their rented land for three years or less. More recently, it has been estimated that around 50 percent of rented land is under short-term lease (1-4 years), with about 70 percent being in modern villages (adopting new technological inputs) and about 30 percent in traditional villages (Rahman 1991). With the adoption of new technology, landlords are increasingly cautious and are ensuring their involvement in decision-making through the practice of cost-sharing in the cultivation of high-yielding crop varieties (A. Rahman 1991; Hossain et al. 1991; BIDS/SRP Study 1992).
EMPLOYMENT TRENDS
We can assess the size and growth of the labour force by examining the sectoral composition of the rural labour force and its changes over time, using the population censuses for 1981 and 1991 and recent labour force surveys. Estimates of the rates of participation of the rural labour force, along with the relative share for non-agricultural sectors, are presented in Table 7. In spite of some limitations of the estimates presented in Table 73, the data presented show an increase in the share of the non-agricultural labour force in rural areas, though not as dramatically as has been found elsewhere (Khan and Hossain 1989; Osmani 1990).
Table 7 Proportion of rural population employed and the share of non-agricultural sectors in employed population, 1981-1995/96
Employment status and sex % of rural population employed: Population census Labour force survey
1981 Both sexes Male Female

Share of non-agricultural sectors in employed population (%):
1991 25.9 49.7 3.6
1983-84 28.9 52.3 4.8
1984-85 29.3 53.2 8.8
1990-91 31.8 53.3 8.8
1995-96 34.1 55.4 12.0
25.8 49.5 2.7
Both sexes Male Female
29.3 27.2 67.3
33.9 30.9 76.1
34.6 28.5 89.5
34.4 29.4 88.8
38.6 35.1 61.3
39.6 36.6 54.1
Sources:Various BBS reports on population census and labour force surveys.
Family labour is the main source of labour in farm and non-farm activities. However, a large portion of cultivating households rely on hired labour (Saha 1997). This indicates an emerging trend towards a substantial wage market. The share of labour force in the non-farm sector appears to have increased not only within the rural economy, but also in the economy as a whole.This observation can be substantiated by the fact that the estimated growth rate of the labour force in the rural non-farm sector is much higher than that of both the rural labour force and the countrys total labour force (Mahmud 1996). As discussed earlier, there has been a significant shift from farming to non-farming occupations.Whilst agricultural labour in the farm sector has declined significantly, labour in the non-farm sector has increased correspondingly. It seems reasonable to suggest that the shift of labour out of agriculture can be largely (if not entirely) accounted for by increased landlessness and not by increased numbers of land-owning (or cultivator) households diversifying their sources of income towards non-farm activities (Quasem 1996). Therefore, this shift has mainly taken place at the lower end of the income scale, since there is a strong correlation between land ownership and income (Hossain et al. 1996). Productivity growth and technology uptake in crop agriculture may have facilitated the release of agriculture workers and their entry into non-agricultural activities (Sen 1996).
TENURE, GROWTH, PRODUCTIVITY AND POVERTY

The inverse relationship between farm size and productivity in traditional agriculture is an old hypothesis. Recent studies have noted that this inverse relationship has not changed as a result of the influence of new technologies. However, productivity differences between large and small farms have narrowed (Saha 1997; Hossain 1989). We have already observed the trend towards landlessness and marginalisation along with the proliferation of small farms and relative consolidation of middle farms within the existing unequal structure of land ownership. This structural pattern suggests that the distressed condition of small farms and the subsistence needs of medium farms contribute towards higher productivity. However, this kind of productivity enhancement cannot generate higher savings and investment (Saha 1997). For this reason, the existing growth pattern cannot be sustainable and a route towards poverty reduction. TENURE AND PRODUCTIVITY A key issue is the production performance of sharecroppers vis--vis owner-cultivators. However, existing studies have failed to establish that sharecroppers are less productive than the owner-cultivators (Saha 1978, 1997; Hossain 1981).
The extent of share tenancy in rural Bangladesh has declined, giving way to fixed-rent tenancy and mediumterm leasing arrangements.These institutional changes are assumed to enable tenants to derive some of the benefits of additional investment in agricultural inputs (Hossain 2000). However, Zohir and Sen (1999) find that tenants refrain from investing in fixed-rented land as compared to investments on their own land. Furthermore, fixed-rent tenancy has been observed to be crop-specific and seasonspecific and has been confined to the cultivation of high yielding varieties of rice in aman and boro seasons (Zohir and Sen 1999). Thus, households associated with fixedrent contracts have the lowest crop diversity. Even if productivity is higher on fixed-rented land than on sharecropped land, as is found in Zohir and Sen (1999), this higher productivity on fixed-rented land would not result in any significant influence on the productivity and growth pattern of the country as a whole, since only a small marginal amount (about four percent of total operated land) has so far been brought under this tenurial arrangement. GROWTH PERFORMANCE AND POVERTY REDUCTION There has been a significant positive growth in the nonfarm sector, while the farm sector has seen very slow growth. Of the non-farm sub-sectors, growth performance in the service sector has been the most striking. In the farm sector, the high-income growth of non-crop sectors (e.g. poultry and livestock) is well recognised.The performance of the non-farm sector is based mainly on self-employment opportunities and less on wage employment opportunities. The share of income in the non-farm sector to total household income for both poor and non-poor households has also increased over time. Growth rates for the non-crop and non-farm sectors have not been confined to a particular class or group of households. Higher growth rates for non-farm sectors, along with their increasing share of income to the total household, nevertheless fail to compensate for the loss of agricultural income (including wages) resulting in overall negative growth rates for the whole of rural economy. According to both the Poverty Trend Survey data of BIDS and Household Expenditure Survey (HES) data of Bangladesh Bureau of Statistics (BBS), there has been pos-
itive growth of per capita income in non-farm sectors. The BIDS survey data indicate that positive growth rates in both the farm and non-farm sectors have favourably influenced the growth of the entire rural economy. In contrast, the HES data show negative growth rates in the farm sector, resulting in an overall negative growth rate for the rural economy as a whole. It is now debatable whether and how the slow per capita income growth rates (as found in BIDS survey estimates) and/or negative per capita income growth rates (as found in HES data) can be improved in order to reduce the incidence of rural poverty. The increasing share of income and positive growth rates that feature in the nonfarm sector may be expected to exert an influence on poverty reduction. Yet in the absence of significant growth in the whole economy, significant reductions in rural poverty are unlikely to be achieved.
CONCLUSIONS
Over the past two decades there has been a major shift in rural livelihoods in Bangladesh from agricultural to nonagricultural sources of livelihood. Increasingly, the landless and rural poor are diversifying into non-agricultural activities as a part of their overall livelihood strategy. There has been an increasing trend towards landlessness and marginalisation along with the proliferation of small farms, relative consolidation of medium farms and relative decline of large farms. These changes have taken place within the existing framework of highly unequal land distribution. The extent of agricultural tenancy has increased.The area under share tenancy has declined significantly, as sharecropping arrangements have given way to fixed-rent tenancy and medium-term leasing arrangements. This demonstrates that even without State intervention, agrarian institutions may change under the pressure of market, social and demographic forces.These institutional changes appear to have taken place as a response to the failure of the existing system to ensure a secure livelihood. There has been a significant decline in agricultural labour in the farm sector with a corresponding increase in the non-farm sector. This sectoral shift of the rural labour
force out of agriculture appears to be a survival strategy of the landless and land-poor to escape the trap of poverty and livelihood insecurity. The inverse relationship between farm size and productivity in traditional agriculture does not appear to have undergone any significant change following the adoption of improved agricultural technology. It has been observed that owner-operators are more productive than sharecroppers. Whilst there has been slow growth and/or stagnation in the whole economy, there has been positive growth in per capita income in the non-farm sector.This indicates that non-farm sub-sectors have been confined to low-productivity enterprises with limited growth potential, which are unable to strengthen the growth of the whole economy. In the absence of rapid economic growth, significant reductions in levels of rural poverty are unlikely to happen. Yet, if the non-farm sector can play a more dynamic role and move towards high productivity enterprises that capture the advantage of economies of scale, there would be improved prospects of sustained economic growth. Ultimately, this process may facilitate substantial poverty reduction.
REFERENCES
ABDULLAH, A.A, HOSSAIN, M. AND NATIONS, R. (1976): Agrarian Structure and the IRDP-Preliminary Considerations, The Bangladesh Development Studies,Vol. 4, No. 2, April. BBS (1977): Land Occupancy Survey of Rural Bangladesh, Dhaka. BBS (1982): Statistical Pocket of Bangladesh. Dhaka. BBS (1986): Census of Agriculture and Livestock, 1983-84. Dhaka. BBS (1984): 1981 Population Census of Bangladesh. Dhaka. BBS (1996): Report on Labour Force Survey in Bangladesh. Dhaka. BBS (1999): Census of Agriculture 1996, National Series, Vols. 1 and 2. Dhaka. BBS (1998): Household Expenditure Survey, 1995-96. Dhaka. BIDS (1992): Benchmark and Evaluation Study: System Rehabilitation Project, BWDB. Dhaka.April. BIDS (1996): 1987-1994: Dynamics of Rural Poverty in Bangladesh.Analysis of Poverty Trends Project,Dhaka.April. GOVERNMENT OF BANGLADESH (1981): Report of the Study for the Development of Future Land Policy and Land Reform Measures in Bangladesh, 1978-79. Ministry of Land Administration and Land Reforms, Dhaka. November. GOVERNMENT OF PAKISTAN (1962): Census of Agriculture 1960,Vol. 1 (East Pakistan). Karachi. HOSSAIN, M. (1981): Land Tenure and Agricultural Development in Bangladesh. Institute of Development Economics, Tokyo. HOSSAIN, M. (1986):A Note on the Trend of Landlessness in Bangladesh, The Bangladesh Development Studies,Vol. 14, No. 2, June. HOSSAIN, M. (1989): Green Revolution in Bangladesh: Impact on Growth and Distribution of Income. University Press Ltd., Dhaka. HOSSAIN, M. (2000): Recent Development and Structural Changes in Bangladesh Agriculture: Issues for Reviewing Strategies and Policies.A paper presented at a seminar on Bangladesh Agriculture at the Crossroads: Current Challenges at Centre for Policy Dialogue, Dhaka. 15 July. HOSSAIN, M., QUASEM, M.A.,AKASH, M.M. AND JABBAR, M.A. (1991): Differential Impact of Modern Rice Technology: The Bangladesh Case, BIDS, Dhaka, July. HOSSAIN, M., RAHMAN, H.Z. AND SEN, B. (1996): What has been happening to Income Distribution and Poverty in Rural Bangladesh? (mimeo.). BIDS, Dhaka.
NOTES
1 The 1983-84 and 1996 national agricultural censuses are full-count-cum sample agricultural censuses in Bangladesh to include farm and non-farm households within a specific and similar definition of rural household. A household with a cultivated land of below 0.02 ha is considered to be a non-farm household. 2 1987, Differential Impact Study (DIS), 1991, System Rehabilitation Project (SRP), BIDS. 3 The above estimates are not strictly comparable due to conceptual and methodological differences.The population census data exclude those workers engaged solely in home-based activities, whereas the labour force survey data include them so that the estimates result in an apparently large increase in the size of the female labour force. Moreover, similar definitions of rural and urban areas are not used in the labour force estimates of population censuses and labour force surveys. However, estimates presented in Table 7 are based on conventional definitions of labour force so as to make them consistent and comparable.
HOSSAIN, M. AND SEN, B. (1992): Rural Poverty in Bangladesh:Trends and Determinants, Asian Development Review,Vol. 10, No. 1. KHAN, A.R. AND HOSSAIN, M. (1989): The Strategy of Development in Bangladesh. Macmillan, London. MAHMUD, W. (1996): Employment Patterns and Income Formation in Rural Bangladesh: The Role of Rural Non-farm Sector, The Bangladesh Development Studies, Vol. 24, Nos. 3 and 4. OSMANI, S.R. (1990): Structural Change and Poverty in Bangladesh: The Case of a False Turning Point, The Bangladesh Development Studies, Vol. 18, No. 3. QUASEM, M.A. (1996): Income Diversification in Rural Bangladesh, Research Report No. 149. BIDS, Bangladesh, Dhaka. RAHMAN, A. (1991): Recent Changes in Rental Contracts in Bangladesh (mimeo.). BIDS, Dhaka.August. SAHA, B.K. (1978): Socio-Economic Effects of Technological Change in Agriculture:A Study of Two Villages in Bangladesh, M.Phil. Thesis, Institute of Bangladesh Studies, Rajshahi University, Rajshahi. SAHA, B.K. (1997): Agrarian Structure and Productivity in Bangladesh and West Bengal A Study in Comparative Perspective. University Press Ltd., Dhaka. SAHA, B.K. (2001): Changing Pattern of Agrarian Structure in Bangladesh: 1984-1996, in Q. Shahabuddin and R.I. Rahman (eds.) Bangladesh Economy 2000: Selected Issues. BIDS, Dhaka. SEN, B. (1996): Rural Non-farm Sector in Bangladesh: Stagnating and Residual, or Dynamic and Potential?, The Bangladesh Development Studies,Vol. 24, Nos. 3 and 4. ZOHIR, S. (1999): Land Rental Market in Rural Bangladesh: An Empirical Study (mimeo.). BIDS, Dhaka. ZOHIR, S. AND SEN, B. (1999): Land Tenure and Land Productivity (mimeo.). BIDS, Dhaka.
57
Agricultural and non-agricultural livelihoods in rural Bangladesh: a relationship in flux Kazi Ali Toufique
This paper elucidates the changing relationship between agricultural and non-agricultural activities and assesses the impacts of these changes for rural livelihoods in Bangladesh. Agricultural growth has traditionally been perceived as the main driver for non-agricultural growth via three links.The first, the consumption link, refers to the incremental income gained from agricultural growth that is spent on goods produced by the non-agricultural sector. Second is the production link, resulting from the demand created by agricultural growth for inputs and farm implements produced by the non-agricultural sector. Finally, the backward link refers to the processing of agricultural produce by the non-agricultural sector. However, data from 1983-96 challenge the assumptions inherent in this explanation.As we shall see, the non-agriculture sector grew in a situation where growth in the agriculture sector virtually stagnated. This suggests that the relationship between growth in the agricultural and non-agricultural sectors is in flux; growth in the nonagricultural sector is clearly not solely dependent on agricultural growth. This paper argues that there were two sets of factors at work and, furthermore, that the importance of these factors is even greater today. Firstly, a set of new external links a result of the increasing connectivity of rural Bangladesh to the outside world came into play.These external links were the result of different developments: massive investment in physical infrastructure served to boost the rural economy and facilitate external movement, which led to the increased integration of rural areas within themselves and with the flourishing urban areas; q migration rates grew spectacularly, and with them remittance flows and inter-industry transactions within the non-agricultural sector.
q
low-growth and low-potential. Activities and livelihoods associated with this low-growth, low-potential sector have thrived on stagnating agriculture. They absorbed unskilled workers particularly those who could not find agricultural employment, produced cheap products and services that the agricultural poor could afford. The high-end part of the non-agriculture sector could grow with or without any significant growth in the agricultural sector because of the external linkages. But the low-end part developed a symbiotic relationship with agricultural stagnation. Both sets of factors (new links and differentiation) have been gradually changing the relationship between agricultural and non-agricultural livelihoods. Livelihood opportunities in the non-farm sector are no longer as dependent as they used to be on good agricultural performance. Note that we are not saying that the link has disappeared, it will remain important as long as agriculture continues to exist as a major economic activity. But new evidence suggests that growth in the non-farm sector is due to the emergence of new economic links, which are increasingly important in the livelihoods of rural people. This raises the question of how policy can best serve poor people in maximising the potential benefits of this process.
DEFINITIONS AND KEY CONCEPTS

There is an increasing tendency towards standardisation of concepts and definitions of agricultural activities and non-agricultural activities. Most data define agriculture so as to include crop, livestock, forests and fisheries. Other activities carried out in the rural sector are considered as non-agricultural. Under this heading we include manufacturing, processing, repairing of manufacturing goods, trading activities, transportation, construction and all other service activities done on a commercial basis in the rural economy. The issue now is, what do we mean by rural? The arguments set out in this paper propose a redefinition of rural to mean small towns and population centres with a natural hinterland (Bakht 1996).
The second set of factors relates to the increasing differentiation within the non-agricultural sector. Part of the sector is high-growth, high-potential and the other is
GROWTH IN THE AGRICULTURAL AND NON-AGRICULTURAL SECTORS

The quantity and quality of official statistics on nonagricultural activities in Bangladesh are limited. This is particularly the case for the second half of the 1990s, for which little information is available after 1996, when the last Labour Force Survey (LFS) report was published. The total size of the labour force in Bangladesh is 56 million, of which 10.2 million live in urban areas and the rest in rural areas.Thus about 82 percent of the labour force is employed in rural areas. According to BBS (1999), more than half of rural households is involved in crop and livestock production. These figures indicate that agriculture continues to form the basis of employment in the rural sector. But this paper argues that we are in a transition phase where the relative importance of non-agricultural activities in rural livelihoods is increasing: 34 percent of rural households were involved in non-agricultural activities in 1996 as compared to 27 percent in 1984 (BBS 1999). Similar trends can be observed by comparing population censuses of 1981 and 1991, which show that more and more rural households are taking up non-agricultural livelihoods (Mahmud 1996). The proportion of income derived from non-agricultural activities broadly follows the pattern of increasing employment in the non-agricultural sector (Sen 1996;Toufique 2001).
sector, there is now a general sense of shared optimism1. However, there is less agreement if we take the rural economy as a whole. Recent evidence shows that whilst the non-agricultural sector is thriving, agriculture is stagnating.This does not fit with conventional perceptions of rural economies, which hold that agriculture is the driving force behind growth in the non-agricultural sector. It is critically important to understand the relationships within the rural sector between the agriculture and non-agriculture sectors and to assess the implications of their changing relationship for livelihood opportunities. So what is actually happening? Table 1 provides some information that reinforces the picture of a poorly performing agriculture; agricultural growth in fact declined during the period under consideration.
THE PERFORMANCE OF THE NONAGRICULTURAL SECTOR

The econometric analysis of HES data suggests that a shift from agriculture to non-farm occupations entails significant income gains among households with similar land ownership and other characteristics. Hossain et al. (1994) found that self-employment in activities such as rickshaw pulling and petty trading generated more income than did jobs in agriculture. Unfortunately there is little information available on wage rates in non-agricultural activities but what evidence there is does show that wages in non-agricultural activities are generally higher than agricultural wages (Toufique and Ahmed 2001; Bakht 1996)2. Sen (1996), again from an analysis of HES data, found that:
IN SEARCH OF MISSING LINKS

A common trait in most literature on the rural economy is that they express either optimism or pessimism (Faruqee 1998). Considering only the non-agricultural
Table 1 Selected indicators of agricultural performance in Bangladesh

Annual rate of growth in 1972-73/ 1994-95 1972-73/ 1979-80 1980-81/ 1989-90 1990-91/ 1994-95
Agriculture sector Crop sector Fertiliser use Total irrigated area Rice production Rice yield
1.98 1.99 9.23 4.29 2.38 2.23
1.85 2.37 13.13 4.27 2.80 2.28
1.83 1.77 9.98 4.05 2.32 2.83
0.86 -0.68 6.63 2.56 -1.30 -0.02
Source: Computed from Shahabuddin and Rahman (1998)
Agricultural and non-agricultural livelihoods in rural Bangladesh: a relationship in flux 59
for the poor as well as for the non-poor, agricultural income as a proportion of total income has declined steadily; q the proportion of non-agricultural income is higher for the non-poor; q the proportion of non-agricultural income to total income is high and increasing for both poor and nonpoor, implying that both groups participate in the nonagricultural activities.
q
poverty has been positive a move from agriculturebased livelihoods to non-agricultural livelihoods will tend to reduce overall levels of poverty.
THE NATURE AND IMPLICATIONS OF DIFFERENTIATION

The set of factors at play in the rural economy today relate to the structure of the non-agricultural sector and the nature of its growth. If we wish to support growth and build on existing strengths, we must identify who is benefiting from the growth, who is losing and the pathways through which winners and losers are created. This will help to suggest more supportive policies to maximise the benefits of the changing rural economy.
Unfortunately the above findings are based on data that omit wages and salaries, which account for a significant proportion of poor peoples incomes but are not classifiable as either agricultural or non-agricultural. But since wages from non-agricultural employment are generally higher than wages in agricultural employment, we can say that the non-poor earn more from nonagriculture based livelihoods. Others have made similar observations. In a household study of eight villages in 1995,Toufique (2001) found that the proportion of nonagricultural income to total income fell as land holding increased.This was also true when household self-assessment of poverty ranking was used.The lower the level of poverty as self-assessed by the respondents, the higher was the proportion of non-agricultural income to total income. In all these cases, no distinction was made between wage and non-wage income. Sen (1996) concludes that: for the land-poor, a shift to non-agricultural activities reduces poverty; q poverty has declined in service, industry and trade within the non-agriculture sector; q workers engaged in non-farm activities contribute more to household income as compared to agricultural workers.
q
DIFFERENTIATION IN THE STRUCTURE OF PRODUCTION

The first observation is that the more capital used per worker, the higher the workers productivity. The Rural Industries Study Project, which BIDS carried out in 1978-80, found that industries that used large amounts of fixed assets per worker and had a higher proportion of units in thana centres and semi-urban areas demonstrated higher productivity. In contrast, industries located in villages that used very little capital per worker and catered mostly to rural demand lay at the lower end of the productivity scale.This relationship between input and productivity can be observed within a given industry as well as across industries. Bakht (1996) makes the following observations: labour productivity is higher in small industries compared to that in household industries; q labour productivity is higher in urban industries compared to that in rural industries; q labour productivities in cottage industries, rural household-based industries, and even urban household-based industries are lower than agricultural industries.
q
We do not have any information on the amount of remittances that are entering the rural economy.A recent study by Khan and Sen (2001)3 has come up with some figures on informal transfers that include remittances and gifts. The figures are taken from Household Expenditure Surveys. In 1991-92, informal transfers represented ten percent of rural income, whereas in 1995-96 it dropped slightly to about nine percent. Thus we see that non-agricultural income is generally higher than agricultural income. However, its impact on
Bakht (1996) concludes that overall productivity in the 1980s remained almost unchanged. This was particularly true for a long list of activities. Only a limited segment showed higher productivity, higher capital-labour ratio, greater use of hired labour and locating in semi-urban areas. However, there is a positive relationship between
growth of an industry and labour productivity. Bakht (1996) compared productivities of 151 disaggregated industries for two years for which information was available (1980-81 and 1989-90). About one-third of these industries experienced positive growth in labour productivities. Thus, the higher the inputs, the higher the productivity. Enterprises are stratified; household enterprises employ more women, unskilled, cheaper labour, cost less to set up, serve the local population and are least productive. Small businesses in small towns and rural centres demand higher skill levels and pay more for them, serve urban and peri-urban populations, employ fewer women and are more productive. Table 2 summarises the various aspects of differentiation that exist in the non-agricultural sector.
landless and functionally landless households (<0.2 ha) are heavily involved in transport, rural industry, construction and trade. Their participation in the formal and informal services sector is limited; q marginal and small farmers (0.2-1 ha) are more involved in the formal and informal services sector. They are also, to some extent, involved in construction and services (formal and informal); q medium and large households (>1 ha) are more involved in the formal and informal services sector.
q
Thus there is a differentiation of participation in the nonfarm sector. Medium and large landlords do not really participate in any activity other than services (they do not participate at all in transport and construction), whereas the landless participate heavily in activities where the large and medium land holders hardly intervene. The skill requirement in the service sector is high about two-thirds of the households have secondary and above secondary levels of education. On the other hand, in the transport and construction sector, about two-thirds of the households have no formal education. Data presented by Sen (1996) show that most of the participants who are land-poor are also skill-poor, whereas most of the households who are land-rich are also skill-rich. Hossain (1988) studied non-agricultural employment in two sets of villages: progressive villages with developed agriculture and infrastructure; and backward villages with less-developed agriculture and infrastructure. He found the share of non-agricultural income to total income higher than that of non-agricultural employment to total employment, except in underdeveloped villages and for the low-income groups (landless, marginal and small farmers).Thus, in developed areas, the proportion of income from non-agricultural activities is higher than the proportion of non-agricultural employment for land holding categories. This implies that, in the developed areas, income from non-agriculture activities was higher than income from agricultural activities. In the underdeveloped villages, this is true only for medium and large farmers but not for landless and small land holders. The landless and small land holders provide proportionately more labour to the non-agricultural sector in return for proportionally less income.
Table 2 Differentiation in the non-agricultural sector

Indicators of differentiation High-end industries Low-end industries
Capital intensity High Labour productivity High Wages/Returns High Participation of women in the work force Low Main consumers Urban, peri-urban by location Main consumers Rich by income status Income Elasticity Medium to High Skill requirement High Set-up capital requirement High Geographic dispersion Limited Type of establishment Permanent
Low Low Low High Rural Poor Low Low Low Dispersed Household
DIFFERENTIATION OF PARTICIPANTS IN THE NON-AGRICULTURAL SECTOR

We have limited information on the make up of nonagricultural workers. Differentiation is observed in asset ownership (most visibly in land ownership) and in skill (education or human capital) differential. Sen (1996) makes the following observations:
Table 3 Demand for various goods according to income group

Indicators High-income/land group Low-income/land group
Average and marginal budget share of expenditure on non-farm goods Average and marginal budget share of expenditure on urban and imported goods Average and marginal budget share of expenditure on food (cereals and staples)
High High Low
Low Low High
The implication of these findings is important.The poor are also skill-poor; they are mainly involved in low-wage residual employment and a large proportion of their income is derived from participating in the non-agricultural sector. It is precisely this part of the non-agricultural sector that experienced extensive growth over the last decade.
POLICY RECOMMENDATIONS
The general guideline is simple: we need to strengthen the links within the rural sector between agriculture, non-agriculture and the missing third dimension the external sector (migration, urbanisation, etc.). Broadly speaking, we need investment in infrastructure, education, credit, agro-processing and at the same time we have to strengthen the system of local governance. These are the entry points to improve livelihoods based on non-agricultural activities, migration and urbanisation. Investment in infrastructure will encourage specialisation and division of labour. It will expand the exchange of inputs and outputs between villages and small towns, urban areas and possibly export markets.Trade, commerce and marketing will also thrive. Investment in education will enhance productivity, improve entrepreneurial capabilities and help migrants to secure better jobs at destination. This will also reduce differentiation in the non-agricultural sector to some extent because, as we have seen, participation in the non-agricultural sector is often constrained by lack of appropriate skills. Investment in agro-processing (such as Bangladesh Rural Advancement Committee (BRAC) dairy production) will create backward linkages. Similar results come from experience of Grameen Uddog. These are the demandside constraints that have to be relaxed for the growth of the non-agricultural sector. Self-governing local governments may stimulate growth of the non-agricultural sector by facilitating the development of physical, social and human infrastructure at the local level. For this, the local governments need decision-making, implementing and financial powers.
DIFFERENTIATED DEMAND
Table 3 shows that the consumption demand for nonfarm goods is likely to come from large and medium farmers (and also high-income groups) and not through the demand of the poorest or marginal farmers (or lowincome group). Rural-produced consumer goods have a niche market in the rural areas because of their lower price, even though they are of inferior quality compared to urban consumption goods. Rural non-farm goods are protected not only by specialised rural demand for such goods, but also by transport costs involved in marketing in rural areas (Islam 1997). Thus we see that, whereas the high-growth part could grow without commensurate growth in agriculture, the low-growth remained reliant on what became a stagnating agriculture. The low-end part could absorb a large number of unskilled labourers who could not find employment in the agricultural sector. It required less capital and hence labour productivity and wages were low. It produced low-quality products which a stagnating agriculture could afford to buy. This weakened the traditional link between the agricultural and the non-agricultural sectors.
NOTES
1 Some early scepticism on the growth potential of the non-agricultural sector was, however, expressed by Osmani (1990), amongst others. On the basis of data pertaining to the mid-1970s and early 1980s, he claimed that: (i) the shift of labour from the agricultural to non-agricultural sector increased work-sharing in the latter and therefore labour productivity declined; (ii) non-farm workers were poorer than their agricultural counterparts and (iii) declining productivity increased poverty in the non-agricultural sector. It is now generally agreed, however, that many of these fears were unwarranted. 2 The wage gap could be even more significant if we take account of risk and seasonal variation.Agricultural wages fluctuate strongly over seasons and according to production levels.Therefore the real risk-adjusted wage differential could be even higher. 3 I am grateful to Dr Binayak Sen for sharing this information and for the follow-up discussion I had with him.
REFERENCES
BAKHT, Z. (1996):The Rural Non-farm Sector in Bangladesh: Evolving Pattern and Growth Potential, The Bangladesh Development Studies, 1996, 24 (3 and 4), pp.29-73. BBS (1999): Census for Agriculture 1996, National Series. BBS, Dhaka. BIDS (1996): 1987-1994: Dynamics of Rural Poverty in Bangladesh. Analysis of Poverty Trends Project, Dhaka, April. FARUQEE, R. (ed.) (1998): Bangladesh Agriculture in the 21st Century. University Press Ltd., Dhaka. 1998. HOSSAIN, M. (1988): Nature and the Impact of the Green Revolution in Bangladesh, Research Report 67. IFPRI, Washington, D.C. HOSSAIN, M., RAHMAN, M. AND BAYES, A. (1994): Rural Non-Farm Economy in Bangladesh: A Dynamic Sector or a Sponge for Absorbing Surplus Labour?, SAAT Working Papers, ILO-SAAT, New Delhi. ISLAM, N. (1997):The Non-Farm Sector and Rural Development: Review of Issues and Evidence, Food, Agriculture and the Environment Discussion Paper 22. IFPRI, Washington, D.C.,August, 1997. KHAN, A.R. AND SEN, B. (2001): Inequality and Its Sources in Bangladesh, 1991-92/1995-96: An Analysis Based on Household Expenditure Surveys, The Bangladesh Development Studies,Vol. 27, No. 1, March, 2001. MAHMUD, W. (1996): Employment Patterns and Income Formation in Rural Bangladesh, The Bangladesh Development Studies, 1996,Vol. 24, Nos. 3 and 4, pp.1-27. OSMANI, S. (1990): Structural Change and Poverty in Bangladesh: The Case of a False Turning Point, The Bangladesh Development Studies, 1990, 55-74. RAVALLION, M. AND SEN, B. (1995): New Evidence on Agricultural Wages in Bangladesh?, Policy Research Department: Poverty and Human Resources, World Bank, Washington, D.C. SHAHABUDDIN, Q. AND RAHMAN, R. (1998):Agricultural Growth and Stagnation in Bangladesh, MAP Focus Study Series, No. 6, July, Centre on Integrated Rural Development for Asia and the Pacific (CIRDAP). TOUFIQUE, K. (2001): Sustainable Rural Livelihoods in Bangladesh, IDS Research Report 45, Institute of Development Studies at the University of Sussex, UK, September, 2001, pp.1-132. TOUFIQUE, K. AND AHMED, N. (2001): Livelihoods and Institutional Assessment for DFID Bangladesh: Further Processing of the Fieldwork Information. Dhaka.
YUSUF, S. (1996): The Non-farm Road to Higher Growth: Comparative Experience and Bangladesh Perspective, The Bangladesh Development Studies, 1996, Vol. 24, Nos. 3 and 4, pp.195-248.
65
Frontiers of change in rural Bangladesh: natural resources and sustainable livelihoods Mohammed Asaduzzaman
Major changes have occurred in the economy and society of Bangladesh over the past 30 years. GDP has increased, albeit slowly at times and, during the last few years, there has been cautious optimism regarding future upward trends. Crop production, despite fluctuations, has risen to keep pace with population growth, thanks to technological change. More recently, livestock and fisheries output has also increased. Despite such positive trends, however, there is little room for complacency. Poverty remains widespread; while the proportion of people living in poverty may have declined, the absolute number of poor people continues to grow as a result of increasing population. Food security at the aggregate level may have been achieved, but the prospects for sustaining the tempo in the future are uncertain. Future uncertainties are inextricably linked to the process of globalisation, which will shape the development of natural resource-based livelihoods in the future. Formally and operationally, globalisation is embodied in the rules of trade and investment established by the World Trade Organization (WTO). Briefly, the rules will allow only limited support to agriculture (through subsidy and other means) and few trade barriers.Bangladesh pared down subsidies during the 1980s and has already liberalised her trade substantially. In short, the implications are that agriculture in Bangladesh, and other economic activities that are linked to agriculture, will have to be efficient to survive. The challenge for policy-makers over the coming decade is to support those poor households whose livelihoods depend on natural resources to adjust to the new context and take advantage of new opportunities in the natural resources sector.This does not mean preserving the status quo.What it does mean is ensuring that poor households can make the adjustment from a low-level subsistenceoriented and partially closed economy to a medium-level, market-driven, open economy. This paper begins by outlining the nature and potential of the natural resource base in rural Bangladesh. It then looks at the organisational and technical nature of production. This is followed by an examination of some of the opportunities for improving the productivity of natural resources. Finally, an assessment is given of the potential of natural resources in relation to future livelihood opportunities.
RURAL LIVELIHOODS AND THE STATUS OF NATURAL RESOURCES

Agriculture remains the major sector of the economy and is critically dependent on the management and exploitation of natural resources through crop production, fisheries, livestock and forestry. Much of the industrial sector also depends on the processing of agricultural products, while a significant part of the services such as trade and transport provide marketing services to agriculture. Natural resources thus play a key role in the economy and the livelihoods of rural people. CROP PRODUCTION Despite the declining trend, crop production remains the main economic activity. Two types of natural resource land and water play the most important role in crop production and, for both resources, quantity and quality are changing. The increasing population has substantially lowered the ratio of land to people and, more importantly, has resulted in land-use moving from directly productive purposes, such as crop cultivation, to other uses such as housing and roads and urban development. While official statistics sometimes portray a conflicting picture, this loss of cultivable land is a reality as repeat visits to a village over successive years testify (BBS 1999, 2000).This trend of land loss for homesteads and other purposes is expected to continue in the medium-term future. Families will continue to divide their own homesteads; new schools and roads will have to be built and this will take more land away from crop production. One result of this process is a trend towards smaller farm sizes. There is not only less land available for crop production, but the land that is available is of lower quality. Despite a
lack of firm estimates, analyses indicate annual production losses might be as high as one percent of the crop production contribution to GDP (Asaduzzaman and Toufique 1997), with consequent losses elsewhere in the economy through linkage effects. If present trends continue, the quality of land is likely to fall further and, with it, the productive capacity of the land. But this need not be so; the experiences of other countries indicate that there is potential for doubling the present yield. Unfortunately, two factors militate against reversing the present trend. Firstly, a lack of awareness of the gravity of the problem at the policy-making level and its immediate and ultimate causes (both scientific and societal). Secondly, the lack of any integrated and general programme for improving soil quality (Roy and Quasem 2000). There are serious concerns too over the management of water resources. Bangladesh enjoys one of the highest rainfall regimes in the region but awareness has dawned rather slowly that managing the competing uses of water is not an easy task. In fact, until recently, all major interventions in the water sector generally had the sole purpose of increasing crop production; there was little reference to other water uses. Bangladesh has yet to develop a water-management plan that takes the various competing demands for water into consideration. The regional distribution of available water, both surface and ground, is uneven and regions also vary greatly in their susceptibility to flooding. For example, completely flood-free land ranges from 57 percent of the total cultivable area in the northwest, to only 23 percent in the south and central districts (Shahabuddin 2000).There is a similar unevenness in the regional potential for the development of irrigable land. Of a total potential irrigated area of 6.9 million ha, 2.32 million ha of this area (34 percent) is in the northwest and two million ha (29 percent) is in the northeast. Much of this potential, however, remains to be exploited (Shahabuddin 2000). In fact, the spread of irrigation has been slowing down since the mid-1980s (Shahabuddin 2000).This appears to be due to economic factors; deep tube wells (DTWs) are the only technical means of irrigation in many areas yet, in the absence of subsidisation, they are too costly to
operate compared to shallow tube wells (STWs) and lowlift pumps. A final important point in relation to water resources is arsenic contamination. In addition to the problem of safe drinking water, there is the issue of the contaminated water entering the food chain through crops or their byproducts fed to livestock. The source of water for irrigation is therefore critical.The shift to the use of surface water for irrigation, in response to arsenic contamination of ground water, has both cost and coverage implications as surface water availability is more location-specific than is ground water. In either case, livelihoods will be affected due to potential rises in the cost of food and loss of employment opportunities. FISHERIES A balanced development strategy for water resources is critical for the development of the fisheries sector. Much of the social discord generated during earlier development interventions in the water sector centred on the issue of vanishing water bodies inside the flood-protected areas, the silting up and dying of small rivers and creeks and the loss of the flood plain for fisheries. Indeed, conventional economic criteria of cost-benefit analysis have indicated that many projects are not viable if fisheries losses are considered along with the gains from crop production. This loss refers to that from open water fisheries, which in many cases were common property resources accessed by the poor for subsistence fishing, and by traditional fishermen. In some cases, culture fisheries in ponds and beels1 leased from the Government have filled the gap. However, only pond owners and those who can lease in the beels and other water bodies benefit. Due to the high elasticity in the demand for fish, capture fisheries production is expected to continue to increase.The question is how far this development takes place in an equitable manner. LIVESTOCK The density of cattle and buffalo head per hectare in Bangladesh is 2.6, one of the highest in the region (Asaduzzaman 2000). The density has remained stable since 1988, when the last livestock survey was undertaken, and densities are unlikely to change in the short to medium term, barring some catastrophe. However, changes might be expected in the quality and composition of the herd.
Frontiers of change in rural Bangladesh: natural resources and sustainable livelihoods 67
This is particularly true for the case of large ruminants, which have traditionally made an important multidimensional contribution to the household economy. They provide draught power for tillage, dung for use as fertiliser or fuel, milk and meat for direct consumption and/or sale and hides and bones for sale. They are also used for transport. Generally, the bullocks are used for draught and transport and cows for milk. As tillage takes place only 2-3 times a year, keeping a pair of bullocks all year round may become uneconomic. Bullock exchange, or using cows for tillage, has therefore been a frequent strategy. In the last decade a new option has emerged: the use of power tillers.While comparatively few households own power tillers, many hire them for short periods. In 1996, 88 percent of the land tilled using power tillers was tilled with hired machines (BBS 1999). The upshot of this is that farmers are increasingly turning away from keeping bullocks and are more likely to keep cows for milk. The data show that small farmers are more likely to use power tillers because they cannot afford to keep draught animals.The herd composition (ratio of males to females) for different farm sizes reflects this: the ratio is 1.62, 0.9 and 0.82 for small, medium and large farmers respectively. Small farmers are, therefore, twice as likely to keep adult female cows.This offers potential for a vigorous livestock development programme that can benefit small farmers. As the interest of small farmers in keeping milk cows has increased, so has the dairy industry grown. By 1998-99, nearly 34,000 private dairy farms had been established (MoF 2000). Another aspect of livestock that needs careful attention is that of poultry and small ruminant development. By 1999, 116,000 poultry and duck farms had been established (MoF 2000: 55) and any visitor to the market in Dhaka would have been struck by the growth of vendors selling broiler chickens, slaughtered and dressed upon request. Shops are supplied by the hundreds of poultry farms that dot many of the major highways from Dhaka. FORESTRY Two forms of forestry need to be considered: (i) public forests managed largely by the public authorities and mainly for the purpose of growing timber (with firewood
as a by-product); (ii) village forests, which to be more accurate are not true forests, but tree cover. Of the total land designated as forest, 90 percent is public forest land and the rest is village forest.As the products of public forest land are available only through a licensing system, people have to depend on fuel wood from the village forest. As the population increases and the demand for fuel wood rises, it is likely that people may increasingly turn to tree plantations. But, given that there are trade-offs between output from the trees and benefits from other uses of land, the trend towards increased village forestry is unlikely to be strong in the near to medium-term future.
FUTURE DIRECTIONS IN THE USE OF NATURAL RESOURCES

The previous section outlined the utilisation of natural resources. The future directions of exploitation will depend largely on how rural people interact with the resource base: both the way production is organised and how technology is applied. Both are becoming important factors in determining the nature and limits of change. As crop production remains the major economic activity, much of the discussion focuses on crop agriculture. ORGANISATION OF PRODUCTION Land is by and large held as private property. Ownership laws have been well defined and the core elements of the present system of land administration have been in place for the last 400 years. From the early 1980s to the mid-1990s, the overall numerical dominance of marginal and small farms increased at around 2.8 percent per annum while the number of medium and large farms fell (Saha 2001).The proportions of land owned and operated by the respective groups show a similar pattern. There is a substantial regional variation in the proportion of holdings by size: the proportion of small farms varies from 43-93 percent, that of medium farms from 6-47 percent and of large farms from practically nil to 10 percent2. Since 1983, there has been little change at the aggregate level in the tenurial distribution of farms. The reduction in size of farm, coupled with no change in tenurial distribution, indicates intense competition in the land market, even for tiny pieces of land. Such changes corroborate the earlier observation that, rather than holding on to tiny pieces of
land, people may part with them against the assurance of a certain return. In some areas it has been noted that fixed-rent tenancy is gaining prominence in place of share tenancy (Saha 2001). Who among the farmers are losing or gaining in the process of land transactions? Both on the basis of changes in patterns of ownership and operation, it may be concluded that there is a redistribution of land from the higher to the lower size categories, a kind of land pauperisation. This is a result of both tenurial transactions and, perhaps more importantly, due to the subdivision of land among sons. For other resources, issues of access and ownership are less clear.Water is largely a common property resource. Open water bodies such as rivers, canals and beels are under public ownership, but with access that is sometime free and sometimes regulated through a licensing system.The ownership of fisheries or fishing rights depends upon the rights of use of specific water bodies. Fishing rights are reserved in private fishing ponds, which are now becoming more common. Ground water is owned by the public authorities but can be almost freely exploited by the people, although restrictions have been imposed at one time or another through control over number of tube wells that can be sunk within a given area. Formal forests come under the domain of public authorities in most cases, whereas the village forest remains the preserve of citizens. What do these ownership patterns mean for livelihoods? People, including the poor, clearly have certain customary rights over common property resources. However, the tragedy of the commons situation is far too often a reality and in such a free-for-all, the wealthier and more powerful groups often gain access to the resource at the expense of the poor.Two key areas for policy to focus on include the establishment of institutions that ensure the poor a fair share of common property resources and support to groups of poor households to enable them to bid for access to resources for establishing use rights. TECHNOLOGY Technological change has played a crucial role in the development of agriculture, particularly food grain production, over the last 25 years. The advent of High Yielding Variety (HYV) technology in rice production
allowed peasants to feed the country, generate employment and income and, through linkage effects, generate employment and income in the rest of the economy. True, it probably did not deliver all that it promised and technological advances in Bangladesh are still modest compared to those other countries but the blame falls on social and economic forces that prevented the full potential of technological change to be realised, rather than on peasants. So what scope is there for technology to deepen development efforts? Mechanisation will continue to increase in importance. There are important regional variations in the adoption of technology in these areas: while about 27 percent of net cultivated land is ploughed mechanically, in certain districts the percentage is as high as 72 percent (BBS 1999). q A similar situation pertains in case of water use technology, particularly irrigation. The net sown area irrigated is just about 45 percent whereas, by district, the proportion varies from 1-93 percent (BBS 1999). q Bangladeshi farmers use less fertiliser per unit of land than their neighbours. q New production technologies, such as super rice, may be introduced. This is still, however, one or two years away from being released and even then it will need to be tested in farmers fields. So it may be up to five years before such new technologies can be adopted on a large scale. q The application of technology to other crops beyond food crops. q Processing and marketing improvements there is great scope here as the current marketing of crop and other agricultural products results in substantial losses of output and a loss of income for producers.The application of technology to livestock and fisheries products also has potential.
q
Overall, technology has much to offer in terms of increasing the productivity of natural resources.The need, however, is not so much for exotic technology, but to disseminate existing and proven technology and address the organisational aspects of production. In terms of the latter, information technology should be used as far as possible. While there are certain infrastructure-related constraints, these may be overcome, given enough ingenuity: tele/information centres in various parts of the world
Frontiers of change in rural Bangladesh: natural resources and sustainable livelihoods 69
have allowed small producers to get up-to-date information on technology, weather conditions and markets.
FRONTIERS OF RURAL CHANGE

This section draws together the issues of natural resource status, the nature of production and the dimensions of future demand to identify areas of likely future opportunities in relation to natural resources. Several criteria need to be considered: the present and future demand; equitability issues how to ensure maximum benefits to the maximum number of people, either directly or through linkage effects; q location issues how far benefits are location-specific or independent of location; q the extent to which the present system can readily respond to the immediate needs for development of these activities; q the technological opportunities.
q q
made and a marketing support system can be developed, poor people may be able to benefit from cultivation of some of these products; vegetables, for instance, can be grown in small plots.This mode of production would be feasible all over Bangladesh and could therefore bring widespread benefits. The linkage effects are substantial, particularly as many of the crops need to be processed. The need for technological change is fairly high. Such changes mainly involve more optimal use of currently available production techniques to avoid further ecological damage, but may also necessitate a switch to more sophisticated technology. Marketing channels would need to be developed particularly if export markets are to be exploited (technological development, particularly in packaging, and changes in collection, processing and dissemination of market information may be necessary). Livestock and poultry hold probably the most potential for the most people, as the demand for products is highly income-elastic (i.e. demand is responsive to changes in income). Positive changes have already occurred; small farmers, particularly, are taking advantage of dairy development opportunities, with few adjustment costs. On the other hand, for a real dairy revolution, far more sophisticated processing and marketing development will be needed.This will entail fairly high investments in organisation, technology for chilling plants and pasteurisation
Table 1 shows which natural resources offer the most potential. There are several opportunities for crop production. Some crops are in high demand and have an export market. Provided that proper technological changes can be
Table 1 Criteria for prioritising natural resource management

Criterion Crop production Livestock and poultry Fisheries Forestry
Demand Export potentials Equity Location-specificity Processing needs Flexibility Technological change Marketing needs Adjustment costs Institutional needs NGO involvement Ecological benignness
Overall priority
Low to fair Fair Low to fair Low Fair to high High Low to fair Low to fair Low to fair Fair to high Low to fair Low to fair
2
High Fair to high High Low Fair to high High Fair to high High Fair to high Fair to high High Fair to high
1
High Fair High High Fair Fair Fair to high High Low Fair to high Low to fair Low to fair
3
Fair Low Low to fair High Fair Low Fair to high Fair Low Low Low to fair High
4
and specialised transport equipment.As the experience of the Anand dairy in India has shown, the pay-off can be very high. Poultry production also has potential both domestically and internationally. However, if the poultry business is to gain a foothold in the export-market, far more stringent hygienic and other technical requirements will have to be met. These will necessitate institutional development of a far higher order, but lessons can be taken from the experiences of the shrimp industry in Bangladesh (Toufique 2001). Compared to crop and livestock, fisheries and forestry development stands to benefit a comparatively smaller group of people because of location specificity and smaller linkage effects3. The rankings in Table 1 may seem somewhat subjective, particularly as the criteria have been set subjectively. Indeed, it is not argued that fisheries and forestry need not be developed, but that their overall scope for sustaining livelihoods for the majority of people is limited compared to other sectors.
REFERENCES
ASADUZZAMAN, M. (2001a): An Employment Strategy During the Sixth Five-Year Plan of Bangladesh (final draft), unpublished.A report prepared for the Bangladesh Planning Commission. ASADUZZAMAN, M. (2001b): Bangladesh Agriculture in the Era of Globalisation: Constraints and Opportunities. under print as part of BIDS Review of the Economy, 2000. ASADUZZAMAN, M. (2000): Livestock sector, economic development and poverty alleviation in Bangladesh in M.A.S. Mondal (ed.) Changing Rural Economy of Bangladesh, Bangladesh Economic Association. 2000. ASADUZZAMAN, M. AND TOUFIQUE, K.A. (1997): Rice and Fish: Environmental Dilemmas of Development in Bangladesh in CPD, IRBD 96: Growth or Stagnation, University Press Ltd., Dhaka. 1997. BBS (2000): Year Book of Agricultural Statistics of Bangladesh, 1998, Statistics Division, Ministry of Planning, Dhaka. BBS (1999): Census of Agriculture,Vols. 1 and 2, Dhaka. MOF (MINISTRY OF FINANCE) (2000): Bangladesh Economic Review 2000 (in Bengali), Dhaka. ROY, P.K. AND QUASEM, M.A. (2000): Land Degradation in Bangladesh: Its Extent, Causes and Policy Suggestions, unpublished draft. SAHA, B.K. (2001): Changing Pattern in Agrarian Structure in Bangladesh: 1984-1996 in BIDS Review of the Economy, 2000 (under print). SHAHABUDDIN, Q. (2000): Position Paper on Water Resources Development for The Sixth Five-Year Plan (2002-2007), unpublished draft. General Economics Division, Planning Commission. TOUFIQUE, K. (2001): Sandwiched between us and them: dilemmas and contradictions in the shrimp-processing export sector of Bangladesh in the globalisation process, South Asia, Special Issue on Regional Responses to Global Economic Changes: West Bengal and Bangladesh, New Series, Vol. 24, No. 1, June, 2001, pp.185-200.
CONCLUSIONS
Natural resources will remain a major factor in ensuring and sustaining rural livelihoods in the next decade. However, the management of the resources will need to develop, both technologically and institutionally. More efficient and effective use of natural resources can lead to substantial improvements in employment and earnings of poor and small farmers. But that is the direct effect.The consequent agro-processing and marketing activities will result in further employment and earnings for rural and semi-urban people.This potential will be more profitable and sustainable if the country can take advantages of the processes of globalisation and specifically some of the rules of the present trading system and the envisaged changes (Asaduzzaman 2001b).
NOTES
1 A beel is an area of low-lying land that is seasonally or perennially inundated. 2 Estimated by the author on the basis of information in the BBS (1999; Vol. 1). 3 Asaduzzaman (2001a) estimated the employment multipliers for various sectors per unit of output.These (in person years/mn taka of output) are, rice: 47, livestock: 99, poultry: 79; shrimp: 22; other fish: 31; forestry: 18.
71
Rural development policy in Bangladesh Sattar Mandal

Agriculture and rural development are generally intertwined in government policy formulation as well as in academic discourses in Bangladesh. Although one can differentiate between these two at a conceptual level, they have been virtually synonymous in the traditional Bangladesh, a country with a large rural population that is directly or indirectly engaged in a wide range of agricultural and non-agricultural activities. Bangladeshs successive Five-Year Plans have continued to emphasise poverty alleviation as their core objective and have underscored agriculture and rural development as the main means of reducing widespread rural poverty. Agriculture contributes about a quarter of the countrys GDP, accounts for one-third of export earnings and provides employment to two-thirds of the civilian labour force (MoA 1999). In the past three decades, however, there have been significant changes in policies related to agriculture and rural development that have spawned numerous governmental and non-governmental programmes. These included policies and programmes on rural institutions, the distribution of agricultural inputs, food assistance to the poor, micro-credit operations as well as innovative commercial ventures by the private sector. This paper attempts to provide a brief historical analysis of rural development policies in the post-liberation period after 1971, mainly with reference to cooperatives, mechanised irrigation, farm power, fertilisers, rural credit and non-farm enterprises. The intention is to highlight how the broad features of the policies have led to a number of recent initiatives and actions by public, private and NGO sectors, which we believe have immense potential to accelerate rural development and stimulate the rural economy. Major concerns and policy options are also suggested in the paper. masses of unemployed and under-employed rural workers and for creating physical and social infrastructure to support production. The first Five-Year Plan (1973-78) envisaged the building block of Bangladeshs rural development as broadening the base of farmers co-operatives Krishak Shamabaya Samity (KSS) which comprise small and marginal farmers, landless labourers and resource-poor men and women. In line with the basicneed strategy of the first Plan, agricultural policy promoted new seed-based technologies in order to boost productivity and attain food-grain self-sufficiency.To this end, the Bangladesh Agricultural Development Corporation (BADC) a public-sector agency was given sole control over the distribution of heavily subsidised chemical fertilisers and irrigation pumps. Although BADC played a key role in popularising these technologies, the growth of food production was not fast enough to mitigate the chronic food deficit.This prompted the Government in the early 1980s to opt for large-scale privatisation of agricultural inputs. The other important policy of this period relates to land reform.Although this resulted in the release of some land to landless and marginal farmers, implementation of the land reform policy failed due to administrative weaknesses, inadequate financial resources and resistance from vested interests. A decade later, the Land Reform and Agricultural Labour (Minimum Wages) Ordinance of 1984 fixed a ceiling on the maximum land holding, this time varying according to the quality of the land, but it amounted to little more than political sloganeering (Elahi and Mandal 1991; Rahman 1998). The second (1980-85) and third (1985-90) Plans sought to drive rural development by increasing labour and farm productivity through the intensification of technology packages. This was coupled with local-level planning, privatisation of agricultural input supplies and rural infrastructure.The fourth Plan (1990-95) stressed participatory bottom-up planning with a view to engaging the rural poor with the development process more effectively. But it could hardly be put in practice due to poor local
RURAL DEVELOPMENT POLICIES IN FIVE-YEAR PLANS

Each of the countrys Five-Year Plans have prioritised rural development as a means of increasing food production, generating employment opportunities for large
government institutions, scant funding of local government and the absence of any effective means to integrate local needs and priorities with national development plans.The on-going fifth Plan (1997-2002) also prioritises poverty alleviation and rural development.The fifth Plan adopts a policy of partnership, seeking to engage mainstream government agencies as well as NGOs and the private sector through support to technical inputs, skill training, infrastructure and credit. Common to all of the Five-Year Plans is the goal of attaining self-sufficiency in the production of food grain, (especially rice).The impetus for this has been the persistent annual deficit of 2-3 million tonnes of food grain. In a sense, it was asking a lot of the public sector to manage tasks such as input distribution which, left to the private sector, presents much less of a problem. The unilateral emphasis on cereal production also led to a rapid reduction of non-cereal crops such as pulses and oilseeds, which have consequently reduced the availability of protein to the poor. Development of non-crop enterprises, i.e. fisheries, poultry and livestock, was neglected by earlier Plans. In fact, inland capture fishery has suffered considerably as open water bodies have dried out, a result of the expansion of pump irrigation for the cultivation of boro rice. These sub-sectors were highlighted as the thrust sectors in the policy and programmes of the fourth and fifth Plans. Unfortunately the intended benefits of rural development projects set up to support non-farm enterprises are severely hampered by fluctuations in allocations and delays in the release of quarterly instalments. Until recently, the private sector has been shy of investing in rural enterprises, even when encouragement and support were committed. As a result, neither agricultural growth nor employment has increased fast enough to achieve the desired degree of rural development or significantly alleviate rural poverty. It is against this backdrop that there have been significant policy shifts in the recent plans towards creation of selfemployment through skill training, credit supports, marketing facilities, empowerment of the poor, womens participation in non-farm activities, etc.
THE CO-OPERATIVE APPROACH TO RURAL DEVELOPMENT

During the 1970s, KSS were the focus of rural development policy. KSS groups under the Governments Integrated Rural Development Programme (IRDP) were the official channels for distributing government credit, subsidised low-lift pumps and deep tube wells for irrigation and other agricultural inputs. One vital policy goal of IRDP was to develop rural physical infrastructure and create employment opportunities for the rural unemployed labour force. For this, the Rural Works Programme (RWP) combined with Food for Works was launched. The most significant aspect of the integrated rural development policy was to promote subsidised low-lift pumps and DTWs to farmers groups, many of which were later formalised as KSS.The major weakness of rural development policy through IRDP was that it did not pay adequate attention to non-farm or non-agricultural employment creation. Although IRDPs promotion of high-yielding rice did result in the creation of some additional employment opportunities, very few of these were accessible by landless and marginal farmers. The Comilla model of rural development benefited the landed rich more than the small and marginal farmers and these heightened the socioeconomic and spatial inequalities (Jones 1978; Blair 1978; Wood 1979; Hye 1996). Given the dominance of the landed rich in co-operatives, KSS management of the DTW programme turned out to be an imperfect institutional innovation, which not only slowed irrigation expansion but also worsened the prevalent rural inequality (Mandal 1987). The limited success of its programme to create co-operatives amongst the land-owning classes prompted the Bangladesh Rural Development Board (BRDB) to change its policy. Since the early 1980s, BRDB has focused on credit programmes for small, marginal and landless wage labourers.The new policy also provided for resourceless and poor women in income-generating activities, which led BRDB to facilitate the formation of Bittaheen Shamabaya Samity (BSS) and Mahila Shamabaya Samity (MSS).The other conspicuous shift in government policy at that time was to permit groups other than KSS, and even individuals, to purchase tube wells for irrigation. Another change in rural development policy is that it now addresses a felt need to ensure peoples participation
Rural development policy in Bangladesh 73
in government-supported development efforts (Planning Commission 1997).The Kishoreganj model, sponsored by the local government administration is a recent example of this. The model used the existing social structure and local government institutions to mobilise local resources and encourage communities to identify their own needs and problems and identify feasible solutions. The underlying aim of the policy is to make people aware of their needs, abilities, resources and options in order to nurture their self-reliance and ultimately empower them. The model supports this learning process with advisory and technical inputs. While this resembles the bottom-up planning promoted by the fourth Plan, it remains to be seen whether it is any more successful in resolving the conflicts between the objectives of community resource management and the private interests of the local rich and landed elite.
FOOD-ASSISTED DEVELOPMENT PROGRAMMES

Side by side with the mainstream rural development programmes, the Government policy has been to support safety net programmes for the poor.These aim to guarantee food entitlements to the poorest of the rural population, especially distressed women and children, who remain the most vulnerable to natural hazards such as floods and droughts. The programmes include Food for Work (FFW) programme, Rural Maintenance Programme (RMP), Food for Education Programme (FFEP), Vulnerable Group Development (VGD),Test Relief and Gratuitous Relief programmes. The programmes have generally been well targeted to the poor and largely benefited them, especially those who depend on daily wages for their livelihood. The safety net programmes gained importance during and after the 1998 flood, when government spending on this programme increased. In recent years, some food-assisted rural development programmes have integrated skill training, awareness-building and capacity-building for the women beneficiaries. The policy has also addressed some environmental concerns, for example, through roadside afforestation and allowing fish passes in village roads. This policy can therefore be well integrated with the overall development efforts (Planning Commission 2000).
DEVELOPMENT OF RURAL INFRASTRUCTURE

In the early 1980s, the second Plan stressed the development of growth centres to support rapid growth of farm and non-farm sector production and marketing activities.Thus the development of rural infrastructure including roads, bridges and culverts, market places, drainage and irrigation channels was underlined as the major element of the new rural development strategy. This prompted the Local Government Engineering Department (LGED) to initiate numerous projects for developing feeder roads, upazilla connecting roads and nascent market/growth centres throughout the country. The road development project, connecting 1,400 of the 2,100 growth centres/markets, contributed to increasing farm and non-farm output, employment and income, especially of the rural poor and women. Moreover, the functionally landless and small farmers gained a larger share of the increases from crops, wages, livestock and fisheries (World Bank 1996). Employment impact of rural infrastructure development was huge. Other positive impacts of the rural infrastructure development policy included the rapid growth of non-farm sector employment, roadside shops, petty trading, rural transport, etc. In tune with the policy, increased public spending is currently channelled through the Annual Development Programmes for rural infrastructure development programmes.
NGOs FOR RURAL DEVELOPMENT

Government policy in the 1970s responded to the failure of institutional credit to accelerate rural development by facilitating NGO-provision of micro-credit to the rural poor, especially to women. Except for a period in the mid-1990s, when the Government-NGO relationship became more fragile, the Governments supportive role has helped to expand NGO activities in poverty alleviation and rural development substantially. The main policy stratagem of NGOs is the formation of social capital through skill training and education, awarenessbuilding, group mobilisation, micro-financing and empowerment of the poor. NGO support has focused on income-generating activities related to fisheries, livestock, poultry, social forestry, homestead gardening sericulture, bee-keeping, plant nursery development, irrigation development, etc. Some NGOs, such as BRAC and Proshika (and also agro-business companies such as Aftab Bahumukhi Farm and PRAN) have ventured into
production of high-value vegetables, poultry, milk and hybrid seeds on a commercial scale through a contract growing system. Under this system, the companies ensure credit, extension and technological supports to the contracted farmers and buy back products from them at predetermined prices. Other notable NGO programmes include supporting the landless to sell irrigation water, organic farming and food processing.The Grameen Bank gradually extended its micro-credit operations to subsidiary activities such as low-cost housing, sanitation, rural telephone services, etc. Apart from direct income and employment generation, one significant impact of NGO involvement in rural development is that it has induced many government agencies such as BRDB or the Social Welfare Department to take up similar micro-financing policies and programmes for the rural poor. The NGO sector has made a notable contribution to rural development, but it is not without its shortcomings. Lack of transparency, little or no interaction beyond NGO groups and duplication of programmes with fellow NGOs are examples. GO-NGO collaboration is much talked about, but there are few examples in practice.The Palli Karma Shahayak Foundation, which channels government funds to the rural poor through small NGOs, is cited as a success story of GO-NGO collaboration in micro-credit operations (Ahmed 2000).
The most significant impact of the market liberalisation and privatisation policy has been a marked reduction in prices of engines and other irrigation equipment, which means that irrigation equipment is now much more affordable to farmers. The number of local workshops has grown rapidly to manufacture spare parts and provide repair services, as has the number of rural mechanics for irrigation equipment (Mandal 2000). This has led to a rapid expansion of ground water irrigation, especially through STWs. Growth of low-lift pump irrigation was moderate, while there was virtually no growth in DTW irrigation. Private-sector participation in minor irrigation has directly contributed to rural development by accelerating food-grain production. This has meant attaining foodgrain self-sufficiency for the first time.The privatisation of irrigation equipment has resulted in cheaper tube wells and pumps, which has had significant equity implications as more and more small and medium farmers were able to benefit from irrigation. The growth in minor irrigation facilitated the emergence of a fast-growing market for irrigation water, which has largely been competitive and broke the monopoly formerly enjoyed by the landed rich or water lords (Mandal 2000).
FERTILISER DISTRIBUTION POLICY IRRIGATION AS THE LEADING INPUT FOR RURAL DEVELOPMENT
The start and early growth of programmes for minor irrigation relied heavily on the support of BADC, the main public-sector agency involved in irrigation development. Rice production suffered as a result of the very slow growth of irrigation achievable through public-sector programmes. Since the late 1970s, a series of policy reforms have been pursued by the Government to liberalise the market for irrigation equipment and to create opportunities for the private sector to play a larger role in developing minor irrigation. The major policy reforms in irrigation included: (i) liberalisation of imports and distribution of irrigation engines and spare parts; (ii) rationalisation of duties and taxes on the importation of irrigation equipment; (iii) removal of engine standardisation restrictions; (iv) withdrawal of tube well spacing and siting regulations; and (v) withdrawal of subsidies on irrigation equipment. From the early 1960s to the end of the 1970s, fertiliser distribution was under the absolute control of BADC, which marketed subsidised fertilisers. Small and marginal farmers benefited marginally from fertiliser subsidies, but the greatest benefits accrued to middlemen and rich farmers. Partly to reduce the burden of subsidies and partly to ensure fertiliser supplies at reasonable prices, the distribution of fertilisers began to be privatised in 1979. By 1991, fertiliser subsidies were completely withdrawn and distribution was completely managed by the private sector. Government, however, maintains a stock in order to avert a crisis in supply when demand for fertilisers is at its highest.The privatisation of fertiliser marketing has led to an increase in total sales of fertilisers. However, although maintaining a buffer stock of fertiliser seems logical (given the uncertainty in the market), involvement of local officials in monitoring of fertiliser distribution is not only administratively expensive but also hinders the free movement of fertilisers.
AGRICULTURAL MECHANISATION POLICY

Government policy since the early 1980s has been to support the growth of the crop sector by liberalising the import policy and credit supports, which have facilitated the mechanisation of agriculture (Hossain 2000).The private sector has imported thousands of diesel engines, power tillers, four-wheel tractors and sprayers every year. Numerous small hardware companies now manufacture irrigation pumps, pipes and spare parts, sprayer machines, threshers, hand weeders, etc.The policy of withdrawal or reduction of import duties has served to increase imports of power tillers (Planning Commission 2000). Simultaneously with the market for irrigation water, a market for the hire of power tillers has developed. Boro rice farmers have benefited from increased yields as a result of preparing their land using power tillers. Although power tilling affords fewer employment opportunities, total employment increased significantly as a result of the increased cropping intensity and the larger volume of rice produced as a result (Miah 2000; Islam 2000).
mainly due to inadequate marketing facilities and the withdrawal of special credit facilities formerly provided by the Bangladesh Krishi Bank.
DEVELOPMENT OF POND FISHERY

Major policy supports to the sector included provision of credit on easy terms, free training in production, processing, storage and marketing, and production and marketing of fish fingerlings. Private pond fishery has grown to generate substantial productive employment and incomeearning opportunities in the countryside (Miah 2000). Fish farms are linked with hatcheries and with small vendors for fingerlings. Professional fishermen and fish traders are engaged in harvesting and disposing of fish through their marketing network, which is not very efficient in many locations. The provision of fisheries training by the Ministry of Youth Development has further facilitated development of pond fishery, which has largely benefited rural unemployed/under-employed men and women.
POLICY FOR POULTRY AND DAIRY DEVELOPMENT

The Government has given substantial support to private poultry and dairy farming since the 1980s. Special measures included cash support to livestock farmers, distribution of credit on easy terms, tax concessions on importation of poultry farm equipment and poultry feeds, concessional air transportation of improved breeds of poultry chicks and eggs from abroad, vaccination and free training. As a result of this supportive policy, the number of poultry/duck farms and support services such as poultry feed mills, manufacture of vaccines and poultry farm equipment have increased and poultry farming has proved very profitable (Begum 2000). Side by side with the commercial poultry sector, backyard poultry raising by rural householders has also increased significantly. This has positive equity implications as it involves largely the poorer sections of the rural poor and women. A number of agri-business companies offer a contract growing system; day-old chicks, vaccines, credit, extension and technological support is provided to individual poultry farmers and the business buys back products from them at guaranteed prices. The growth of commercial dairy farms was significant up to the mid-1990s, but it showed a declining trend towards the end of the decade,
DEVELOPMENT OF THE RURAL NON-FARM SECTOR

The recent Five-Year Plans have outlined a strategy for non-farm sector development as a part of the rural development policy. Increased public-sector spending on rural road infrastructure, market place and growth centres and rural electrification has contributed to rural non-farm sector development. The non-farm sector includes rural manufacturing, construction and services. Although cottage industry constitutes the bulk of rural manufacturing, its relative importance is declining in terms of the number and volume of products now being produced. New rural-manufactured products include furniture, farm implements, spare parts for machines, hand-knitted products, processed food, etc. According to a recent survey, rural manufacturing employed 2.5 million, or 21 percent, of the rural non-farm labour force in 1996 (World Bank 1996). Employment in rural areas has risen significantly with the growth in rural technologies. As agricultural output grows, road networks intensify and rural electrification expands, the rural non-farm sector is set to be a prime mover of rural development through its backward and forward linkage effects.
CONCLUSIONS AND POLICY OPTIONS

Although Bangladesh has attained self-sufficiency in food-grain production, there is concern about the annual and seasonal fluctuation in food production. Bangladesh has to pursue strategies that promote diversification of its rural economy (Mandal and Palmer-Jones 2000). A number of policy options are highlighted below: Reorganisation and reorientation of mainstream rural development agencies like BRDB needs careful attention. To increase efficiency and competition in project implementation, decentralisation or division of BRDB management in the regions is worth consideration. Projects should be selected primarily on the basis of financial and economic criteria, their agricultural potential and their potential for poverty reduction. Agroecologically-disadvantaged areas and the Chittagong Hill Tract regions should be given more attention. A new form of production organisation should be developed to reflect declining farm sizes and a changing natural resource base. Contract farming is one possibility and seems promising in Bangladesh. NGOs contribute positively to rural development through social mobilisation, micro-credit and skill training to the poor, who may be overlooked by mainstream government programmes. Government should continue to facilitate NGO development initiatives. For sustainability, NGOs should be more accountable, transparent and interactive, as well as more responsive to overall development efforts of the Government. The other requirement for NGOs in business (e.g. seed businesses) is that, as non-profit making organisations, they should compete with the private sector on equal terms and conditions. The present policy of liberalised import and privatesector distribution of small-scale irrigation engines and power tillers/tractors should continue. However, the question of ground water quality, especially in the light of recent arsenic contamination, deserves serious attention and effective and affordable solutions. The untapped potential of augmenting surface water irrigation should be prioritised. A comprehensive policy should be drawn up for augmenting and conserving surface water in small water bodies, khals, drainage channels and roadside pits. Many of these newly-created open-access resources can
be integrated with rice-cum-fish culture, through NGO group mobilisation or even through open franchising by local government.The privatisation of fertiliser marketing should continue, but care should be taken to ensure that the zoning and monitoring of fertiliser delivery by the local administration does not jeopardise free market play. The involvement of officials of the Department of Agriculture should be kept to a minimum in order to allow them to provide their extension advisory inputs properly. The present policy supports to non-crop sub-sectors, e.g. poultry, dairy, pond fishery and aquaculture, should be continued. Significant improvement of agricultural exports is needed to boost the growth of the non-crop sector.The present restrictive policy governing the air-freight, cargo handling, transportation, etc., should be rationalised together with improvement of port management. Development of rural infrastructure mainly roads, rural transports and market places has proved critically important in facilitating agricultural growth and establishing upstream and downstream linkages. The present policy of supporting rural infrastructure should continue. For a meaningful rural development strategy, Bangladesh needs to give special attention to agro-ecologically constrained areas. Finally, the safety net programmes need to be continued. Food assistance should be recognised as an integral part of the resources available for the overall development and its impact needs to be enhanced by combining it with nonfood assistance programmes. However, food aid policy needs to be carefully synchronised with food-grain procurement policy (and with food import) so that the quantum of food-grains distributed through safety net programmes does not affect incentive prices for rice producers.
REFERENCES
AHMED, S. (2000): PKSF The Apex National MicroCredit Fund in Bangladesh (case study). BEGUM, S. (2000):A Comparative Analysis of Broiler and Layer Production in Some Selected Areas of Mymensingh District. Masters thesis, Department of Agricultural Economics, Bangladesh Agricultural University, Mymensingh. BLAIR, H. (1978): Rural development, class structure and bureaucracy in Bangladesh, World Development, 6(1), pp.65-82. ELAHI, K.Q.I. AND MANDAL, M.A.S. (1991): Implementation of land reforms and agricultural labour (minimum wages) ordinances, 1984- a closer look, Bangladesh Journal of Political Economy, 11(2), pp.55-68. HOSSAIN, M. (2000): Growth and structural change in Bangladesh agriculture: implications for strategies and policies for sustainable development, in M.A.S. Mandal (ed.) Changing Rural Economy of Bangladesh. Bangladesh Economic Association. HYE, H.A. (1996): Below the Line Rural Poverty in Bangladesh. University Press Ltd., Dhaka. ISLAM, M.S. (2000): Socioeconomic impacts of power tiller adoption on small farming in Bangladesh, Journal of Agricultural machinery and Mechanization, 4(1), pp.77-85. JONES, S. (1978): Bangladesh: a critical evaluation of recent rural development programmes (mimeo.). University of Cambridge. MANDAL, M.A.S. (1987): Imperfect institutional innovation for irrigation management in Bangladesh, Irrigation and Drainage Systems, 1(3), pp.67-84. MANDAL, M.A.S. AND PALMER-JONES, R.W. (2000): Agricultural growth and rural poverty in south Asia: lessons learnt and issues to be addressed. Review paper prepared for DFID, Bangladesh. MANDAL, M.A.S. (2000): Dynamics of irrigation water market in Bangladesh, in M.A.S. Mandal (ed.) Changing Rural Economy of Bangladesh. Bangladesh Economic Association. MIAH, M.T.H. (2000): Socio-economic Impacts of the Adoption of BFRI Evolved Aquaculture Technologies, A report to the Bangladesh Fisheries Research Institute. MINISTRY OF FISHERY AND LIVESTOCK (undated): Livestock Development Policy. Government of Bangladesh. MINISTRY OF FISHERY AND LIVESTOCK (1998): National Fisheries Policy. Government of Bangladesh. MOA (MINISTRY OF AGRICULTURE) (1999): National Agriculture Policy, 1999. Ministry of Agriculture, Government of Bangladesh. PLANNING COMMISSION (1983): The Second Five-Year Plan, 1980-85. Government of Bangladesh. PLANNING COMMISSION (1995): The Fourth Five-Year Plan, 1990-95. Government of Bangladesh. PLANNING COMMISSION (1997): The Fifth Five-Year Plan, 1997-2002. Government of Bangladesh. PLANNING COMMISSION (2000): Mid-Term review of the Fifth Five-Year Plan, 1997-2002. Government of Bangladesh. RAHMAN, H.Z. (1998): Rethinking land reform, in R. Faruque (ed.) Bangladesh Agriculture in the 21st Century. University Press Ltd., Dhaka. WOOD, G. (1979): Rural Development in Bangladesh Whose Framework? (mimeo.). Institute of Development Studies, University of Sussex. WORLD BANK (1996): Bangladesh: Rural Infrastructure Strategy Study. University Press Ltd., Dhaka.
79
Rural poverty: patterns, processes and policies Rushidan Islam Rahman

Poverty trends and their determinants are a much researched subject in Bangladesh. The objective of this paper is not to synthesise the large number of studies or draw conclusions on their varied findings, but to analyse the relationship between the livelihood strategies of poor rural households in the context of current economic and institutional forces. This will enable us to develop better strategies and interventions for addressing poverty. however, that Table 1 also highlights the fluctuations in poverty incidence over the period shown and therefore we cannot say conclusively that improvements during the early 1990s are likely to be sustained. Other poverty estimates come from the Analysis of Poverty Trends (APT) carried out by BIDS (2000). This shows that the incidence of rural poverty (both extreme and moderate) increased between 1987 and 1989-90 and then declined to 1994. These data also highlighted the significant decline in the non-income indicators of extreme poverty during 1990-95. For example, the rural population without basic clothing declined from 15 to four percent, and households living in extremely vulnerable housing (jhupri) declined from nine to two percent. Households whose members live above the poverty line but are vulnerable to poverty, are also cause for concern. This category is described as tomorrows poor, and is defined as non-poor households whose land ownership is 0.6 ha or less, and who have just sufficient food, day to day. Zillur Rahman (2000a) estimated that 21 percent of rural households fall into this category. These statistics have come out of quantitative studies and surveys and are subject to the usual criticisms of such methods. A more accurate view of the level of rural poverty may be obtained from participatory studies, based on poor peoples own perception of their situation. However, such studies are also subject to a number of problems: comparison over time is difficult as peoples perceptions of poverty change with the growth and availability of consumer goods; q participatory studies on poverty tend to combine the notions of absolute and relative poverty and the symptoms of poverty and causes of poverty.
q
RECENT CHANGES IN RURAL POVERTY

The multi-dimensional nature of poverty is well recognised. Poverty is the result of a range of factors such as income, calorie intake, health and education status and other aspects of material and social well-being. However, multi-dimensional composite indicators of poverty are difficult to compare, year to year. This is where expenditure/calorie-based measures (Head Count Ratios or HCR1) of poverty retain value, due to their simplicity and the ease with which comparisons can be made over time. Table 1 provides estimates of rural poverty incidence provided by BBS (various years) and the World Bank (1998). Table 1 shows a modest reduction in poverty levels and one can safely conclude that there has not been, at least, a deterioration in the rural poverty situation. Note,
Table 1 Trends in rural poverty (HCR) 1983-96 and 2000

Year HCR Poor % point reduction of HCR Extreme poor HCR % point reduction of HCR
1983-84 1985-86 1988-89 1991-92 1995-96 2000
61.9 54.7 47.8 47.6 47.1 42.3
3.6 2.3 0.1 0.1 1.06
36.7 26.3 28.6 28.3 24.6 18.7
5.2 -0.8 0.1 0.9 1.3
As a result, participatory poverty assessments tend to provide higher figures of poverty incidence. Shamunnays (2000) participatory assessment, for example, showed that
75 percent of rural households were poor in 1996, considerably higher than the BIDS surveys estimate of 47 percent in 1996. In fact, the two methods should be seen as complementary, with qualitative studies playing an important role in suggesting indicators, which may then be used in quantitative studies. CHANGES IN HDI AND INCOME POVERTY The Human Development Index (HDI) rose faster during the 1990s, compared to earlier decades. The annual rate of increase was 2.7 percent during 1960-92 and 9.3 percent per year during 1992-98. The reduction of human poverty is, to a large extent, the result of public investment in the social sectors. What is interesting is that the HDI increased at a relatively faster rate than the rate of decline in income poverty. However, the relationship between these two indices must be interpreted with caution. Human poverty can decrease as a result of two sets of factors: public investment in social services or higher incomes, which households can invest in their human capital (either through purchasing adequate and better food or paying for private health and education services, for example).The fact that reducing human poverty does not seem to translate into reduced income poverty is troubling. There are several possible explanations for this. Social-sector expenditure might be flowing to geographical regions and to groups that have already made gains in economic growth; there may be a lack of employment opportunities to mop up the improved human capital; or there is a time lag and in the future we should expect to see gains in human capital increasingly translating into income generation. Further research is needed to understand what forces are really at work.
1998) emphasise that graduation occurs when there has been a sustained increase in income. In general, however, it is agreed that graduation requires not only enough income to move to a less poor status but also the means to resist pressures of downward mobility so households can remain at that improved level. Poor households do not possess enough land or capital to generate sufficient income. A sustained flow of income usually requires diversified economic activity. Moreover, many rural activities are subject to seasonal fluctuation in addition to natural factors (such as floods or hurricanes, etc.), which may lead to income- and asset-loss in some years. A diversified set of activities therefore has more chance of withstanding such vicissitudes and providing a sustained and regular flow of income.Various studies show that many households attribute deteriorating poverty status to risk factors2 and upward mobility to external economic factors. Optimal diversification may be impossible to achieve because of the lack of productive resources. Even if some rural economic activities required little or no capital investment or cultivable land, supplementary resources such as homestead areas, land for storage, etc., are essential. For instance: for raising livestock or poultry, some sort of shelter is required for the animals.While the poorest households often find shelter in a rich relative or neighbours house, they may not be allowed to keep their livestock with them. Some very poor families keep a cow or some goats in their own shacks. But, clearly, very few animals can be accommodated in this way; q a rickshaw puller requires a safe place to keep his rickshaw. The space around the house is usually used for this purpose; q unused public land (khas) alongside roads is often used by poor people for conducting business. But their occupation of khas land usually relies on connections with local power brokers and law enforcement agencies.
q
FACTORS THAT KEEP PEOPLE POOR

Opinions differ on the concept of graduation out of poverty (Sen 2000; Quasem 2000; Husain 1998) and it is useful to start with an operational definition in the context of rural Bangladesh. Some studies (Sen 2000; Quasem 2000) emphasise the sustainability of an increase in consumption and the role played by assets in helping households to overcome unexpected shocks such as illness or loss of employment, which would otherwise push them back below the poverty line. Other studies (e.g. Husain
Moreover, the following supply-side problems also constrain diversification, especially for the extreme poor:
q
the monsoon period disrupts certain types of outdoor
Rural poverty: patterns, processes and policies 81
activities (e.g. petty trading, selling cooked/processed food, etc.); q female labour flow may be interrupted by domestic workload, childbirth, etc.; q sickness and poor health status may lead to the loss of labour days. Capital assets not only provide the means for production and the scope for income diversification, they also act as a form of insurance. Assets may be sold or mortgaged to meet unforeseen demands for large expenditure. Since there is generally a lack of formal institutional insurance against risks faced by rural households, such an asset base is a precondition for graduation. Credit institutions, including the micro-finance institutions (MFIs), do not generally provide emergency short-term credit (though some of the larger MFIs have recently initiated such schemes). The only other option is informal borrowing, which incurs high rates of interest and can trigger off a process of downward mobility. The extreme poor find it virtually impossible to escape poverty because they usually lack even basic resources, such as labour, that are required to make a living. The labour force among the extreme poor may lack education and skill or good health.There may be a predominance of female workers, the elderly or children among the work force.All of these features contribute to low productivity. Paradoxically, most of the extreme poor rural households depend on agricultural wage employment.
However, land distribution patterns do not support this. The extent of sale and purchase of agricultural land has not been significant and changes in the land ownership pattern occur mainly as a result of demographic processes. Rural non-agricultural activities are generally considered as beneficial in reducing inequality, therefore the diversification that characterises the rural economy might be expected to reduce income inequality. The paradox can be explained through an analysis of the access of poor households to the emerging sources of power and the new sources of profit and income. The following processes are evident: remittance flows have been increasing. However, those who receive remittances are not usually the poorest, because they are unable to make the initial investment for overseas migration (Mahmud 1996b;Afsar 2000); q within agriculture, the use of farm machinery is growing rapidly. This leads to income gains by farmers adopting such technology, but may be accompanied by the loss of employment opportunities for poor wage labour households (Rahman, R.I. 2000); q both of the preceding two processes create opportunities for technology-related roles and roles as agents but, again, these are captured by the non-poor; q all of these processes require interaction with the political, administrative and trading systems that are emerging in and around the rural areas and the poor are not capable of establishing or utilising these links. To tap into these systems requires access to social, human and financial capital.
q
INCREASING INCOME INEQUALITY

Income and expenditure inequality increased in Bangladesh during the early 1990s. HES reports show that between 1991-96, the Gini coefficient3 of expenditure in rural areas increased from 0.26 to 0.29. In 2000, it declined to 0.28. Income inequality is even higher with Gini values of 0.36, 0.38 and 0.37 in 1991-92, 1995-96 and 2000 respectively. Although there are no comparable data for recent years, there is no reason to expect a reduction in income inequality as the factors causing inequality (see note 3) have recently intensified. Land being a major productive asset, we might be able to explain rising inequality in rural areas as an expected result of increasing concentration of land ownership.
WHERE THE POOR LIVE

District-level HCR data (BBS 1996) highlight important regional variations in income poverty.These data identify nine districts in the northern part of the country as the worst performers with high HCR values of above 50 percent.These are: Jamalpur, Sherpur, Panchagarh, Rangpur, Sirajganj, Kurigram, Gaibanda, Lalmonirhat and Nilphamari. Dhaka and Chittagong have the lowest HCR of below 30 percent.This can be attributed to the large urban populations of these districts. The worst performing districts are located in areas with low land productivity, most of them subject to regular river erosion. As well as the northern and river-belt districts,
there are concentrations of poverty in some of the lowlying haor4 areas of Sylhet and in Barisal. It should also be pointed out that the identification of divisions or districts with high HCR is not sufficient to identify pockets of severe distress. It is also important to look beyond income and material dimensions of poverty. Security indicators, for instance, are also important. Some villages close to the border and to some big cities experience extreme forms of violence, lawlessness and insecurity. This particularly affects the poorest households.
participation in agricultural growth is therefore not synonymous with poor farmers participation. Farm mechanisation is proceeding rapidly in Bangladesh and is viewed as an important factor in the recent acceleration of agricultural growth (although there is little empirical evidence as yet to prove this). Little is known, however, of the impact of mechanisation on poor households. Both wage labourers and tenants may be displaced by the use of farm machinery, with disastrous effects on the poor. The labour displaced through farm mechanisation may be absorbed in the rural non-farm (RNF) activities. However, the scope may be limited due to the low proportion of hired labour used in the RNF sector and also because of the overall surplus of labour that currently prevails. LIMITS TO POVERTY ALLEVIATION THROUGH AGRICULTURAL GROWTH Before concluding this section, it should be emphasised that crop sector growth alone cannot be relied upon to achieve a significant reduction in poverty in the coming decade.The limits are set by the following constraints: the shortage of land and water resources. Land constraints apply across the whole country whilst the availability of water is a constraint in certain ecological zones; q most poor farmers will base their choices about crop production on considerations of food security rather than on maximisation of household income from the crop sector; q the labour-displacing effect of agricultural growth, which prevents wage rates from rising.
q
AGRICULTURAL GROWTH AND POVERTY

The poverty alleviating impact of agricultural growth is channelled through both macro- and micro-processes. In terms of macro-level processes, there are several forces at work. Firstly, agricultural growth can lead to higher wage rates for farm labourers, who are normally members of the poorest households. Empirical evidence on seasonal wage increases has, sometimes, been used to demonstrate the positive impact of agricultural growth on poverty reduction (Rahman, R.I. 1995; Hossain 1988). However, the impact may not be significant as higher rates usually prevail only for short periods (especially during aman and boro harvesting). Thus the average wage rate over a crop season or year may not be significantly increased. Moreover, farm mechanisation may have a negative impact on labour demand, thus leading to greater competition for labour opportunities and lower wage rates5. Secondly, agricultural growth helps to keep grain prices low. How far cheaper food will benefit households relying on agricultural labouring will depend on trends in real wages, and not merely on the food prices. Small farmers who depend on food purchases to make up food deficits will benefit from the declining real price of food-grains (Abdullah and Shahabuddin 1997). As for micro-processes, the contribution of agricultural growth to poverty alleviation will depend on whether poor farmers with smallholdings are able to participate in the process of agricultural growth. Most empirical studies provide positive evidence on small farmers adoption of technology.This is not, however, a sufficient indicator of poor households benefit, as the small farmer grouping is likely to include those farmers who own agricultural capital in the form of irrigation equipment, etc., as well as poorer households who do not. Small farmers
Agriculture will still play an important role in poverty alleviation through its impact on food prices. If food grain production can keep pace with population growth, real food prices can be kept stable and the poor will benefit.
GROWTH IN THE NON-FARM SECTOR AND ITS EFFECTS ON POVERTY

The crop sector dominates the rural economy of Bangladesh and large numbers of the rural poor are engaged as wage labourers in agriculture. But the prospect
of reducing the poverty of this group as a result of an increase in employment opportunities and/or wage increases in the crop sector is not bright. An assessment is needed therefore of whether non-agricultural growth offers an alternative pathway to poverty alleviation. During the past 15 years, the non-crop sector6 grew strongly. The rate of growth of the non-crop sector was about 4.88 percent per annum, compared to the crop sector growth rate of 1.19 percent per annum (during 1987-96) (Rahman, H.Z. 2000).The small manufacturing sector grew at a rate of 7.7 percent per year from 198995 and at a rate of 5.5 percent during 1994-2000 (Bakht 2000). However, within the RNF sector, the potential for growth varies among the different activities. Bakht (1996) found that small rural industries that had a higher than average number of employees and high capital intensity experienced higher growth. Similarly, the 20 sub-sectors that experienced the highest growth in workforce and profitability within the cottage industry group were nontraditional.The non-traditional cottage industries or high capital intensive small industries are not likely to be owned by poor people. The RNF sector can be seen as a poverty-reducing force if the marginal income gained by the poor is higher in this sector than income gained in agriculture. The econometric analysis of HES data (World Bank 1998; Mahmud 1996) suggests that a shift from agriculture to non-farm occupations entails a significant income gain for households within each land ownership group and who share similar characteristics. According to these studies, if a landless farm worker were to become a transportation worker, his family would enjoy a 16 percent increase in per capita consumption; if he were to take up petty trading, the consumption gain would be about 23 percent. The 1995-96 HES provides strong evidence that people in non-farm occupations are less likely to be poor. This was established by comparing the incidence of poverty among landless agricultural labourer households, other rural households and non-farm wage labour households. The BIDS poverty study obtained similar results (2000).
The questions, then, are why do inequality and income differentials persist and, indeed, increase in the rural economy and why have policies for the accelerating growth in the RNF sector not been adopted? Firstly, there are some structural constraints or barriers to entry that impede poor peoples access to the RNF sector.These barriers are directly related to poverty and include: access to a minimum amount of capital, and the ability to take risks; q some RNF activities require special skills that are obtained through family tradition; q even if some RNF activities do not require specific skills, the organisational and entrepreneurial ability that are needed are not common among the poor. RNF activities often require numeracy, literacy and management skills.They may also require knowledge of the outside world, such as when and where to buy raw materials and capital equipment, how to market output, etc. These are the ingredients of psychological and human capital, in which the poor are usually deficient.
q
Certain regional characteristics and macro-features also constrain the development of RNF activities in poorer regions. For example: a lack of infrastructure, especially all-season roads and easy links with market centres; q a lack of market centres with storage, communication and other facilities; q lack of a power supply; q incidence of natural calamities; q unfavourable law and order conditions; q low growth of crop sector; q a shortage of internal demand for RNF products due to poverty.
q
There is a circularity in the above constraints.Whilst nonagricultural growth offers the promise of poverty alleviation, poverty tends to be reinforced through constraints acting at both an individual level (via limitations in human capital) and also at the regional level, through regional characteristics and demand shortages. To build on the potential of non-agricultural growth as a strategy for poverty alleviation, the above constraints need to be addressed. Policies may focus on the removal of area constraints and on the development of individuals
capabilities.While the first route is a necessary precondition, without the development of entrepreneurial ability and investment capacity of individuals,little will be achieved.
MICROFINANCE AND POVERTY

The initial experience of micro-finance (MF) in Bangladesh looked very promising for poverty alleviation. Loans were used mostly in productive activities, generated high rates of return and household incomes significantly increased. Proponents of MF claimed that the provision of investment funds helped break down the vicious circle of poverty (low income low savings low investment low productivity low employment low income). MF coverage has expanded dramatically but many now question the scale of its impact on poverty alleviation. During the early years, MFIs were largely successful in targetting the poor. Hossain (1988) estimated that more than 90 percent of the clients were drawn from poor households. A number of recent studies, however, have shown that success in targetting poorer households has declined during the 1990s (Matin 1998; Husain 1998). A special feature of MF has been its targetting of femaleheaded households, which are often among the poorest of the poor. However, the coverage of female-headed households attained remains proportional to their overall share in total households (Husain 1998; IMEC 1999). A number of changes occurred as MFIs scaled up their coverage: the average size of loans increased and clear boundaries on the use of loans gave way to encouraging independent decisions by borrowers.As a result, visions of the purpose of MF loans and their actual use diverged: loans were not always used for productive purposes and this in turn often led to indebtedness. Increasing competition and efforts to achieve financial sustainability often led to MFIs adopting policies not conducive to poverty alleviation (Khandker and Rahman 2000). For example, some MFIs raised the average size of loans and effective interest rates increased. Most MFIs adopted a rule of compulsory weekly savings and required a fixed amount of money to be deposited. These policies discouraged the poorest households from becoming members. Poorest borrowers may prefer smaller loan sizes and may not have other sources of income from which they can repay the weekly instalments (Rutherford 1995). Larger average loans often attracted non-poor clients who may in turn crowd out the poor households.
MIGRATION AND POVERTY

Migration is an increasingly important phenomenon in Bangladesh. In terms of its impact on poverty alleviation, an important question is which group of the poor has the greater propensity to migrate: the extreme poor, the moderate poor or tomorrows poor? Data suggest that the moderate poor comprise the largest component of migrants, while the percentage of extreme poor who migrate is low (Afsar 2000; Mahmud 1996). Migrants have a stronger skill and resource base when compared to non-migrants. As a result, they improve their social and economic condition more rapidly than do non-migrants (Mahmud 1996). Migration enables these households to generate a higher return because their enhanced capabilities are combined with better employment opportunities. Such employment opportunities originate mostly in the urban informal sector. During the last decade, many migrants, especially female migrants, have found employment in the RMG sector. Focusing specifically on international migration, it should be clarified that the poverty alleviation impact of remittances is not currently as large as the potentials and/or expectations of such impact. This is due to a number of reasons: remittances are not being invested in productive rural sectors or labour-intensive industrialisation that could generate large-scale employment for the rural poor; q remittances create demand for imported luxury goods rather than locally produced goods; q remittances may appear as windfall gains and raise expectations about making such gains. Those who receive remittances are unaware of the effort required by migrants; q the return of migrants exposes the rural society to various social and health problems, including HIV infection and AIDS; q the process of migration takes place through middlemen which exposes the poor households to the risks of exploitation and reduces returns on investment in migration.
q
NATURAL RESOURCES AND POVERTY

During the last three decades, awareness of environmental degradation has led to concerns that the poor in Bangladesh are both victims and agents of environmental damage (Rashid et al. 1994). An over-emphasis on the link between poverty and environmental degradation has been quite common: poverty and environmental degradation form a trap from which there is little chance to escape it is on the environment in which the poor live and from which they draw their sustenance (Adams 1990). Such arguments appear to miss the fact that poor and rich both derive their livelihood from natural resources, especially land.Whether the poor cause more degradation of natural resources than the non-poor remains to be proved. Any discussion of the relationship between natural resources and poverty has to include a consideration of access by the poor to common property resources (CPRs). Zillur Rahman (1996a) showed that consumption items from CPRs accounted for a significant percentage of their income but that the scope for such expenditure savings through collection of firewood, jute sticks, fruits, vegetables and fish decreased substantially between 1991 and 1995. Gradual encroachment of wetlands by private owners, and the commercial leasing of water-bodies, are reducing poor peoples access to CPRs and have a significant adverse effect on livelihoods and, specifically, on the nutritional status of the poor (Rahman, H.Z. 1996b; Beck and Ghosh 2000).There are important gender dimensions to this process. Adnan and Rahman (1976) showed that women and girls not only collected fuel for household use but also sold a part of it, generating cash income for the family. Gleaning paddy from fields after harvest forms an important source of consumption for many destitute women. It is clear that systematic efforts are required to increase poor households access to CPR. Also, local government institutions might be able to take the lead in developing integrated resource plans. However, it also should not be forgotten that greater access to natural resources is by no means a sufficient condition for graduation from poverty. Production relations based on natural resources are often exploitative and perpetuate poverty among communities. Some of the poorest groups in rural Bangladesh are tribal and fishing groups who eke out a living from natural resources.
HOW TO CHANGE THE POVERTY SITUATION

REMOVAL OF ACUTE DISTRESS IN DESTITUTION-PRONE AREAS Poverty alleviation policies must begin by attaching highest priority to the removal of the worst forms of poverty: hunger, destitution and crisis. Relief, rehabilitation and targetted food distribution programmes are some of the forms of direct intervention. However, more effort is needed to properly identify the destitution-prone pockets, affected areas and the crisis period of the year. AGRICULTURAL GROWTH Agricultural growth can help to reduce poverty if it is labour generating and if an appropriate environment prevails to ensure that marginal and small farmers can participate in such growth. As mechanisation is proceeding rapidly, its impact on labour displacement and wage rates needs urgent examination. The agricultural growth potential of the poorest areas needs to be assessed, and to realise its potentials the provision of subsidised inputs, and the implementation of small infrastructure projects and marketing support projects to these areas, may be considered. Targetted credit for farmers growing specific crops, and for poor and marginal farmers, may help them to use inputs more efficiently. Such programmes should be based on an understanding of patterns of input use among the poorest farmers in contrast to prevailing studies, which compare large with small farmers. NON-FARM GROWTH The daily earning of an unskilled worker engaged either as an agricultural wage labourer or as an RNF labourer (or even in a formal sector job such as the ready-made garment sector) is insufficient for supporting a threemember family above the poverty line. In such a situation, crossing the line will be possible only if there is a shift in the productivity and wage rate of non-farm activities. This will be possible only through an increase in the scale of activity and the introduction of modern technology in the case of RNF self-employment.The provision of adequate infrastructure is also a prerequisite. A combination
of these facilities may be created in the appropriately chosen hubs of non-farm activities that may be developed into peri-urban-centres of rural industries. REDUCING INCOME POVERTY BY REDUCING HUMAN POVERTY A reduction of human poverty is, to some extent, an end in itself. But conscious efforts need to be made to ensure that advances in human development lead to a reduction of income poverty. This requires actions on two fronts: firstly an improvement in the quality of education and provision of new skills and secondly to initiate a growth process that will utilise the skilled labour force.There are limitations on the extent to which skilled labour should and could move to urban areas, so growth centres need to be established in rural locations. REDUCING INEQUALITY A reduction in inequality can have a noticeable impact on poverty trends and processes but efforts to redistribute wealth are fraught with political difficulty. Efforts to reduce inequality should therefore concentrate on giving the poorest households access to new resources and assets that can help them to take advantage of new opportunities. These include specialised skill training to generate marketable skills, preferential access to credit, support in the form of specific inputs for production activities and in the form of services for accessing product markets and mechanisms for creating pro-poor bias in institutional (both Government and NGO) service delivery. ACCESS TO INSTITUTIONS A significant upward shift in income among the poor requires new initiatives, new activities and new modalities of investment. Such changes require a restructuring of the institutional relationships and linkages. In the past, many institutions both formal and informal not only bypassed the poor but sometimes acted against them. It is time to change the institutional processes and provide services in such a way that the poor can participate and are provided with security against theft, violence and illegal toll collection. FINANCIAL SERVICES It is clear that a disaggregated approach is needed. Rather than allowing the random entry of the non-target group, MFIs should clearly target the tomorrows poor as a
distinct group and provide a slightly modified set of financial services for this group.This may in turn help MFIs to expand the scale of activity at a lower cost and costsavings may be used for operating credit services for the poorest households. While the credit needs of rural poor have been recognised, their need for other financial services has not been taken into account. Most poor households with cash income would save and generate their own investment capital if savings services were available in close proximity. COUNTERACTING THE FORCES OF DOWNWARD MOBILITY Special attention is needed to address the risk factors that erode the income and asset base of poor households. Health problems, for instance, increase the vulnerability of all groups of poor. More effective health services may reduce the expense and workdays lost due to the sickness episodes (Begum and Sen 2000). Among man-made problems, dowry requires immediate attention with a planned agenda to be implemented through social actions and awareness raising. SAFETY NET PROGRAMMES Recent safety net measures undertaken by the government (in the form of housing facilities, allowance for old and widowed, etc.) deserve appreciation first and foremost for their welfare impact.Their impact on the graduation process can only be assessed after a longer interval. Effective policies must be adopted for alleviating the extreme poverty of rural households that are constrained by lack of initiative, entrepreneurship and human capital and therefore cannot be expected to graduate from poverty through self-employment. Safety net programmes consisting of direct transfer or targetted employment programmes are often the only hope for this group. Special employment schemes based on food aid have proved successful and should continue, but there has been a recent decline in donor resources for this purpose.
NOTES
1 Headcount Ratio (or Headcount Index) refers to the percentage of individuals or households that fall under the poverty line.The poverty line is often set at the expenditure required to buy a bundle of food items that will provide a given amount of calories, plus expenditure for purchasing basic non-food items. 2 A Gini coefficient is a measure of inequality (of income, expenditure, etc.). The higher the value, the higher the inequality. 3 Risk factors may include natural calamity or man-made disasters and insecurities that create an unforeseen demand for expenditure in cash, kind and time.They also include life cycle factors including natural processes of change in family size, composition, age structure and so on. Life cycle factors affect both the scope for income generation and the consumption needs of a family. 4 A haor is a depression in the floodplain. 5 Further research is needed to identify the exact nature of what is going on in the wage labour sector. 6 For the purposes of this paper, all non-crop activities (livestock, fishery, etc.) and rural services, industry and processing activities are included in the category of RNF. Livestock and fishery are components of agriculture in a technical definition.
REFERENCES
ABDULLAH, A. AND SHAHABUDDIN, Q. (1997): Recent Developments in Bangladesh Agriculture. Crop Sector, in Growth or Stagnation?, in R. Sobhan, CPD and UPL (eds.) A Review of Bangladeshs Development 1996. Dhaka. ADAMS,W.M. (1990): Green Development, Environment and Sustainability in the Third World. London, Routledge. ADNAN, S. AND RAHMAN, R.I. (1976): Social Change and Rural Women: Possibilities and Participation, in M. Hossain et al. (eds.) Role of Women in Socio-Economic Development of Bangladesh, 1976. Also Reprinted by Land Tenure Centre, University of Wisconsin; Reprint No. 137,April, 1979. AFSAR, R. (2000): Rural-Urban Migration in Bangladesh. Causes, Consequences and Challenges. University Press Ltd., Dhaka. BAKHT, Z (2000): Growth Performance of the Manufacturing Sector. Review of the Revised Industrial GDP, in A.Abdullah (ed.) Bangladesh Economy 2000: Selected Issues. BIDS. BAKHT, Z. (1996):The Rural Non-farm Sector in Bangladesh: Evolving Pattern and Growth Potential. The Bangladesh Development Studies, Vol. 24, September-December, Nos. 3 and 4. BBS (1996): Report on Labour Force Survey in Bangladesh 1995-96. Government of Bangladesh, Dhaka. BECK, T. AND GHOSH, M.G. (2000): Common Property Resources and the Poor. Findings from West Bengal, Economic and Political Weekly, 15 January, 2000. BEGUM, S. AND SEN, B. (2000): Not Quite, Not Enough: Financial Allocation and the Distribution of Resources, Research Report 167. BIDS. BIDS (2000): The Face of Human Deprivations. Bangladesh Human Development Report 1999 (mimeo.). BIDS, Dhaka. HOSSAIN, M. (1988): Nature and Impact of the Green Revolution in Bangladesh, Research Report 67. IFPRI, Washington D.C. HUSAIN, A.M.M. (1998): Poverty Alleviation and Empowerment. The Second IAS of BRACs Rural Development Programmes. BRAC, Dhaka. IMEC (1999): Proshikas Impact Assessment Study 1998-99. Proshika, Dhaka. KHANDKER, S.R. AND RAHMAN, R.I. (2000): Editors Introduction, The Bangladesh Development Studies, Special Issue on Microfinance and Development, July-September,Vol. 26.
MAHMUD, W. (1996): Employment Patterns and Income Formation in Rural Bangladesh:The Role of Rural Nonfarm Sector, The Bangladesh Development Studies,Vol. 24, September-December, Nos. 3 and 4. MAHMUD, R.A. (1996): Migration and Poverty, in H.Z. Rahman, M. Hossain and B. Sen (eds.) 1987-1994: Dynamics of Rural Poverty in Bangladesh (mimeo.). BIDS. QUASEM, M.A. (2000): Graduation of NGO Participants: Some Preliminary Findings (mimeo.). BIDS. RAHMAN, H.Z. (1996a): Ecological Reserves and Expenditure Savings, in H.Z. Rahman; M. Hossain and B. Sen (eds.) 1987-1994: Dynamics of Rural Poverty in Bangladesh (mimeo.). BIDS. RAHMAN, H.Z. (1996b): Crisis, Income Erosion and Coping, in H.Z. Rahman; M. Hossain and B. Sen (eds.) 1987-1994: Dynamics of Rural Poverty in Bangladesh (mimeo.). BIDS. RAHMAN, H.Z. (2000): Poverty in Bangladesh:An Overview of Trends and Issues (mimeo.). BIDS. RAHMAN, R.I. (2000): Tractor Use, Irrigation and Agricultural Productivity in Bangladesh, in A.A.Abdullah (ed.) Bangladesh Economy 2000: Selected Issues. BIDS, Dhaka. RAHMAN, R.I. (1995): Impact of Irrigated Agriculture on Rural Employment, Wage and Household Income in Bangladesh, Paper presented at the Workshop on Agrarian Structure and Agricultural Productivity in West Bengal and Bangladesh, organised by Queen Elizabeth House, University of Oxford and Centre for Social Studies, Calcutta, 9-12 January, 1995. RASHID, M.A., BUIYAN, A.R. AND AHMED, S. (1994): Environmental and ecological Aspects of Poverty and Implications for Sustainable Development in Bangladesh. MAP Focus Study Series No. 3. CIRDAP, Dhaka. RUTHERFORD, S. (1995): ASA The Autobiography of an NGO.ASA, Dhaka. SEN, B. (2000): Impact of MF on Poverty. Preliminary notes presented at an internal seminar at BIDS. SHAMUNNAY (2000): The Budget and the Poor. University Press Ltd., Dhaka. WORLD BANK (1998): Bangladesh. From Counting the Poor to Making the Poor Count.
89
Migration and rural livelihoods Rita Afsar

Migration is both a cause and effect of socioeconomic change. Internal migration is one of the dominant forces contributing to the rapid urbanisation of Bangladesh and is likely to increase in scale, complexity and diversity. Along with reclassification1, migration contributed between three-fifths and two-thirds to urban growth since independence. External migration is important too. Bangladesh is a country with a huge labour surplus, and the growth of the labour force far out-strips the growth rate of the population. Estimates suggest that the magnitude of unemployment and under-employment is at around one-third.That this is fairly stable is in part due to the continuous outflow of migrants and inflow of remittances that generate employment opportunities, directly or indirectly (World Bank 1996). Despite its importance, migration is poorly understood and often viewed as a single event of permanent nature, with negative consequences for the distribution of population and resources (Afsar 1999a). However, this is only one scenario among many.This paper attempts to clarify some common misunderstandings outlining the importance of migration to rural livelihoods and examining its causes and consequences. of commitment that migrants have to their place of origin or destination. Commitment may be gauged by factors such as the migrants asset base and expenditure pattern on housing, intensity of social contacts, the proportion of the migrants urban income that is saved or sent as remittances (Afsar 1995). The possibility of permanent relocation does not arise with regard to international labour migration, due to restrictions in receiving countries. Non-permanent international migration is becoming more important, due to the massive improvement in road and transportation networks, information technology, the restructuring of the global economy and the growing tendency towards greater selection of migrant workers. Estimates suggest that nearly three million Bangladeshi nationals were employed abroad between 1976 and 1999 and all were temporary migrants (Afsar et al. 2000). Side by side, there are a growing number of undocumented workers who are not registered with the Bureau of Manpower Employment and Training. The diversity of migration types calls for improved sources of data and also precludes any expectation of a simple and uniform relationship between migration and development (Afsar 1999a).
DEFINITION, PATTERNS AND ATTRIBUTES

Migration describes a large spectrum of movement from commuting or temporary absence from the home location for a couple of days to several years to seasonal migration or permanent relocation (Afsar 1999a). Problems associated with the categorisation of migration are described in Afsar (1995); here, it is sufficient to say that information on place of birth is needed to distinguish migrant from non-migrant and that a 10-year cut-off is generally used to distinguish the long-term from recent migrants2.Two variables can be used to classify permanent and temporary migrants; the first is whether the migrant is living with the family of procreation and, the second, whether housing is owned, rented, boarding house or mess. Whether temporary internal migration is likely to lead to permanent relocation depends mainly on the level
PROFILE OF MIGRANTS
AGE SELECTIVITY Migrants are generally young adults and temporary migrants tend to be younger than permanent migrants. Household surveys at destination revealed that threequarters of temporary migrants,half of permanent migrants and only a third of non-migrants are between 15-34 years of age. Migrant labour in the RMG sector is even younger, with more than 90 percent of them under 30 (Afsar 2000b). Paradoxically, among overseas migrants, most male workers are over age 30;25 percent of this group are over 40 (Afsar et al. 2000). However, youths (under 30) comprised nearly one-third of male overseas migrants and
more than 70 percent of female overseas migrants.Whilst the age of migrants may be related to the degree of risk they are willing to take, age selectivity in female migrants is put down to demand factors such as dexterity. GENDER SELECTIVITY The sex ratio is an indicator of gender selectivity in the population of a particular area and of that areas sending and receiving potential. Areas with more men than women (a high male:female ratio) are conventionally considered as receiving areas and are said to have a high sex ratio. Those areas with more women than men are considered sending areas. An examination of sex ratios for different cities over time shows that cities that experienced higher urban growth had relatively higher sex ratios than cities that experienced slower growth. As a result of male migration, the sex ratio becomes more feminine in the active age group in rural areas adjacent to sample towns (Afsar and Baker 1999). The high sex ratio in urban areas can be attributed to social and economic factors. Prior to the 1980s, the majority of poor women in rural areas worked as unpaid family helpers and housewives, whereas their husbands migrated to cities to work in the transport, service, manufacturing and construction sectors. Since the 1980s, however, the sex ratio has declined in all metropolitan cities as a result of the increasing incidence of familybased migration and an increase in womens independent migration.The demand for cheap labour generated by the RMG industries paved the way for womens migration (Afsar 2000c). Evidence suggests that the RMG industry absorbed 1.5 million workers in the past decade, more than 90 percent of them migrants from rural areas (Afsar 2000b).Young women, largely from landless families, are migrating to the capital to earn a livelihood that was hitherto considered a male prerogative. SOCIOECONOMIC CHARACTERISTICS Chaudhury and Curlin (1975) and Afsar (1999a; 1996; 2000a) showed clearly that migrants to slum and squatter settlements were predominantly from landless families and nearly half of them were agricultural labourers. In contrast, migrants to non-slum areas were largely students and unemployed (nearly three-fifths) and nearly one-fifth of them had jobs in the public sector. Owner-cultivators made up the smallest group.
On average, household heads in slums had two years of education and were illiterate at the time of migration, compared to counterparts in non-slum areas of the same age who, on average, had eight years of education. The reasons for migrating were also different. Whilst 40 percent of non-slum migrants left home to seek higher education or to take up a job transfer, nearly 66 percent of respondents from slums said they migrated to Dhaka for employment-related reasons (Afsar 1996). People also migrate from rural areas to marry or seek education (Afsar and Baker 1999). Certain spatial trends have emerged: the poor are leaving the central and southern districts of Comilla, Faridpur, Barisal and Dhaka: more than 90 percent of poor migrants belong to these districts.Although these are not the most impoverished districts in terms of their agricultural growth performance, most are highly flood-prone. Perhaps more importantly for this group of people is that these districts are well connected to Dhaka City by waterways and road networks and therefore have had a long history of migration. Almost 90 percent of poor migrants who settle in slums and squatter settlements came directly to Dhaka from their areas of origin. In contrast, one-third of their nonpoor counterparts came via small towns and, in some cases, from other cities (Afsar 1999b). MIGRATION AS A HOUSEHOLD LIVELIHOOD STRATEGY Rural-urban migration in Bangladesh is a householdbased strategy for maximising income or well-being and is partially dependent on the support of other family members, and/or district-based acquaintances, friends and neighbours.Whether men or women, nine out of ten migrant labourers to RMG factories arrived as part of a group of family members or friends (Afsar 2000b). Similarly, temporary migration is perceived as a way of maximising the familys income and minimising its risks. Migrants and families both benefit from the migration process; the migrant gains membership of a support network and the familys resources enlarge through the economic efforts of the migrant (Afsar et al. 2000). Of overseas migrants, 86 percent used family members to look after their wives and children (Afsar et al. 2000). The proportion is similar for temporary migrants to Dhaka. Indeed,
Migration and rural livelihoods 91
strong support among family members is now more prevalent for migrant families than in general for rural Bangladesh (Afsar 2000). The benefits of a strong support network among migrants should not be underestimated; it offers relocation support, opportunities for education and to learn new skills, courage to bargain for higher wages and protection from threats such as physical assault, sexual harassment, dismissal from jobs, etc. Far from the self-motivation and anonymity sometimes associated with the migrant way of life, poor migrants generally value their families over selfinterest and live in a close-knit community of family and friends.
Nonetheless push migration does occur. It is often due to a lack of year-round employment in rural areas and/or a result of natural calamities such as floods, cyclones, etc. Surveys reveal that almost two-fifths of rural households sent their adult members to the nearest towns due to lack of year-round employment.The data show that respondents who left their homes immediately after the great floods tended to be from the poorer sections of the population: they viewed migration as a temporary measure (Afsar 1999b). Only people who lost their homesteads and/or assets such as livestock were forced to become permanent migrants. MYTHS AND REALITIES OF PULL MIGRATION Neo-Malthusian demographers and labour economists are generally pessimistic about urbanisation and view immigration to cities with alarm.They have been strongly supported by Todaros model of labour migration and urban unemployment (1969).Todaros model fitted well with three prevalent stereotypes: high wages in the modern sector with presumptions of mass unemployment; permissive or overly generous policies and/or articulate militant labour movements (Kannappan 1985: 703). But is it relevant in the Bangladeshi context? It is important to answer some basic questions. Firstly, Does the dichotomy between modern and traditional sectors really exist and are urban immigrants unemployed for long periods? In short, no. Firstly, the view of urban economies as comprising discrete capitalist and traditional sectors is obsolete; modern livelihood strategies straddle the rural-urban divide. For example, rural producers need bigger urban markets to get better prices for their products.The majority (53 percent) go to the nearest district town as they get better prices there than in their own villages (Afsar and Baker 1999). Similarly, urban businesses need a cheap and loyal labour force, which rural migrants provide. The growth of the RMG industry is a case in point. Secondly, with regard to unemployment, data from secondary sources and household surveys show that the rate of unemployment has always been low, particularly for poorer households, which cannot afford to remain unemployed. Unemployment among recent immigrants is lower than prior to migration and lower than that among long-term immigrants or non-migrant counterparts (Afsar 1999b).
CAN PUSH AND PULL FACTORS EXPLAIN RURAL-URBAN MIGRATION?

MYTHS AND REALITIES OF PUSH MIGRATION It is generally argued that the high rate of population growth and the consequent increasing pressure on limited cultivable land is creating landlessness and pushing labourers to the cities. However, this argument seems flawed. Firstly, the agricultural productivity of districts that record consistent flows of emigrants to Dhaka City, namely Dhaka, Faridpur, Comilla and Barisal, are generally higher than the national average.The hypothesis might hold were these districts ones in which land is concentrated in the hands of large land owners. In fact, there are few instances of land holdings of more than 2 ha in any of these districts, nor is landlessness higher in these districts. These districts are highly productive Green Revolution districts. However, there were winners and losers in the Green Revolution. Hossain (1988) notes that the top 20 percent of per capita income benefited the most, the middle-income band were squeezed and the bottom 40 percent remained unaffected. This, and the fact that the growth of agricultural income was only around two percent as opposed to more than seven percent per annum for non-agricultural income (1988-95), meant that landless agricultural workers had little incentive to remain in agriculture (Rahman et al. 1996). Instead, they chose to commute to nearby district towns or migrated to larger towns and cities in search of more remunerative nonagricultural jobs.
Table 1 Primary occupation of the migrant head of the household (slum) prior to migration to Dhaka, immediately after coming to Dhaka and at the time of survey
Primary occupation Before migration Immediately after coming to Dhaka Current occupation
Agriculture* 44.4 Industry 0.6 8.5 9.2 Trade 9.2 16.6 28.4 Salaried Service 3.1 25.7 18.5 Construction 3.1 25.7 19.7 Transport 4.7 18.4 20.4 Rent/Transfer/Income/Pension 1.2 Housewife 8.0 3.0 2.5 Unemployed 7.4 Not Working** 21.5 1.8 All Occupation (No.) 162 162 162 *Agriculture includes farming, agricultural labourers who, in fact, constituted the dominant majority and livestock/fisheries **This category includes children and students Source: Afsar (1999b)
Similarly, new immigrants find jobs more quickly, particularly the poor and unskilled immigrants. Almost half of these groups who settled in slums and squatter settlements found work within seven days and three-quarters of them found employment within a month. This shows that Todaros perspective of the urban labour market is invalid. Are earnings in the urban informal sector less than those in the modern sector and in rural areas? Most poor rural migrants settle in urban slums and squatter settlements. Slum dwellers generally find jobs as manual labourers: they account for 60 percent of the urban labour force in the construction, transport and manufacturing sectors (Hossain et al. 2000). Earnings from those sectors contributed 18 percent of the income of slum households in 1991. Over time, poor immigrants diversify their livelihoods and enjoy better incomes than had they stayed in rural areas (Table 1). Jobs in these sectors contributed less than 10 percent of household income for households in small towns and rural areas (Afsar and Baker 1999). Hossain et al. (2000) found that poor migrants to Dhaka City have been able to improve their incomes at a much
higher rate than those remaining in rural areas. Ample economic incentives exist therefore for the rural poor to migrate to Dhaka City. Public policies and NGO efforts to alleviate poverty and promote rural development have failed to offset market forces that overwhelmingly favoured the concentration of social and economic activities in large metropolises, particularly in Dhaka.Therefore, it is not the pull of social services that brings people to Dhaka and which makes planners apprehensive of immigration. Rather, it is the inability of planners to equip other big cities, medium and small towns with sufficient infrastructure and facilities to attract investment and markets that has led to the disparity between Dhaka and the rest of the country in their ability to provide income and employment opportunities. POLICY PULL Some commentators blamed the statutory rationing system for accelerating migration from rural areas to big cities. As a result, the system was withdrawn (except for selected groups such as the army, police, etc.). The Government began to focus on policies and programmes which were intended to alleviate poverty, develop rural areas and curb the growth of large cities (Afsar 1999b).
VGD and FFW Schemes were launched to cater for the rural destitute and vulnerable groups.The Reorganisation and Decentralisation Policy of 1982-83, and the subsequent development of thana-upazilla complex and transportation networks linking thana to district headquarters and Dhaka, helped to revive decaying middle-sized towns with populations of 25,000-100,000.The number of such cities increased from 20 in 1961 to 102 in 1991 (Afsar 2000c). A study of the medium-sized towns of Faridpur and Rajbari showed a rapid increase of population from 66,579 to 74,000 and 38,645 to 44,000 respectively between 1981 and 1991. Immigrants constituted 60 percent and 50 percent of population of those towns respectively (Afsar and Baker 1999). Following decentralisation, nine thana headquarters became municipalities and 392 thana headquarters were reclassified as urban in 1982-83. More importantly, it resulted in investment at local levels, transfer of civil and judicial administration and, subsequently, employment opportunities in the construction, services and trading sectors increased (Afsar 2000). In contrast, in 1991, nearly a quarter of urban centres were upgraded from thana headquarters to municipalities, and 46 urban growth centres were created. However, the move was neither followed by adequate investment on infrastructure development, nor were there any transfers of power.As a result, it failed to attract the upper- and middle-classes with the ability to invest and create employment opportunities. NGOs began to work in urban areas only in the 1990s. They focused on micro-credit for income generation and human capital development through education and health services.Their policy is to serve the poor and disadvantaged through a service-giving approach. They do not address broader areas of decentralisation and infrastructure development that really affect the process and pace of urbanisation. Moreover, their target-oriented welfare approach is not likely to create any pull effect, given their recent arrival and low coverage in urban areas. Thus, NGO policies have probably had little impact on rural-urban migration or urbanisation. Bangladesh embarked on a structural adjustment policy in the 1980s, resulting in the expansion of private investment and rapid expansion of private informal and formal sectors, whilst the public and autonomous sectors shrank
(Afsar 1999a). In contrast to Dhaka City, private-sector development has been much slower in small and intermediate towns, however, and the pull of Dhaka City as a source of employment remains strong.
CONSEQUENCES OF MIGRATION
MIGRATION, POVERTY AND GROWTH Migration contributes to the alleviation of poverty. Existing estimates on poverty show that Dhaka and Chittagong districts have fewer people living below the poverty line. Using the HPI to identify relative deprivation at district-level3, we find that, with the exception of Faridpur, all other sending districts have HPI values of 35 percent or less (an HPI value of over 45 percent represents worst performance) (Sen and Ali 1999)4. Rahman et al. (1996) noted that three-quarters of rural migration occurred from landless households. Owner-cultivators now predominate in such areas as they are able to afford the costs of agricultural innovation necessary for higher growth. Afsar et al. (2000) estimated HPI values for sample households in these districts at 21 percent prior to overseas migration. Post-migration, the proportion of such households dropped dramatically to seven percent. USE OF REMITTANCES IN RURAL AREAS Empirical studies reveal that consumption expenditure of rural households constituted 37-90 percent of overseas and urban remittances (Afsar 2000a; Afsar et al. 2000). A significant proportion of remittances are spent on house repairs, considered to be the basic unit of both production and reproduction for poorer families. Buying land and agricultural inputs such as irrigation and fertiliser to cultivate HYV rice are common ways of using remittances in rural areas. Remittances also help to strengthen human capital development through childrens education and treatment of sick members. On average, temporary migrants remit respectively 40 percent and 45 percent of their urban and overseas income to their families in rural areas. SCOPE FOR THE DEVELOPMENT OF HUMAN CAPITAL IN BIG CITIES Schools are generally located in better-off areas and the quality of education acquired by rich and poor students is markedly different. More importantly, there is a
conspicuous difference in the enrolment rates between age-cohorts in slum and non-slum areas (this is significantly larger at the tertiary level where very few counterparts from slum and squatter settlements enrol). Nor did enrolment seem to improve. Figures show that higher education does result in higher income and seems to be a significant factor in determining urban income (Hossain et al. 1999). They further suggest that, unless the household can afford to invest in the higher education of children and/or can provide financial support to the educated members to undertake business activities, the income disparity between the rich and the poor is likely to continue. Women and children of the poorer migrant households are more likely to work when compared to counterparts in better-off migrant and temporary migrant households. This puts them at greater risk of remaining uneducated and in poorly paid jobs. Slum dwellers suffer as a result of lack of access to water and sanitation facilities; to make matters worse, slums often spring up in vulnerable areas such as along drains, around garbage dumps or adjacent to ditches or floodprone areas.The overall morbidity rate is much higher in slum (56 percent) than in non-slum areas (35 percent). VULNERABILITY AND EMPOWERMENT Temporary migrants face the day-to-day insecurity of work and financial security. Permanent migrants in slum and squatter settlements have suffered largely from sickness, followed by the death of the main breadwinner. Both groups often resort to borrowing from friends/neighbours, co-workers and even informal money-lenders to avert crises. Existing literature provides contradictory evidence on the impact of migration on womens empowerment, depending mainly on their migration status and the nature of migration. In generating employment opportunities for poor rural women, the private sector is playing a positive role in poverty alleviation. Semi-literate and unskilled women from rural poor families are entering the formal manufacturing sector, which was previously open only to a few educated and upper- or middle-class women (Afsar 2000b). It is a way of enabling young and unmarried women from landless families to defer marriage. But on the other hand it may burden women with a double workload of home and factory work.
CONCLUSIONS AND POLICY RECOMMENDATIONS

Evidence shows clearly that migration offers better livelihoods, greater income and improved life-chances but the potential for reaping these rewards varies greatly according to the migrants support networks, endowment of capital and assets in urban areas and human capital content of the worker. There is a profound disparity in the income distribution and human capital development between the rich and the poor in Dhaka City. If recent trends continue, income distribution will worsen further unless positive action is taken. Attempts should be made to generate demand for higher education (up to secondary and high school level) among the poorer population.This can be done by providing educational credit and preferential loans for entrepreneurs along with technical advice for marketing and assuring product quality. The current elderly and widow stipend programme needs to be extended in both rural and urban areas. Human deprivation, which is primarily the result of inadequate health and educational entitlements, must be addressed with urgency by the Government, NGOs and the private sector. In Bangladesh, rural development and urbanisation are often seen as two distinct processes. Seldom is there any attempt to link rural development with big infrastructure development projects. Such projects concentrate largely in big cities, widening the gap between smaller towns and villages (with little infrastructure and few facilities) and urban areas (with the full range of modern amenities such telephone networks, electricity, roads, and housing). Rural towns and villages also lack sufficient high-quality educational institutions and health services. Until these towns can enjoy comparable earning opportunities and quality of services, the incentive of the rural poor to migrate to Dhaka City is likely to prevail. From this perspective, the upazilla system should be revived to transfer civil and judicial administrations.This would encourage the upper and middle socioeconomic groups to move to those towns and help create demand for investment in infrastructure and social services. NGOs, alongside the Government and private sector, can help provide high quality health, education and financial services in these towns. Government and donor-giving agencies must realise that sustainable development cannot be achieved by confining
people to rural areas. They must acknowledge the inevitably of urbanisation and foster both forward and backward linkages to integrate villages into the overall development process.
REFERENCES
AFSAR, R. (2000): Addressing urban problems: The issue of human settlement and well-being. Keynote paper presented at South-Asian Workshop on Settlement of the Urban Poor: Challenges in the New Millennium organised by the Coalition of the Urban Poor and Association of Development Agencies in Bangladesh, Dhaka. AFSAR, R. (2000a): Rural Urban Migration in Bangladesh: Causes, Consequences and Challenges. University Press Ltd., Dhaka. AFSAR, R. (2000b): Gender, labour market and demographic change: a case study of womens entry into formal manufacturing sector of Bangladesh, in B. Garcia, R. Anker and A. Pinnelli (eds.) Women in Labour Market in Changing Economies: Demographic Issues. Oxford University Press, London (forthcoming). AFSAR, R. (2000c): Mapping Urbanisation and Urban Change for the Twentieth Century:The Case of Bangladesh, a case study prepared under Better Understanding of Urban Change. IIED, London (forthcoming). AFSAR, R. (1999a): Rural-urban dichotomy and convergence: emerging realities in Bangladesh, Environment and Urbanisation,Vol. 1. AFSAR, R. (1999b): Is migration transferring rural poverty to urban areas?, in H.Z. Rahman (ed.) Bangladesh: Poverty Dynamics in the 90s. University Press Ltd., Dhaka (forthcoming). AFSAR, R. (1996):Application of demographic analysis in assessing gender-biased urban poverty in Bangladesh. A paper presented at the national conference of the Australian Population Association,Adelaide, 2-6 December. AFSAR, R. (1995): Causes Consequences and Challenges of Rural-Urban Migration in Bangladesh, Ph.D. Thesis. The University of Adelaide,Australia. AFSAR, R. AND BAKER, J. (1999): Interaction between rural areas and rural towns. A study on Prerequisites of Future Swedish Support to Rural Towns Development and Poverty Alleviation. Agder Research Foundation, Norway. AFSAR, R.,YUNUS, M. AND ISLAM, A.B.M.S. (2000): Are migrants chasing after perilous illusion? A Study on CostBenefit Analysis of Overseas Migration by the Bangladeshi Labour. A Report prepared for the International Organization for Migration (IOM), Dhaka. CHAUDHURY, R.H. AND CURLIN, G.C. (1975): Dynamics of migration in a rural area of Bangladesh. Bangladesh Development Studies, 3(2): 181-230.
NOTES
1 The 1982-83 Government decentralisation policy reclassified urban areas to include not only thana (police districts) headquarters but also rural markets and peripheral areas lying beyond municipal boundaries. It is estimated that reclassification contributed nearly a third to the growth of the urban population between 1974 and 1982-83. Between 1982-83 and 1991 reclassification added another one million to the urban population. 2 Residential and temporal criteria are most commonly used to distinguish between different categories of migrants. Exclusive use of residential criteria often bypasses circular migrants, sojourners and those migrants who retain several places of residences, such as nomads or migratory labourers. Paradoxically, by defining a migrant as a person who changes his/her residence for a period of six months or more, the Bangladesh Bureau of Statistics excludes most non-permanent movers such as commuters, seasonal migrants and temporary migrants (BBS, 1989:160). 3 HPI is composed of three indicators to capture three types of deprivation: deprivation in health, knowledge and in overall economic provisioning. Deprivation in health is indicated by vulnerability to death at a relatively early age quantified as percent of people likely to die before reaching 40 years. Deprivation in knowledge is represented by the percentage of adults who are illiterate. Deprivation in overall economic provisioning is represented by percentage of people (i) without access to safe drinking water, (ii) without access to health services and (iii) percentage of children under 5 who are moderately and severely underweight 4 Northern districts have the worst HPIs and income poverty but migration to big cities is, however, uncommon from these districts.
HOSSAIN, M., AFSAR, R. AND BOSE, M. (2000): Growth and distribution of income and incidence of poverty in Dhaka City, in H.Z. Rahman (ed.) Bangladesh: Poverty Dynamics in the 1990s. University Press Ltd., Dhaka (forthcoming). KANNAPPAN, S. (1985): Urban employment and labour market in developing nations, Economic Development and Cultural Changes, 33(4): 699-730. RAHMAN, H.Z., HOSSAIN, M. AND SEN, B. (1996): 198794 dynamics of rural poverty in Bangladesh. Final Report (mimeo.). BIDS, Dhaka. SEN, B. AND ALI, Z. (1999): Poverty and Human development in Bangladesh, Background paper presented for the Bangladesh Human Development Report. BIDS/Planning Commission, June (draft). TODARO, M.P. (1969): A model of labour migration and urban unemployment in Less Developed Countries. American Economic Review, 59(1): 138-148. WORLD BANK (1996): Bangladesh: labour market policies for higher employment. University Press Ltd., Dhaka.
97
The informal institutional framework in rural Bangladesh S.Aminul Islam

Eight out of ten people in Bangladesh live in rural areas and have little access to formal institutions. For them, and therefore for most of the countrys population, informal institutions exert the most influence on their day-to-day lives and livelihoods. Despite 50 years of research into informal institutions, the resulting body of knowledge still has serious gaps.This paper draws on that body of knowledge and new research findings to provide a fresh perspective on informal institutions. More specifically, the paper explores the nature of informal institutions, institutional change, characteristics of informal leaders, the nature of samaj and shalish and the relationship between informal governance and livelihoods. European observers was the lack of corporate institutions, in contrast to other regions of South Asia.They saw it as an atomised society or, more correctly, an example of institutional atomisation (Tepper 1976) a structural deficiency that accounted for a broad range of negative societal characteristics such as lack of public spirit, factionalism, etc. Scholars both outside and within the country have held this view. Recently, Khan (1996) has shown that the predominantly deltaic ecology of Bangladesh has influenced rural settlement in the country. Bangladeshi villages are small, and can be called open villages with a dispersed, linear pattern of settlement, different from the close or corporate villages found in other parts of South Asia.Villages in Bangladesh are not administrative units; they have no officials except watchmen in some places. Settlement is based on the para a cluster of hamlets surrounded by trees and separated by low-lying fields. This pattern militated against tight integration of the population and a common identity within a local space. This vision of the village led to the coinage of the phrase elusive villages (Bertocci 1970). It does not mean that the community at the village level is incapable of corporate action. The structural characteristics merely reflect the limited need for concerted action for common good within the larger social space of the village.
CONCEPTUAL ISSUES
In anthropology and sociology, institution refers to an endurable status and role, sets of which collectively shape the behaviour of a group of people (Wallis 1985; Jary and Jary 2000). More broadly, institutions are valued, stable, and recurring patterns of behaviour (Huntington 1968; Uphoff 1986).A second view of institution has emerged within economics; here, institutions are seen as the rules of the game that govern peoples interactions (Eggertsson 1990; North and Thomas 1995; Ostrom 1996). Informal institutions refer to small, indigenous bodies with some specification of role and status. Informal institutions are more prevalent in rural life and shape the way that people behave in rural society. Research indicates that informal institutions in Bangladesh were viewed as a throwback that would eventually disappear with the triumph of modernity. Here, we focus on two key institutions samaj and shalish. Building on earlier work (Rahman, H.Z. and Islam, forthcoming), this paper maps out the underlying patterns, processes and possibilities.
SAMAJ Bertocci (1996) argues that the Bangladeshi village has few institutions beyond the kinship grouping but the samaj is one institution that is present in every village. Samaj is an institutional space for collective worship, performance of rituals and festivals. It has three overriding characteristics. First, samaj is a group. Secondly, it is autonomous, having no connection with any higher religious or secular authority.Thirdly, its social control mechanism is psychological coercion or manipulation according to socially constructed notions of honour and shame. Bertoccis description, although long, is worth quoting (Bertocci 1996: 17):
It is in such groups, rather than in what is otherwise designated as the village per se, that
HISTORICAL ORIGIN
One of the most striking aspects of Bengali society that captured the attention of the colonial government and
Bengali Muslims act out what is morally and ritually meaningful as community, celebrating important religious festivals, coming together in prayer on critical days of atonement, and also reinforcing when necessary the normative standards to which all are supposed to adhere. It is noteworthy that these groups are commonly called samaj, which is the generic term for society in Bengali and other Indo-Aryan languages, and in some parts of Bangladesh they are also known as reyai which connotes follower or citizen and by extension implies the protection of those who submit. Samaj is the moral universe of people scattered over the forest clearings and the cascading floodplain.The two key normative idioms of this universe are takdir and tadbir, or divinely ordained destiny and human agency, respectively (Hara 1967).The peasant life is determined by the will of Allah to whom all must surrender totally.The developed notion of individualism in Bengali culture is manifested in tadbir individual striving and achievement. It refers to action that is goal-oriented; and opens before the peasant the horizon of power.
a major function of samaj is conflict resolution through shalish an indigenous judicial system that settles disputes quickly and inexpensively through minor sanctions; q it confronts external agencies and negotiates with them on behalf of the group; q it plays an important role in local elections by influencing the voting behaviour of the members. It may even turn out to be a vote bank; q samaj provides a social basis for new institutions in the village. People recruited from different samaj groups can reintroduce factional conflicts in these new institutions and render them ineffective.
q
Along with samaj, a faction is another pervasive solidarity group in rural Bangladesh. If samaj represents the moral universe of the peasants, the faction is the domain of power articulated through leader-follower or elite-subordinate relationships.Whilst samaj embodies the members willingly submitting to the idea of collective identity and action, the faction is formed out of the power or khamata of an individual or group of individuals, which is exerted on members and which casts a shadow in which nonmembers live (Bertocci 1996). Some critics, however, suggest that the image of samaj as a model of solidarity should not be over-emphasised because it does not ensure a high level of social control in the rural society (Khan 1996).
Hossain and Westergaard (1998) show the role of samaj in the UP election of Boringram in Bogra. In Boringram, the villagers were interested in getting a person of their village elected. But there were three candidates from the village and none of them were likely to get elected if the villagers votes were split. So a meeting was called in the premises of the big mosque to resolve the problem.The samaj leaders were instrumental in deciding upon one candidate and others were paid their expenses.
RECENT CHANGES IN SAMAJ

Recent studies of samaj tend to suggest that it has shown a great deal of resilience in the face of social change. Adnan (1997) reports that the ten villages he revisited showed remarkable stability in the organisation and function of samaj. Most samaj groups identified in 1975-76 still existed in 1985. Some, formed in the 1940s, had also endured. But samaj groups had also suffered from growing too large or from internal conflicts and factionalism, which led to their breakdown. In 1947, the villages had ten muslim samaj.This figure increased to 17 in 1975-76 and 34 in 1985. In Boringram, Hossain and Westergaard (1998) found that initially samaj was coterminous with the village and the richest man was the headman. Factional conflict after the death of the headman led to the break up of samaj and its splintering into ten new samaj groups in the village.
Samaj has several functions ( Bertocci 1996):

q it
constructs and maintains the mosque, mandir or other place of worship, organises collective worship and other religious ceremonies; q it regulates social ceremonies and life-course events. It decides who is to be invited to marriage and other feasts;
The informal institutional framework in rural Bangladesh 99
Although the authority of ten new groups is less meaningful than the authority of the previous single samaj, they nonetheless retain some degree of relevance, as described earlier. In Jagatpur, a village in Narail district that Siddiqui (2000) revisited, samaj seemed to have lost its moral authority and, over the past 20 years, been replaced by factions. In work in char lands of Kurigram, Hossain found two types of samaj a lineage samaj comprising the extended kinship group and another type comprising neighbourhood members (Hossain 2000). Samaj was of particular relevance in the char lands, providing support in crises and playing a key role in conflict resolution in an area that was little served little by the State machinery.
Improved infrastructure, increased mobility of labour and the growth of the non-farm sector have loosened patronage ties. But as a result of this process the destitute widows, the old, the infirm, divorced women and other marginalised groups have found themselves exposed to greater vulnerability (Jansen 1988). The patronage system, in spite of its limitations, provided the poor with a safety net.
SHALISH Shalish is so central to rural society that it deserves separate treatment. It can be thought of as a village judicial system in which a panel of representatives adjudicate over disputes. It is best described by one of Adnans (1997) respondents:
Shamaj means unity. In order that people may live in peace, the operation of a shamaj is required.Those living in a shamaj have to observe certain rules If any person belonging to the samaj commits an offense, it is judged by the shamaj. Apart from the leader of the shamaj concerned, other shamaj leaders and important persons of the village may be called upon to preside as shalishkars. The verdict is usually verbal and declared publicly, in the presence of all. If someone is judged to be guilty in the shalish, he is punished. Should any party not agree with the judgment, attempts are made to persuade him and his supporters to accept it. Hossain and Islam (forthcoming) observe that shalish takes place quite frequently; it is quite popular and an effective mechanism for conflict resolution at local level. It is very inexpensive; disputes are mostly resolved in three sittings by an assortment of local leaders from different social strata. The traditional shalish is changing. The rich no longer monopolise it and a variety of strategic actors, such as elected UP representatives, now play a greater role in the shalish. New mechanisms for conforming to the verdict of the shalish are being developed.Although it had a reputation for unjust treatment of the poor and for closing its doors to women, the shalish has potential as a powerful instrument for local justice.
PATRON-CLIENT RELATIONSHIPS IN RURAL SOCIETY

The patron-client relationship in rural Bangladesh is born of high insecurity both psychological and material on the part of the poor client. In deltaic ecology where life, as one scholar (Abecasis 1990) describes it, hangs by a thread, the need for patronage is strong. Traditional patronage emanates from extreme scarcity of land and endemic land conflicts. Land owners permit sharecroppers to cultivate part of their land in exchange for a share of the harvest. Jansen (1988) highlights the pervasive ties of the patronclient relationship in rural Bangladesh, which was created and sustained by the zamindari system during the colonial period. After 1950, patronage became the prerogative of farmers with surplus land. Jansen (1998) notes several factors that were responsible for the emergence of the patron-client relationship: a multiple inheritance system, individual property rights to land, scarcity and unequal distribution of land, scarcity of employment and other sources of livelihood, need for protection and lack of access to resources provided by the State. The patron-client relationship in rural Bangladesh has declined significantly.The emerging form of patronage is more to do with the penetration of macro-politics into the rural space and peoples need for protection against escalating violence.
LEADERSHIP IN RURAL SOCIETY

Community leadership at the grassroots level is increasingly recognised as critically important, given the crisis in formal institutions, particularly in the sphere of local justice and conflict management.The excerpt below encapsulates the research on the changing pattern of rural leadership in Bangladesh (Islam 1974; Barman 1988; Jansen 1988; Islam 1989; Karim 1990; Rahman, H.Z. and Islam, forthcoming). In pre-colonial Bangladesh, leadership in rural society fell to four categories: the zamindar and/or his officials, religious leaders; caste leaders; and lineage elders. The zamindars role was concerned mainly with collecting taxes and so zamindars generally did not interfere with village autonomy. q Between 1757 and 1869, zamindars assumed greater decision-making powers on all matters bar religion in the village. The role of the other groups remained as before. q The setting up of Union Boards in 1870 led to the emergence of the dichotomy between formal and informal institutions at village level. From this time a new type of leadership emerged which had formal authority backed by the Raj. The traditional samaj remained a strong springboard for accessing formal authority.
q
setting up new organisations and participating in local government. Some younger people from the category of middle farmers or below were being able to take up leadership roles; q factional leadership was increasing; q brokerage roles emerged through which rural people could negotiate with the external world; q occasionally brokers became touts who charged high fees for their services, or even cheated poor people (Rahman 1989); q patrons as leaders were gradually losing their influence; q political parties were establishing themselves in villages, which led the emergence of party leadership at the local level; q development activities were taken up more by new leaders who were younger and better educated.
NEW RESEARCH ON INFORMAL INSTITUTIONS

New research shows that informal institutions have proliferated at the village level. Micro-credit groups, local credit societies, community-based organisations, clubs, cultural organisations, bazaar committees, water users associations, etc., have contributed to a new institutional density at village level. Moreover, the leaders of formal bodies also play out their roles, either informally, or within the informal space. There is a tremendous pool of governance capacity among people at the grassroots level. This capacity is displayed spontaneously and within the framework of short-term collective action. It was found that people could identify eight functional domains pertaining to governance and development: dispute resolution; maintenance of law and order; disaster coping; protection from harassment; development; social asset maintenance; environment management; and social welfare. People could also identify leaders and institutions most appropriate for each of these domains. The governance capacity of the local community was dramatically visible during the flood disaster of 1998, which galvanised public spirit and stimulated leadership initiatives and collective management of disaster (Rahman, H.Z. and Islam, forthcoming). An important change is the gradual recognition of women as leaders, through their membership of union parishads.
This pattern held until the early 1960s, when the Comilla model of rural development initiated a new process of institution-building and modernisation of agriculture. The process of modernisation in rural society led to changes in the pattern of leadership. Research between the 1960s and 1990s found: leadership roles rested with the rich families. Leadership was founded on high land holding, high status, seniority in age, education, and rhetorical ability and skill in shalish; q a new type of leadership was emerging in the Comilla area, and later in other parts of the country, which consisted of cooperative managers, model farmers, schoolteachers, tractor drivers, etc.; q in other areas there was a slow process of change in which more educated and younger people of the dominant families were entering the leadership arena by
q
RULES OF THE GAME IN RURAL SOCIETIES

Following Bertocci (1996), it is possible to develop two models of the rules of the game in rural society (the actual rules are much more complex). The first is a model of peasant solidarity, grounded in trust, reciprocity, deference and equality. A man gains social status and capital as he redistributes his material assets. Property rules are diffuse and the notion of common property is strong. A second set of rules operates in the power domain. It refers to a social terrain where people are pitted against one another for securing livelihoods in an extremely resource-poor situation. It subverts the world of peasant solidarity without having a normative framework of its own. It creates an anomic landscape in which society is fractured by competition and the deployment of violence for securing resources (Hartman and Boyce 1983) a landscape most visible in the char lands of the country. These rules of the game and their complexity in the lived peasant world of Bangladesh can be viewed more closely through the ethnography of Khan (1989) and the fieldwork of the present study. In the three villages of Daudkhandi near Dhaka, three different sets of game rules operated. In Malikhil, the undisputed headman represented the solidarity of the village.There was no competitor to challenge him. He was also the UP chairman and, in that capacity, he was able to secure resources for the village. So he had high legitimacy.The rules of the game were framed in terms of the traditional authority. In Shigula, there were factional struggles after the death of the last major leader. Eventually a leader emerged who fomented violence and litigation in the village in order to benefit from it. His main strategy of resource generation was exploitation of the poor and the vulnerable. Raipura provided a typical scenario of factional politics. Here the last undisputed leader had died quite some time ago and five families were competing for dominance that resulted in endemic feuds and a string of legal cases. As the faction followers were the poor and the vulnerable, they suffered most from physical injury or litigation. A field study conducted by this author in four villages presents a more complex scenario.The four villages (not their real names) Maniknagar in Nawabganj, Haorpur in Sunamganj, Ratanpur in Shariatpur and Lalmai in
Comilla have different constellations of normative and institutional rules. MANIKNAGAR Maniknagar had three samaj groups in 1947. Now there are ten. Most of the new samaj groups have been formed out of the old ones. Factional conflicts were mainly responsible for the break up of the old samaj groups. The moral authority of elders has declined greatly. Younger people are increasingly assuming leadership roles. The national political divide, characterised by mistrust and conflict between the two major political parties (the Bangladesh Nationalist Party (BNP) and the Awami League) has impinged upon the village. There is some effort in constructing a solidarity group modelled upon the Islamic umma1 and in establishing the normative system of orthodox Islam. It is an attempt at constructing a new moral order on the basis of Islamic high culture in the face of substantial decay of the tradition of the village. Thus several normative patterns co-exist and have led to fuzzy rules of the game. HAORPUR The traditional samaj is powerful in the village.There are five samaj groups in the village and all of them date from the pre-1947 period. These samaj groups are generally lineage-based and there is considerable reciprocity within samaj. People help each other in looking after cattle in the field, lend carts and agricultural implements and provide interest-free credit.The wells in the haor2 that contain fish in the winter are leased and the money used for reconstructing earthen roads, which are washed away every year. The village shows some traces of peasant solidarity. But many of the samaj leaders are losing land and consequently authority. Upon this declining samaj there emerged a Bangladeshi who had returned after some time in England. He built legitimacy in the village by displaying his wealth and distributing his resources and eventually he became the UP chairman. He gathered around him a group of strongmen and deployed them to grab land from the Hindu families of the village. Some of these families are now totally dependent on him. This is a new form of patron-client relationship.The UP chairman has abandoned the formal office of the chairman and the office is wherever he is.Thus he transformed formal authority into informal domination. As in Max Webers work (1978), the process can be described as
traditionalisation of legal-rational authority in the direction of patrimonialism played out at micro-level, but more evident at macro-level and connected with it. RATANPUR Ratanpur is an example of a traditional pluralist community. There are five samaj groups, which exist in a weakened state. There is a famous syncretistic saintly complex in the village which has many devotees from all over the country. But the majority of the villagers do not belong to the sect or tariqua (way) of the saint. As a consequence the saint or his subordinates do not interfere in the routine affairs of the village. There is also a strong tradition of indigenous Bengali culture, which partially balances out the other forms of domination. The village has been subject to political attention in recent years. Macro-level party politics is increasingly subverting the pluralism of the village as external actors vie for control of its affairs. A recent murder was a flashpoint for party politics.The rules of the game, which were largely positive for the promotion of livelihoods, are becoming blurred. LALMAI Lalmai has two congregations during Eid-ul-Azha; sacrificial meat is distributed separately, hence one can technically talk of two samaj groups in the village. But for most purposes the village has one samaj. The villagers belong to cults of different saints, but these worlds do not collide with one another or the secular order. In fact, the activities of the samaj have been taken over by a CBO a village credit society. It can be described as an emerging civic institution. It has contributed substantially towards setting up a high school, promoted sanitation in the village and is trying to bring a gas supply to the village.The followers of different parties remain united for the common interest of the village. Lalmai represents the normative pattern of a modernising community. The four examples, although brief, show that there is no deep structure that imprisons human behaviour in rural Bangladesh, either within a matrix of under-development or within a high culture that limits prospects for improving livelihoods. Rather, there is a great deal of heterogeneity within rural society. Some have dysfunctional consequences and some have positive outcomes for livelihood gains of the poor.
CONCLUSIONS AND POLICY IMPLICATIONS

In the past, informal institutions have been viewed in anthropology and sociology mainly as dysfunctional throwbacks destined to disappear as modern formal institutions render them obsolete.This paper, based partly on new research, shows that informal institutions are remarkably resilient; they are complex, dynamic and with diverse forms that produce different outcomes for livelihoods. The institutional space is being constituted and reconstituted as external forces impinge upon them. There is now a growing realisation that informal institutions have significant potential for enhancing livelihoods. The critical strategy is to change the rules of the game. In Bangladesh, there is a crisis in governance. Informal institutions are therefore particularly relevant for village communities. Institutional development from the bottom up is a complex process that demands a great deal of knowledge about the existing normative and institutional framework and appropriate intervention strategies that can change the rules of the game to be more relevant and beneficial to more people.The example of Lalmai hints at a new institutional framework. But there are other possibilities that can be tapped and moulded through an institutional development policy. Shalish is an example of indigenous governance capacity. It is inexpensive, quick and effective. It reinvigorates the moral order of the community. NGOs are now trying to make shalish more egalitarian and gender-sensitive.This is a major area of policy intervention. The impact of rapid changes in infrastructure development and mobility has rendered the old samaj and its control mechanisms vulnerable in many areas. A new or modified shalish system can reinvigorate these methods of social control and help to create a new environment of governance that is more sympathetic to the needs of the poor. The development of community coping mechanisms at times of crisis is another potential area of policy intervention. Community policing, for which anthropological evidence is available, is a third area of potential intervention, especially in the context of declining law and order in rural areas.The fourth important area for policy intervention is in forging a synergy between formal and informal institutions. UP leaders and even party and trade union leaders are taking part in the shalish and institutionbuilding in rural society. Training of formal leaders for informal institution-building (changing the rules of the
game towards efficiency and equity) may be a promising area for exploration. The training and involvement of women UP members in shalish is critical here. Good governance and poverty reduction in Bangladesh may well rely on the adequate further development of informal institutions.
REFERENCES
ABECASIS, D. (1990): Identity. Islam and Human Development in Rural Bangladesh. University Press Ltd., Dhaka. ADNAN,S. (1997):Class,Caste and Samaj Relations Among the Peasantry in Bangladesh: Mechanisms of Stability and Change in the Daripalla Villages, 1975-86, in J. Breman, P. Kloos and A. Saith (eds.), The Villages in Asia Revisited. New Delhi. Oxford. BARMAN, D.C. (1988): Emerging Leadership Patterns in Rural Bangladesh:A Study. Centre for Social Studies, Dhaka. BERTOCCI, P.J. (1970): Elusive Villages: Social Structure and Community Organization in Rural Bangladesh. Ph.D. dissertation. Michigan State University. BERTOCCI, P.J. (1996): The Politics of Community and Culture in Bangladesh: Selected Essays. Centre for Social Studies, Dhaka. EGGERTSSON,T. (1990): Economic Behaviour and Institutions. Cambridge University Press. HARA,T. (1967): Paribar and Kinship in a Moslem Rural Village in East Pakistan. Ph.D. dissertation. Australian National University, Canberra. HARTMAN, B. AND BOYCE, J. (1983): A Quiet Violence: View from a Bangladesh Village. Zed Press, London. HOSSAIN,A. (2000): Livelihoods in Char Lands of Northeast Bangladesh; An Analysis of Governance Issues and Institutional Arrangements, Annexe 3. The Chars Livelihoods Assistance Scooping Study. DFID, Dhaka. HOSSAIN, A. AND WESTERGAARD, K. (1998): The 1997 Union Parishad Election in Bangladesh: A Case Study of a Village in Bogra. Paper presented at ENBS Workshop in Bath, 16-18 April. HUNTINGTON, S.P. (1968): Political Order in Changing Societies.Yale University Press, New Haven. ISLAM,A.K.M.A. (1974): A Bangladesh Village: Conflict and Cohesion An Anthropological Study of Politics. Schenkman, Cambridge. ISLAM, S.A. (1989): Social Change, Power and Legitimacy in Rural Bangladesh. Journal of Local Government,Vol. 18, 1:98-141. JANSEN, E.G. (1988): Rural Bangladesh: Competition for Scarce Resources. University Press Ltd., Dhaka. JARY, D. AND JARY, J. (2000): Collins Dictionary of Sociology. Harper Collins, Glasgow. KARIM,A.K.Z. (1990): The Pattern of Rural Leadership in an Agrarian Society: A Case Study of Changing Power Structure in Bangladesh. Northern Book Centre, New Delhi. KHAN,A.A. (1996): Discovery of Bangladesh: Explorations
NOTES
1 Umma means the community of Muslims, that is, the totality of all Muslims in the world. 2 A haor is a depression in the floodplain.
into Dynamics of a Hidden Nation. University Press Ltd., Dhaka. KHAN, S.A. (1989): The State and the Village Society: The Political Economy of Rural Development in Bangladesh. University Press Ltd., Dhaka. NORTH, D.C. AND THOMAS, R.P. (1995): The Rise of the Western World:A New Economic History. Cambridge University Press, Cambridge. OSTROM, E. (1996): Incentives, Rules of the Game, and Development.Annual World Bank Conference. RAHMAN, A. (1989): Rural Power Structure and Development:A Review Study (mimeo.). BIDS, Dhaka. RAHMAN, H.Z. AND ISLAM S.A. (Forthcoming): Actors and roles; Community Capacities and Local Governance in Bangladesh. University Press Ltd., Dhaka. SIDDIQUI, K. (2000): Jagatpur 1977-1997: Poverty and Social Change in Rural Bangladesh. University Press Ltd., Dhaka. TEPPER, E.L. (1976): Administration of Rural Reform: Structural Constraints and Political Dilemmas. In R.D. Stevens et al. (eds) Rural Development in Bangladesh and Pakistan. University of Hawaii, Honolulu. UPHOFF, N. (1986): Local Institutional Development. Kumarian Press, Connecticut. WALLIS R. (1985): Institutions. In A. Kuper and J. Kuper (eds.) The Social Social Science Encyclopedia. Routledge Kegan Paul, London. WEBER, M. (1978): Economy and Society.Translated and edited by G. Roth and C. Wittich. Chicago University Press, Chicago.
105
The formal institutional framework of rural livelihoods in Bangladesh Paul Thornton

The distinction between the formal and informal aspects of the institutional framework is somewhat arbitrary. Informal forms, that is, social structures such as class, caste and gender, power relations; shalish, samaj and other traditional decision-making systems, directly influence the way in which formal structures within the public or private sectors, or within civil society, are organised and operate.This paper should therefore be read alongside that on the informal institutional framework and the two, together with the linkages between them, should be seen as describing a single overall framework. The institutional framework within which people pursue livelihoods simply means the structures and processes that overtly or covertly govern the way life is organised. At a formal level, these structures are formed and governed by statute, law or other transparent, accepted and legitimised systems. The processes that determine how those structures operate are again the formal and accepted ways of working. These may be official policies and practices or they may be the policies in practice: those that have been adopted as normative within the given context. This paper focuses on what happens in practice; with the rules of the game and the ways in which formal institutions are interpreted and negotiated in response to the influences of power, personality, process and external influences. The paper begins by reviewing the development of the policy and institutional map of rural Bangladesh. The current formal institutional features and their relationships are then identified. Finally the institutional trends and the implications of the emerging patterns for livelihood interventions are assessed.The paper concludes by identifying interventions that could catalyse these changes and bring livelihood benefits, especially for the poor. local community, dates from the earliest history. The district structure with administration, taxation and law combined under the representative of the colonial authority or the State dates to the British period. Rural self-government in the sub-continent is as old as the villages themselves. The earliest records refer to a model of village-based self-government. Rural communities were more or less left to themselves until the Moghul period; substantive change came during British colonial rule (Siddiqui 1998).A somewhat idyllic myth is portrayed (Khanna 1977) of a gracious community spirit.The reality was a peasant culture ridden with class and caste differences in which political and social organisation took a variety of forms but was characterised by feudal domination and patronage (Tinker 1989). The Moghuls introduced the first serious attempt at reform with the establishment of a structure beyond the village. Four administrative levels were introduced initially for tax collection. Revenue collection was linked to the responsibility for maintaining law and order with some judicial and civic functions. Responsibility at village level rested with the headman, who was assisted by the earliest form of chowkidar to perform the policing duties.The taxation system led to the emergence of a class of revenue farmers (zamindars) and other middle-class citizens between the tax administration and the taxpayers. The British focused predominantly on urban areas and only turned to the rural communities later.They institutionalised the zamindari system as a permanent middleclass, with absolute rights to land to ensure tax collection. Traditional village self-government became inoperative. The present structure and style of local government largely reflects the British model, with an imposed administrative system that linked tax collection with the rule of law (the magistrate and the collector) and provided an administration that focused on the infrastructure to support both. Independence and partition have not changed the fundamental nature of government structures. The district
THE ROOTS OF INSTITUTIONAL TENSIONS

Aspects of local governance systems in Bangladesh date back to the Moghul period and much earlier.The pattern of externally imposed rule of law and taxes from an authoritarian ruler, with other responsibilities left to the
administration still represents State authority with no local accountability. Local bodies were introduced towards the end of the British period with a mixture of elected and nominated members. They have continued intermittently with a series of experiments determined by political whim, but have yet to become effectively established.The Upazilla Parishad Act of 1998 and the recommendations of the Local Government Commission (GoB 1997) set the context for the next round of initiatives. Thus within this historical context lay the roots of tensions between formal institutional actors.The first relates to the institutional dislocation of the micro-level (village/union), where a long tradition of self-governing informal institutions confronts more formal structures emanating from outside. The representatives from outside (public servants) are often absent and the representatives of self-government, even when elected to local bodies, do not yet have effective power or voice. A second tension exists between politicians and the bureaucracy. The long tradition of bureaucratic dominance of the policy process and implementation management is founded on a strong power base that is deeply dysfunctional. Politicians, in contrast, have relatively little experience of democratic politics and even less of policymaking. Relations between the two are mediated by personal relationships and access to resources, rather than political policy and accountability. Many politicians still define their role in terms of access to power and patronage, rather than policy formulation. The bureaucracy therefore continues to dominate, and with the political elite divided over the rules of the game, this is set to continue. Corruption and inefficiency on the part of the bureaucracy is matched by petty politics and dissonance within parliament and outside on the part of politicians. The result is a lack of leadership and direction from either group, which is leading to popular disenchantment. A third, but to date quite weak, tension exists between politicians and the emerging civil society. The rapid growth of the NGO community has created a new knowledge base on policy issues. Other actors, including the print media and some key policy institutes and pressure groups, have also emerged within the policy arena over the last decade.This has resulted in the growth of a more aware civil society, but as yet impact has been
limited. Popular disillusionment with politicians and the State seems to result in a fatalist response rather than fuelling this emerging tension.Thus, particularly at village level, the absence of teachers and other public servants, the self-interest of MPs and local politicians, and the mismanagement of development is accepted and ignored and livelihoods are pursued despite these blocks.
THE CURRENT STATUS OF FORMAL STRUCTURES AT THE LOCAL LEVEL

LOCAL STATE INSTITUTIONS Currently, only elected UPs are in place. Legislation for zilla and gram (village) parishads has been passed, but there have been no elections to date.The 1997 elections saw a return to a fully-elected UP and women members elected for the first time. Some union chairpersons are fulfilling an evolving democratic function while others continue to function as new elements of the elite establishment. It is too early to draw firm conclusions: the roles of the UP and other agents at the local level, and the relationships between them, have yet to emerge in any cohesive form. As members of the Upazilla Development Committee, the official responsibilities of Union Chairpersons extend only to making recommendations on the development budget. However, there is already evidence of influence by chairpersons on main programme planning and budgeting, especially by the Upazilla Engineer (infrastructure) and also of influence over government officials deployed at union level (e.g. block supervisors, union health workers, teachers). MPs act as advisers to the upazilla parishad and use this role, their discretionary powers and party connections, to intervene at the local level. Equally, where there are good linkages between the local politicians and those at the national level, lobbying and favouritism are evident. Some of this is simply part of normal political business, but at times it bypasses established systems, lacks transparency and operates outside the accepted bounds of representational practice. The next round of elections will further establish the UP. It may become more capable of playing a more effective
The formal institutional framework of rural livelihoods in Bangladesh 107
democratic role: one that is transparent, has emerging aspects of gender equity and plays a central role within a system of local governance that has centres of power other than the traditional ones. LOCAL BUREAUCRATIC INSTITUTIONS The lowest bureaucratic tier of management is the upazilla level.The Upazilla Nirbahi Officer, effectively the chief executive and co-ordinator at the upazilla level, is on deputation from the Centre, as is the Deputy Commissioner at district level. All other officers are managed through district and divisional levels to their respective Ministries in Dhaka. The Nirbahi Officer has a co-ordination role for development and disaster, and can influence but has no direct authority over his colleagues from line departments. Most major decisions on planning, budgeting, programme and human-resource management are taken at the district level.The Upazilla Officer makes plans and budgets that are subject to approval at the district level, where line department control also rests. Management and supervision of staff below upazilla level is weak.The bureaucracy is pre-occupied, even at upazilla level, with financial and other resource allocation, including the associated rent seeking, and less with the delivery of responsive services that are accountable to the public. NGOs AND CBOs The growth of NGOs in Bangladesh has occurred in response to the fragile and fragmented development of formal institutions on the one hand, and trends in donor aid on the other (Ahmad 1999).The sector is dominated by a small group of Dhaka-based national NGOs, both financially as well as in coverage and activity span (Ahmad 2000). The movement of many NGOs into the microfinance industry has led to competition over client base rather than collaboration between communities. Institutionally these large NGOs have similarities with the public sector centrally controlled, bureaucratic in nature, weak on accountability and disconnected at local level (Thornton et al. 2000). Some have identified NGOs as the source of new patrimonial relationships (Devine 1999). The current picture is of national NGO presence in a significant number of villages, often with several active organisations in the same locality. There are concerns within the NGO community about overlap and competition, but steps are only just being taken to address
these.The typical situation is of NGOs providing services (social, financial, community organising), but in independent and disconnected ways. The connectedness of national NGOs to local bureaucratic and political systems is weak and the UP is not frequently engaged. There is little evidence of engagement between the emerging peoples organisations developed by NGOs and the representative structures of government. Local NGOs are numerous, but less well-developed. Even these are most often organised at union, upazilla or district level rather than in the village. Below union level there are self-organised groups, co-operatives and other CBOs with no paid staff, essentially self-help in orientation and usually linked to the prevailing social structure. CIVIL SOCIETY Civil society structures (charities, lobby groups, professional bodies, media and other networks) exist at district level, sometimes in quite sophisticated forms. Organisation around policy issues (for instance, the environment in the Southwest and Chittagong Hill Tracts; the shrimp industry in the Southwest; and land tenure issues nationwide) has led to networks within and across districts. Below upazilla level, however, there is little of a formal nature. At union and village level civil society remains dominated by informal institutions and social norms. The movement of NGOs into the micro-finance industry has, in some cases, led to competition for clients and a distancing from local communities at the very time that some of the samity mobilisation was beginning to make a definitive impact. THE PRIVATE SECTOR Until recently at village level, the private sector consisted of local shops with a limited range of goods, artisans and cottage industry. Infrastructure development, notably roads, bridges and electrification has supported a dramatic change. National and international suppliers and wholesalers are active in even quite remote locations. This has led to agencies and networks for supply and marketing that link the village through (or bypassing) the district to national and international levels. Internal quality control and regulation systems seem to be part of these processes. Formal organisation of the private sector, other than at bazaar and hat (country market) level, is still limited below district level.
DYNAMIC RELATIONSHIPS AND PROCESSES AT THE LOCAL LEVEL

The local institutional framework is changing, rapidly resulting in an increasingly complex set of institutional forms and relationships. The only public-sector officials resident at village level remain the teacher, block supervisor and village health worker, and only the teacher has a permanent daily presence (or absence). But NGOs, ward-level politicians and an increasing number of professional traders (chemists, engineers) are now a regular, if not permanent, part of the framework. Education may be delivered by madrassah1, State, NGO, private or community schools; health by a similar range of providers; the private sector is more likely to advise on agricultural extension than is the block supervisor. Market forces are emerging as a more significant element of the rules of the game. Relationships between the private sector, NGO, UP and Government officials are tentatively being explored. Health Watch and other consumer voice groupings supported by NGOs are still rare examples, but these, and other forms of popular voice, are beginning to surface. UP members are playing stronger roles as chairs of formal and unofficial committees and bodies. Decisions that affect the village are increasingly being made in the village (in the UP office, at the NGO centre, by a large employer). These processes are embryonic in their development but there is a shift in the balance between the influence that informal and traditional power bases and the more transparent and formal processes have on the rules of the game. However, informal institutions and deep social norms still prevail and at times are amplified (e.g. rent-seeking and corruption) by the developments that are taking place. The relationships between these varied actors are a potential entry point for change. It is not just a question of who does what, but of building on institutional synergies and maximising the value added by particular actors. Despite some positive institutional developments, however, the village remains disconnected. Access to economic and social space for livelihood gains is determined and controlled at upazilla and district levels. The real scaling up of new livelihood opportunities depends on infrastructure, markets, management and supervision of staff (to ensure teachers are in school), sources of credit
(banks and NGOs), human capital (functioning and staffed schools and health facilities) and development investment.Whilst the emerging institutional framework within the village or union can maximise the space available, it is the institutional, economic and investment support at meso-level that is required to transform that process. Micro-level (union) operational opportunities cannot lead to a transformation of the institutional framework within the public sector without policy linkages through the meso-level (district) to the macro-level (national).
COMPLEX AND CHAOTIC POLICY PROCESSES

Policy emanating from Dhaka is disconnected from institutional practice at the local level: politicians do not own sector policies. Senior bureaucrats, often heavily influenced by donors or other external investment pressures, most frequently frame policy, which is therefore separated from the political process. Implementation depends on efficient management systems,which are absent in most cases. Operationally, the district is the fulcrum, the point at which national policy as articulated is translated into policy for practice. Operational realities (unskilled and poorly motivated staff, leakage of resources, poor supervision and monitoring) and a different set of policy priorities lead to district-level bureaucrats implementing a different policy possibly more achievable, possibly more appropriate but also not as intended.These are not conscious changes, but the institutional process of implementation. Then at village and union level, where again political ownership and understanding is absent, as are staff and resources,policy for practice becomes policy in practice: a long way from the original intent and often bearing no relationship to the felt needs of the institutional voices at the micro-level. This process is not simply the result of corruption or bad planning at the centre. It is more the natural consequence of institutional weaknesses in terms of poor management and supervision, excessively long chains of command with no political interface, the continued assumption that the State at the centre can best determine the service that should be provided by its representatives at local government level.
PROMOTING LIVELIHOODS: THE INSTITUTIONAL CHALLENGES

DEALING WITH COMPLEXITY AND EXCLUSION Institutional complexity can bring confusion and duplication rather than collaboration and co-operation.There are new rules and new systems and structures. This can make the benefits inaccessible to some, often the poor and vulnerable who lack knowledge or skills to operate in a new institutional context.The very poorest (those without skill, credit, land and social support) remain institutionally excluded from most opportunities. What is required are new maps to understand and build on the new institutional opportunities. Co-operation and collaboration in identifying functional synergies should be encouraged whilst establishing processes that minimise inefficiencies and exclusion. INCREASING TRANSPARENCY Contracts for employment and land holding, open systems for credit from MFIs and so on build confidence in formal institutional relationships. If this were to be extended to the attendance and performance of public servants and to the operation of economic and social markets, it would bring positive livelihood benefits, especially for the poor who are often excluded by processes which are opaque and influenced by elite dominance. ADDRESSING HUMAN CAPITAL CONSTRAINTS The benefits of increased non-agricultural opportunities in rural areas can be realised only if there are the skills to match. However, the institutional delivery of human capital (basic health, education, skills, etc.) is currently facing a demand/supply crunch. School enrolment and retention rates, for instance, have improved, but the quality of State-delivered primary education is not responding and NGO and private provision is expensive.Those who can access high-quality human capital delivery are likely to benefit disproportionately from the new livelihood opportunities in turn feeding increased inequality. TACKLING THE MACRO-GOVERNANCE AGENDA This is a precondition to achieving a quantum change. However, this does not imply that institutional solutions lie at the centre. The centre has a role in resourcing and structuring the macro-framework and in moving from a
State that burdens itself with direct delivery of livelihoodrelated services to a State that sets the boundaries, enables and regulates. Attention to long-term human capital needs; infrastructure development; the regulation of financial services; environmental protection; the management of external relations to improve access to markets; and investment within the global arena should be the agenda of a livelihood-focused government.There is also a role for the centre in setting policy priorities that respond to the needs of the very poor who remain excluded and marginalised from the institutional framework. REFORMING LOCAL GOVERNMENT Without real transfer of political responsibility from the centre and a political emancipation at both centre and periphery, the next round of local government elections will be no more effective. Planning and budgeting need to be delegated at least to district level to achieve any real change. Successful reform will need significant support on a sustained basis for politicians, bureaucrats and staff at all levels. Expectations must be realistic: any decentralisation policy takes a long time to become effective and sustainable. At the local level, the challenge is to strengthen UP as the first tier of political representation.This in turn will depend on changes at the next two tiers: devolution and deconcentration at the upazilla level with an emphasis on improved performance of services and community participation would bring livelihood benefits, especially in relation to human capital delivery; q decentralisation at the district level could facilitate a district that is responsive and focuses on setting the framework for economic and social development rather than being preoccupied with project and programme delivery.
q
More appropriate in the short to medium term would be to decentralise management responsibility from the centre to district and upazilla, with an emphasis on supervision and monitoring of performance and outcome.There are indications that a similar shift in both management culture and responsibility would also benefit larger NGOs. ADDRESSING THE DISCONNECTIONS Local communities remain dislocated: the village is the done to institution and the union the unempowered tier of government. Effective financial power and control and
relevant political power and control lie outside the remit of the local. Opportunity for bottom-up participation and influence are essential for a transformation in livelihood opportunities. Change is happening, but scaling up those changes requires appropriate linkages to the meso- and macro-levels.
Economic and social investment and support for development at the meso-level A shift from a development budget and planning focus to an economic and social development facilitating role for the district administration with the necessary skill and support provided. The promotion of locality/area initiatives that addresses urbanisation, infrastructure and enterprise development (e.g. the Barind Tract Development Corporation). District-wide policy development with a livelihood focus (e.g. transport that supports commuting, market centre development that is district-wide and relates to inter-district linkages). The development of sub-district urban centres for industry and enterprise activity.
TOWARDS A FORMAL INSTITUTIONAL FRAMEWORK THAT SUPPORTS LIVELIHOODS

This paper has taken a broad view of the institutional context of livelihoods in rural Bangladesh. Little has been said about the specific institutional and policy options that relate to livelihood improvements. The focus has instead been on giving an insight into the strategic nature of the institutional change agenda. The institutional agenda below is therefore based on a vision of the ideal institutional form and the steps needed to move towards that vision. A vision of the institutional framework that would best support enhanced livelihood outcomes is: functional diversity and popular accountability at the micro-level, economic and social investment and support for development at the meso-level and a more coherent policy framework from an enabling State that has political ownership and vision at the macro-level. At each of the three levels, institutional interventions that build on current trends include:
q
A more coherent policy framework from an enabling State that has political ownership and vision at the macro-level Attention to the livelihood connectivity of existing human capital policies (for instance, the National Education Policy) as opportunities for promoting change. Support to research and innovation in agriculture and non-agriculture that builds on cross-sectoral synergies rather than support to public-sector service provision. Engagement at the political level to encourage vision and leadership. A focus by government on global policy arenas investment, markets, quality assurance, information with specific livelihood outcomes.
Functional diversity and popular accountability at the micro-level Integrated approaches to local economic development that bring together actors from different sectors. Encouragement of political (UP) and popular (NGO samities and other networks) articulation around common livelihood concerns (e.g. education and health service standards, access to transport and financial services). Skill development for UP and upazilla managers to build more effective partnerships for local-level planning and management supported by effective upward linkages. Direct resourcing of innovative initiatives that have specific livelihood outcomes.
NOTE
1 Madrassah is the name for an Islamic religious school.
REFERENCES
AHMAD, M. (1999): Bottom-Up, NGO Sector in Bangladesh. Community Development Library, Dhaka. DEVINE, J. (1999): One Foot in Each Boat:The Macro Politics and Micro Sociology of NGOs in Bangladesh. Unpublished Ph.D.Thesis, Bath University. GOVERNMENT OF BANGLADESH (1997): Report of the Local Government Commission. Government of Bangladesh, Dhaka. SIDDIQUI, T. (1998): Interaction Between International Financial Institutions and Non Government Organisations in Bangladesh, in A. Karim (ed.) Internal Dynamics and External Linkages. University Press Ltd., Dhaka. THORNTON, P., DEVINE, J., HOUTZAGER, P.P.,WRIGHT, D. AND ROZARIO, S. (2000): Partners in Development:A Review of Big NGOs in Bangladesh Commissioned by DFID. DFID, Dhaka. TINKER, H. (1989): South Asia:A Short History. Macmillan, Basingstoke.
113
Re-thinking local governance: towards a livelihoods focus Hossain Zillur Rahman

WHY LOCAL GOVERNANCE?
How critical is decentralisation? The issue is projected mostly as one of what and how.The question of why tends to be taken as self-evident. Recent empirical evidence, however, is beginning to highlight the importance of re-visiting the why aspect of decentralisation. Bangladesh is a compelling case, where significant reductions in poverty have occurred without commensurate advances in the decentralisation agenda. Despite weak local government, Bangladesh has made remarkable strides in a range of areas including food production, safety net programmes, rural infrastructure, credit provision, primary education, child immunisation, family planning and drinking water provision. All of these advances have translated into major gains in the fight to reduce poverty.While the decentralisation agenda has languished, this has not necessarily resulted in a vacuum at the local level, with a number of important initiatives by NGOs and central government agencies. How useful, then, is it to focus on strengthening local government, when central agencies such as the LGED is developing rural infrastructure and NGOs are so effective in primary schooling and health-care? While it may be neater to substitute existing central and NGO efforts with more effective local government, this is not a compelling enough reason for decentralisation; far from being self-evident, the why of decentralisation clearly merits fresh consideration. The rationale for further efforts in local governance arises most critically from the very nature of poverty trends.We have already noted that Bangladesh has made impressive progress in reducing poverty, gains in which NGOs and central agencies have played the larger role. However, the net rate of poverty reduction in the 1990s appears to be stuck at around one percentage point per year (Rahman, H.Z. 2000).Thus, there is a major challenge of scaling up of the rate of poverty reduction. And it is in the context of the need for scaling up that a new engagement with local governance becomes urgent. The focus is not so much on local governments as project-implementing agencies as it is on local governance as a political and institutional process, capable of enhancing developmental choices locally that are more inclusive of all social groups.
THE PERMITTED PARAMETERS OF CHANGE

AN OVERVIEW OF CHANGES IN LOCAL GOVERNMENT Though more than a century old, local government in Bangladesh has been more a laboratory for policy experimentation than a route to stable institutional development. The Bengal Local Government Act of 1885 provided the first framework, envisaging three levels of local government: union, thana and district. Policy discourse over a century or so has essentially oscillated within this framework. Salient features of the development of local government at these three levels are sketched below: Union This is the lowest tier of local government, currently covering an approximate population of 25,000. The unionlevel body has undergone many changes in nomenclature and is currently called the union parishad (UP). These became fully elected bodies in 1962. Among the various tiers of local government, UPs have the longest institutional history, dating back to 1870. This continuity in institutional life, however, is built on a very narrow functional and financial jurisdiction, and on administrative subservience to executive functionaries at thana and district levels. Notwithstanding these limitations, UPs have been the focal point in the local government system, except for a period in the 1980s when the thana became the focal point. Two important policy changes in the 1990s have placed this body more firmly in the spotlight:
q
streamlining of the representational base of the parishad by subdividing a union into nine electoral units, or wards, instead of the previous three, and admitting a representative from each ward as a UP member; this provision, first suggested in 1993, became operational through the union election of 1997;
strengthening of female representation within the parishads through the 1997 provision for reserving three seats for female representatives.
Thana The lowest tier of administration, which dates back to 1793. It currently covers an approximate population of a quarter of a million. Though suggested as a tier of local government by the Act of 1885, there was no attempt to develop an elected body at this level. Instead, a co-ordinating Council was envisaged, which brought together UP chairmen and central functionaries. The Thana Council was presided over by the chief executive officer at this level, currently known as the Nirbahi Officer.The institutional development of the Thana Council drew its impetus from the Comilla model of the 1960s, which significantly expanded rural development functions at this level. A major change was brought about in the status of the thana-level body, now renamed the upazilla parishad, when provision was made for a directly elected chairman. The basic structure of the previous Thana Council was unaltered but now it was placed under the control of a directly elected chairman. Associated with the change of 1984 was a major increase in the flow of Central Government funds through the instrument of the Annual Upazilla Development Plan mirroring the Annual Development Plans of the central bureaucracy. The provision for a directly elected chairman at the thana level was abolished in 1991. The post-1996 Government has recommended the restoration of direct election but such elections have yet to take place. District Within the institutional history of Bangladesh, the district has been the crucial building block of Central Government.The development of local government at this level was constrained from the outset by the institutional preeminence of executive functionaries. Not surprisingly, development of local government at this level has taken the form of supervised bodies under the control of the chief executive officer, i.e. the Deputy Commissioner, with a narrow functional and financial jurisdiction. Institutional attention to the district-level body has been marked by great ambiguity in policy objectives, reflected in long periods of inactivity. There was an attempt to strengthen the district-level body, now called the zilla parishad, in 1976 by an ordinance that provided for a
certain proportion of elected members. However, before any elections took place, another change, in 1982, abolished the provision for elected members. In 1988, there was an attempt to revive these bodies by nominating ruling party members as Chairmen, but the change of government in 1991 put an end to this initiative.The revival of this body was again mooted in the Local Government Commission reports of both the post-1991 BNP government and the post-1996 Awami League government. However, zilla parishads have essentially existed only on paper since independence.
THE REALPOLITIK OF REFORM DYNAMICS

It is easy to get swept along in the normative discourse on local government in Bangladesh. Idealistic solutions abound.The cycles of hope and frustration, however, are regular enough to warrant a closer look at the underlying sociology of the discourse.This reveals certain constants in the pursuit of reform agendas, which can be seen as the realpolitik of reform dynamics. An exploration of this is crucial both to understand the true reach of reform initiatives and to avoid the trap of ineffectual normative debates. Several features stand out. ACCOUNTABILITY VERSUS DEVELOPMENT ADMINISTRATION The demands of accountability have always been balanced with those of effective development administration. The results of this balancing have not necessarily been conducive to consolidating a process of institutional and political growth of Bangladeshs local government. However, there is a certain durability in the official approach to reform.The overriding principle is to ensure that the centre closely supervises local government and controls local development while allowing just enough space for a local political process.The approach born of this principle ensures that accountability issues are emphasised for the lower tier while development administration issues remain the responsibility of the upper tiers.The corollary is that public discourse invariably over-emphasises issues concerning representation while development issues are left to bureaucrats and experts. More often than not, the goals of accountability and development administration thus come to be pursued in isolation from each other, resulting in an in-built barrier to achieving effective local governance.
Re-thinking local governance: towards a livelihoods focus 115
GENERAL VERSUS SPECIFIC POWERS A second expression of the realpolitik is in the interface of local government and central administration. Powers and responsibilities of the former are invariably defined in general and vague terms, while the powers of the latter are enshrined in specific and precise terms. For example, the general power of the UP is to maintain law and order, but the specific powers of arrest, bail and so forth lie with the magistracy and the police.At ground level, it is always the specific power that has the teeth while the general power cannot graduate beyond declaration of intent.This serves to detract from the idea of local government as a potentially powerful and therefore important agency in the public mind. A QUESTION OF BOUNDARIES: THE LIVED REALITY OF STATE-POWER The administrative structure created by colonial rulers had the primary purpose of maintaining political and social control of the populace. It did this by retaining criminal judicial powers, investing those powers in the magistracy and police and thereby ensuring their supremacy over all functional authority.The legacy of this structure fundamentally distorts the post-colonial political and democratic space. On one hand, the discretionary capacity to punish has become a prized resource for political competitors; on the other, the threat of authorities exerting that capacity to punish hinders any consolidation of political and civil initiatives for a democratic transformation of State powers. The States retention of this capacity to punish constitutes a non-negotiable space that effectively limits reform.Again, this does little to raise the public image of local government as a potentially useful agency. THE WHEAT CULTURE OF LOCAL DEVELOPMENT The introduction of food aid in association with the Basic Democracy programme of the 1960s was a landmark in the evolution of a development culture. Wheat became the new currency of development that linked politicians, administrators and local government functionaries in institutionalised corruption.Though food aid went on to play an important role in addressing rural poverty, the corrupt wheat culture established itself and shaped the development visions and personal aspirations of politicians and local government leaders. Much of the criticism here has focussed on the aspect of corruption but perhaps
the more significant and less understood issue has to do with the wheat culture being an imposed style of political accommodation between aspiring local government leadership and the bureaucratic state. AN ABSENCE OF CHAMPIONS Another feature of the realpolitik of reform dynamics is that, while there is no shortage of general advocates, there is a fundamental absence of champions who can, or care to, drive the local government agenda. This is true both of the political and the bureaucratic power centres, and indeed also of the NGO sector for different reasons. National-level politicians fear losing control of the development process; bureaucrats fear losing control of power, whilst NGOs fear losing access to donor resources. Local government itself has failed to generate any effort in creating or encouraging champions for its further development. THE SPACE ALLOWED We can now summarise what may be seen as the reform space allowed within the reality of the realpolitik analysed above. This is best described in terms of four core processes: local function but central action; development administration, not local governance; q tolerating the informal informally; q jurisdictional rigidity but functional pragmatism.
q q
UNDERSTANDING THE LOCAL SPACE

CONCEPTUAL CONSIDERATIONS A peculiar feature of the State-society interface in Bangladesh is that, although the State is aggressive on matters of jurisdiction, it is far less attentive to societal needs. The former is a colonial legacy, but the latter is a product of tensions in the transition towards a developmental State. Many have preferred to see the latter process as one of State failure.The question here though is how best to understand the governance space? Do such functional areas constitute a different category of governance space(s)? Are these then merely complementary to the State or are they opposed to it or both? Are such spaces created systematically or do they arise spontaneously?
The idea of governance spaces beyond the State carries two important implications for our discussion of local governance. Firstly, it challenges us to examine more deeply the nature of the interface between local government and the State. Secondly, it raises the idea of informal governance beyond that of local government. However, there is a prior issue of how one is able to perceive such spaces. EXPLORING THE SPACE FOR LOCAL GOVERNANCE The governance imperatives One way to explore the governance space is to look at the imperatives by which such spaces emerge. These can be seen to fall into two broad dualities:
q q
Table 1 Peoples perceptions of actors most relevant to specific functions

Function Actor perceived to be most relevant
Dispute resolution
Traditional elders Educated class Local elites Chairman/member (UP) Traditional elders Educated class Youths/clubs Community action NGOs Local elites Government agencies Youths/clubs Community action Political leader/worker Traditional elders Educated class Political leader/worker Chairman/member Local elites Govt agencies NGOs Government agencies Business samities Youths/clubs
Law and order
compensatory versus countervailing; systemic versus sporadic.
Disaster-coping
Much of the governance discussion has focussed on the compensatory aspect of the issue, in which local governments and NGOs are seen as extending the traditional frontiers of State action.As opposed to this, there is often also a countervailing imperative at work as, for example, resistance committees against police oppression. The other duality at work is that of the systemic versus the sporadic.An example of the systemic here is the shalish, an indigenous dispute-resolution process. The counter example of the sporadic is the community processes which come into play during times of disaster. Mapping governance needs The other way to explore the local governance space is via an empirical mapping of governance needs as perceived by the rural population.This serves to establish the functional domains of governance, which may, of course, change over time. A nationwide mapping exercise by Rahman and Islam (2001) reveals six broad governance domains: dispute resolution; law and order; q protection from harassment; q disaster coping; q development; q social welfare.
q q
Protection from harassment
Development
Social welfare
THE QUESTION OF AGENCY Having looked at the functional domains, we can now ask who the governance actors are. Is there a challenge in identifying the appropriate cast of actors most relevant to the discussion on governance and social change? The ascendancy of Bangladeshi NGOs in the field of social development has left little room for other would-be development actors. The discussion on local governance provides a particularly useful entry point to re-open this question of agency.
Rahman and Islam (2001) identified the range of actors people perceived as most relevant to given functions (see Table 1). Though some categories clearly overlap, as for example local government leaders who are also traditional leaders or local elites, what clearly emerges is not merely the diversity types but also clearly identified roles. An important caveat to the above is that some categories may also be associated with negative governance attributes such as corruption and violence.The significant
point emerging from the survey, however, is the need to adopt a multi-agent framework of analysis. LOCAL GOVERNANCE: TOWARDS A MULTI-AGENT FRAMEWORK Table 2 outlines a multi-agent framework for local governance. It is not intended as an operational model that distributes distinct roles to the various actors. Table 2 is more appropriately read as an opportunity map of actual and potential actors relevant to local governance and of potential synergies in their respective roles.
Table 2 Governance and agency
INVESTING IN A NEW STRATEGIC FOCUS

Actor-type Role
NGO
Service-delivery Target-group mobilisation Advocacy Justice Political representation Protection Catalyst Infrastructure Service-provision Demand articulation Protection Power issues Broad-based mobilisation Voluntary work Dispute resolution Power issues Value creation Service-delivery Economic growth Capacity-building Constituency-building Catalyst Agenda formulation Performance evaluation
Local government
Government agencies
THE CHANGING IMPERATIVES OF LIVELIHOODS It is essential to frame an examination of the potential of local governance within the context of the realties of modern rural livelihoods. Macro-perceptions of slow progress in poverty reduction masks rather far-reaching changes in the livelihood systems of the poor in Bangladesh. Reading the record right is thus a key first step towards developing a new strategic focus. While the nature and directions of changes have been diverse, the reality of a qualitatively different livelihood context is captured in three key processes as described below: Declining centrality of land Land is no longer the principal basis of power and status; neither does it serve to limit the livelihood opportunities for the rural population.According to a most recent estimate, agriculture now contributes less than 50 percent of household income (Hossain 2001). An emerging rural-urban continuum The expansion of all-weather roads has helped to dissolve the boundaries between rural and urban Bangladesh. Migration and remittances have emerged as dominant factors in household dynamics; the recent study cited earlier shows the share of remittance in household incomes rising from 3.7 percent in 1987-88 to 18.5 percent in 2000. Migration of varying durations to varying destinations is increasingly important in the rural economy. Labour market transformations Labour market transformations have been the third key process impinging on the livelihood possibilities of the
Political parties
Community capacities
Private sector
Support institutions
Think tanks
poor.There is a trend towards an occupational hierarchy for the poor, in which casual daily labour is the least preferred employment. The competition is for piece-rate labour contracts and fixed-rent tenancies in the farm sector and for non-farm employment in rural construction activities, transport operations and in the lower end of trade and service activities.Within the context of such an occupational hierarchy, the capacity to shift becomes the key livelihood concern for the poor. Correspondingly, there is a great emphasis on the finance access, social networking and human capital factors which generate this capacity at the level of the individual household. LOCAL GOVERNANCE FOR LOCAL ECONOMIC REGENERATION These two strands, one charting the changing imperatives of livelihoods, and the other the political parameters of the decentralisation agenda, suggests a new strategic focus: a livelihood-driven local governance agenda. This paper began by citing Bangladeshs progress on the poverty front. Notwithstanding these achievements, poverty reduction is very slow (Rahman, H.Z. 2000). Such a sobering realisation has two critical implications for any new strategy. Firstly, todays poverty challenge is one of scale. Secondly, a scaled-up attack on poverty cannot emerge simply from a more of the same approach to strategy. While we must continue to build on current achievements, the search for new entry points for a scaled-up attack on poverty clearly assumes strategic urgency. It is precisely in addressing this challenge that the livelihood and local governance analyses undertaken earlier come together to produce a new strategic focus, one in which the regeneration of the local economy becomes the new action goal and local governance the vehicle through which such a goal is best addressed. Let us elaborate. Livelihoods analysis suggests that a local economy perspective is a relevant reference frame in which strategies for a scaled-up attack on poverty may be forged. Such a perspective is to be distinguished from prevalent sectoral or micro-household perspectives. A local economy perspective is also not to be confused with village economies; it is more accurately a meso-economy perspective. What imparts the novelty to such a perspective is the changing ground realities, such as the growing
rural-urban continuum, which now define the local. In an earlier era, the programmatic understanding of such a focus would have been captured by terms such as integrated rural development or local-level planning. The current emphasis on a local economy perspective, however, marks a radical departure on such earlier meanings. The concern is less about a sectoral focus and more about positing a new action goal, namely that of regenerating the local economy, as this is the best way of empowering the poor in their pursuit of multiple livelihoods. Insights from the earlier analysis on local governance become of critical relevance when seeking an appropriate vehicle for regenerating local economies. The cast of actors who matter here extend well beyond the traditional focus on local governments as they also cut across traditional sectoral or rural-urban boundaries. The challenge really is of multiple livelihoods, of linkages and a critical expansion of local opportunity frontiers, and of bringing within mainstream attention any categories of missing poor. Such a menu of tasks does not fit easily within traditional sectoral or decentralisation approaches. What is required is a governance focus with a twist, namely a primary orientation to livelihood issues and embracing the possibility of enlisting categories of actors beyond local governments per se. Interestingly, it is precisely in adopting such a livelihood-oriented governance focus that the real possibilities of a decentralisation agenda currently lie.A strategy of up-front struggle over jurisdictions and power for local governments appears to be a non-starter in the current balance of political and administrative power. The politically intelligent approach is to galvanise the local governance potential through investing it with a new functional focus around the goal of regenerating the local economy. THINKING COALITIONS Pursuing a livelihood-oriented local governance focus also requires a new approach to institutional strategy. From the State-centric model of yesterday and the NGOcentric model of today, we are moving towards a multiagent framework. A key mind-set change that this necessitates is towards a coalition approach to institutional strategy. Indeed,thinking coalitions can be seen to be the first step towards a new understanding of institutional strategy.Within the prevailing institutional culture, it is a step requiring an active process of propagation.
We do not mean to imply that the idea of coalitions is absent in the prevailing institutional milieu. However, what tends to go by the idea of coalitions currently is a coming together of the same type of actors, i.e. coalitions among NGOs, etc. What is suggested here are coalitions across boundaries, i.e. local governments and NGOs, local governments and private sector, local government, NGOs, private sector and support institutions, etc. It is also important to emphasise that we are not talking here about an inflexible notion of coalition built around a fixed focal point.Though in many cases, local governments, i.e. UPs and pourashavas, are likely to be the focal points, the general idea is about context-specific coalitions built around specific goals.
REFERENCES
HOSSAIN, M. ET AL. (2001): Changes in Agriculture and Economy in Bangladesh, 1988-2000: Insights from a Repeat Survey of 16 Villages. International Rice Research Institute. RAHMAN, H. AND ISLAM,A. (2001): Informal Governance and Democratic Graduation: Re-thinking Agency and Performance. Power and Participation Research Centre, Dhaka. RAHMAN, H. (2000): Poverty: the Challenges of Graduation, Bangladesh Institute of Development Studies. Dhaka. SIDDIQUI, K. (2000): Local Governance in Bangladesh. University Press Ltd., Dhaka.
121
Gender dimensions of rural change Santi Rozario

Gender roles are strongly differentiated in Bangladeshi society. This paper examines how changes in Bangladeshi society have impacted on womens livelihoods. security for poorer members of the community. The local resources once available to the poorest, such as fuel wood, common grazing lands and fishing grounds are dwindling. Landlessness is increasing.The majority of village households now have no farming land at all and alternative employment in the village remains limited. Extended family networks, traditionally an important safety net for the poorest families, are gradually being eroded. Economic pressures, sometimes caused by natural processes such as river erosion, are driving migration to urban areas. Whilst many changes have made life harder for poor people, some have contributed to considerably changing the traditional picture of a rural womans life in modern Bangladesh. For example: women play a more visible role in the cash economy as a result of the monetisation of the village economy; q villagers are more dependent on credit for all aspects of subsistence; q women are more mobile; they have a greater presence in public spaces and in formal paid labour; q more and more women are migrating to urban areas for work, e.g. in garment factories.
q
CHANGES IN THE RURAL CONTEXT

Bangladeshi villages can no longer be seen as isolated from urban centres, nor indeed from the rest of the world. Roads and transport infrastructure today offer greater accessibility to urban centres and hence to new and emerging livelihood opportunities. Whilst these opportunities are often more accessible to men (given the prevalent bias in favour of men), some governmental and non-governmental interventions have redressed some of that imbalance. For example, an expanded education system has resulted in increased school attendance among females and the poor, and increased their access to secondary and college education.The growing presence of NGOs and MFIs in Bangladeshi villages has helped to increase womens access to finance, generated income-earning opportunities and in some cases led to enhanced social mobilisation. Other forms of aid (health, legal aid, etc.) are also channelled to those who are recipients of micro-finance. Also working in favour of womens empowerment is the pressure (in practice, much of it nominal) mediated through NGOs and micro-finance organisations to modify aspects of gender relations (e.g. payment of dowry). Finally, the intermittent presence of social safety-net provisions such as the VGD Scheme, involving work for food or cash is, to some degree, available to women. Such trends have created a new mix of winners and losers, have led to changes in peoples livelihood aspirations and fears and impacted on the traditional institutions and coping mechanisms that have underpinned rural life. Whilst middle-class expectations are rising, as a result of increased penetration of global and urban culture at the village level (via mass media and improved communication with urban centres), patron-client relationships within village communities are disintegrating, with a consequent loss of
This is a complex range of developments but almost all of them can be seen as aspects of a single overriding process; the progressive incorporation of Bangladeshi villages into the national and global economy and the consequent restructuring of social relationships and livelihood options. Given the heavy dependence of the Bangladeshi government on foreign aid, Bangladesh has been particularly open to this process of incorporation.The globalising process has both positive and negative aspects for village communities, and specifically for their women. POSITIVE ASPECTS OF RECENT TRENDS Better links to the global economy have brought new livelihood possibilities for village women, such as jute work and handicrafts manufacture for overseas markets, work in urban garment factories or labour migration to
the Middle East. The global capitalist system pays little attention to local traditional gender values, and the endemic poverty of Bangladeshi villages provides a strong incentive to accept whatever employment possibilities are available. Globalisation is thus viewed in different ways either as a process which enables or one which forces women to move out of previously restricting, genderdefined roles. In addition, the international development industry has focussed increasingly on women since the mid-1970s, and improving the position of women is an explicit goal of most development projects and NGO activities.The targetting of women for health care, educational and micro-credit facilities forms part of this process. NEGATIVE ASPECTS OF RECENT TRENDS Increased monetisation of the village economy has led to the progressive disappearance of subsistence agriculture and to increasing levels of landlessness, unemployment and under-employment. In general, people now live healthier and longer lives, which puts greater pressure on few and diminishing resources. The environment is deteriorating and traditional social support mechanisms are breaking down.All of these processes hit hardest at the poorest women. Changes to village society brought by processes of globalisation are not altogether welcome. Domestic violence against women, denunciations of village women by Islamic authorities and village courts (shalish), opposition to, and conflict over, womens participation in NGO activities and paid work outside the home, and active resistance to the role of female UP members all show that many Bangladeshi village men find it hard to accept the new order. Whilst there has been no organised Islamic movement in Bangladesh against the imposition of Western materialism, women who transgress their traditional roles are a convenient scapegoat for traditional village men. That many of these men have seen their own livelihoods threatened and are themselves coping with increasing poverty helps to explain their actions but is of little consolation to their victims (Jahan and Islam 1997; Blanchet 2000;ASK 1998, 1999a, 2000).
globalisation, have been continually mediated and structured by the persisting ongoing norms and ideologies of Bangladeshi gender relations. Despite some minor variations between different religious communities, the gender ideologies throughout Bangladeshi society are based on the institution of parda1 and the associated values of purity, honour and shame. These values, and the consequent devaluation of women in relation to men, can only be understood as part and parcel of the structure of patrilineal descent and patrilocal residence2. Whilst parda is commonly seen as an Islamic institution, its practice by women of different religious groups in Bangladesh suggests that it is a fundamental element of Bengali culture. Control over the spatial mobility of women is maintained through the ideologies of purity, honour (izzat) and shame (lazza). Womens movement beyond their appropriate boundary is regarded as entering the mens sphere and as endangering their purity and honour, and thus in turn the honour of their families and community.The boundary refers to both space and time, so that women should not generally be seen outside after dark. Its position is continuously negotiated, and varies for rural and urban, middle-class and lower-class women, but everyone is aware of its existence and women ignore it at their peril. Women are reluctant to pursue livelihood activities that contravene this boundary. Whilst codes of honour and shame refer to the behaviour of both men and women, in practice, honour is seen more as mens responsibility and shame as womens. Mens honour is primarily determined through their role as protectors, whereas shame acts as a restraint on female behaviour. Like women, however, young men are subject to restrictions on their behaviour, and are expected to display sexual propriety. Like women, they are often subjected to arranged marriages against their wishes for financial reasons or to form alliances with other reputable and wealthy families.Although it is theoretically easier for men than women to contravene some of these values and pressures, if they want to maintain their reputation within a village they often have to put up with an unhappy marital union.The consequence of unhappy marriage is often domestic violence and extra-marital affairs, usually on the part of men.
GENDER RELATIONS IN BANGLADESHI SOCIETY TODAY

The ability of women to take advantage of the new opportunities, and their vulnerability to the downside of
Gender dimensions of rural change 123
From childhood, boys are socialised in preparation for their future roles as patriarchs, as providers, and protectors of the purity and honour of women of their families. Girls, on the other hand, are socialised to be caring, selfsacrificing, submissive, and good wives, daughters-in-law and mothers.The significance of marriage for finding an identity, a place of belonging, is implanted in a girls psyche from an early age. Once married, however, young adolescent girls can find themselves in a vulnerable position, socially and economically dependent on their husbands and in-laws. Their parents (i.e. the wife-givers) also occupy a subordinate position in relation to the grooms family. They remain anxious not to antagonise the brides husband or his family. A mature adult womans identity is defined ultimately through her relationship to her husband and children, as a wife and a mother. There is no other acceptable identity for adult Bangladeshi women, particularly in the rural context. Whilst marriage is also important for Bangladeshi men, their identity is not as closely linked to this. Men often do not marry until the age of 30 or beyond. But a woman must be married early, for a mature woman is perceived to be a liability, both in economic and cultural terms. Unmarried women are seen to be dependent on their male guardians who have to feed and guard them lest they misbehave sexually, thereby spoiling their purity and ruining the honour of the whole family. It can be seen that ideally a womans life should be confined to the home and to domestic work. Work outside the household, particularly if it involves moving around or beyond the village on ones own, is undesirable and places the womans honour at risk. In fact, working outside ones home is by no means an entirely new development for Bangladeshi village women, but it has almost always been taken on reluctantly and in response to economic need, and given up as soon as the family can afford to do so. Rural families preoccupation with ideas of honour and shame makes more sense if we appreciate the symbolic capital attached to a familys reputation.The maintenance of honour can ultimately help families to gain economic benefit.Thus a good reputation may be essential for family members in gaining office in local government, village
leadership positions, obtaining a bank loan or contracting marital alliances. However, because poor families cannot afford to maintain their honour by confining their women at home, or by displaying wealth in the form of elaborate feasts and paying large dowries at weddings, purity and honour work essentially to the advantage of the rich. Bengali notions of honour, shame and purity are important not only to maintain a gender hierarchy but to maintain group boundaries, class and status structures, and the distribution of power within rural society. Consequently, it is also in the interest of women of the wealthier sections to perpetuate and maintain the existing system.
CHANGES IN WOMENS LIVELIHOODS

This section looks at the new livelihood activities and economic opportunities that have become available to women in modern day Bangladesh and considers how womens access to these livelihoods has been structured by gender values. WOMENS ENTRY INTO NEW LIVELIHOOD ACTIVITIES There is no doubt that women have begun to penetrate livelihood activities previously dominated or monopolised by men. One new area of female employment within the village is provided by work with NGOs and MFIs, as health and family planning visitors, family welfare visitors, teachers, workers on income-generating projects, etc. These jobs are rarely permanent and some of them go to women from urban backgrounds or from outside the local community, but many are taken by village women. Values, however, continue to structure womens workroles and to affect their willingness or ability to take them on.Women from good (wealthy) families are particularly reluctant to take up NGO work, or are prevented from so doing by their male guardians. Other new livelihood possibilities are effectively closed to most women.The expansion of the service sector has led to many more small shops, tailors and other craft workers, rickshaw pullers, petty traders, etc., in villages and local bazaars. Such work, however, is perceived as mens work, since women cannot maintain their respectable status while working in a local market, though some elderly women may run small shops which form part of their
own houses. Migratory labour overseas is also primarily a male reserve, especially since the Bangladeshi government restricted overseas migration of female nurses and domestic workers in 1998 as a result of fears of sexual violence against migrant women (ASK 1999a). Although women in most parts of Bangladesh do not work directly in the fields, women in some areas have begun to do various kinds of field work, both as agricultural wage labourers and on family land or sharecropped land.This is seen as a temporary expedient, to be ended as soon as the family can afford to (Jordans and Zwarteveen 1997). Garment and other factories mostly urban but some of them in rural locations (e.g. bidi factories) (Blanchet, Razzaque and Biswas 1998) have provided many new employment opportunities for women. Both men and women have migrated in large numbers to cities for employment. Improved communications between rural and urban areas allows many people to travel daily to nearby urban centres. However, while migrant garment workers, especially single women, make an important contribution to the village economy, they are again seen as compromised in terms of female propriety (Siddiqui 1991; Kabeer 1994; Zohir and Paul-Majumder 1996; Rozario 1998). Shrimp farming and processing provides additional employment in some coastal areas, including collecting fry and working in shrimp hatcheries. Shrimp processing is an option both for men and women (Pokrant and Reeves 2000). However, this work is seasonal and insufficient for workers to rely on exclusively. Further outside work is provided by official safety net programmes for mitigation of floods and other natural disasters. There has been little research on the impact of these programmes. Comments made in relation to other new modes of outside employment would, however, seem to apply even more strongly to much safety net work (e.g. earth-cutting), which is not only carried out outside the home but is of low status and an indication of the familys destitution. Overall, a regular pattern emerges in these new livelihood activities: such jobs are taken on reluctantly and never by
families who can afford not to do so; they involve a loss of status and a threat to the women and the familys honour; and they are given up as soon as possible. THE IMPACT OF NGOs AND MICRO-CREDIT The most significant single development in the 1990s in relation to womens livelihoods has undoubtedly been the growth of micro-credit schemes aimed at women. Although advocates for micro-credit place great emphasis on lending to women, it is common knowledge among borrowers and bank workers alike that, whilst loans are taken by women, they are mostly used by their menfolk, husbands, brothers, sons, fathers or fathers-in-law. In recent years, this has been explicitly acknowledged through the practice of requiring the signature of the husband or son who will use the loan before the loan is approved (Rozario 1999; 2000).Women borrowers often take no part in deciding how the money is to be spent. Despite these limitations, NGO/MFI micro-credit has provided many benefits for village women.They use loans to purchase: consumer goods, including food, radio-cassettes, televisions, etc.; q medical care; q materials and labour for building tube wells, toilets and house improvements; q education for sons and daughters; q dowries for their daughters; q supplies for their personal income-generating activities, including poultry or animals, materials for jute work or other handicrafts; q materials for income generation activities for their menfolk, including purchases of land, animals, rickshaws or boats, or setting up tea-stalls or other small trading enterprises; q capital to loan to other villagers at a profit.
q
The purchase of consumption items reflects, in part, the increasing monetisation of village life and the influence of urban consumption patterns.While these may not fit well into the NGO/MFI stated emphasis on income generation, they have a positive effect on womens sense of selfworth and on their status within the family.
However, the form of micro-credit provision developed by Grameen Bank and adopted by most other providers in Bangladesh can also present specific problems for village women, especially those from poorer households. A major issue is the difficulty of meeting repayments, exacerbated by rigorous enforcement and lack of flexibility on the part of micro-credit providers. This often leads to women and their families eating poor food and making other compromises in order to meet the repayments; to domestic conflict and violence when women pressure husbands or sons for repayments; and to fear of joining the schemes by the very poor because of the risk if they are unable to meet payments. In practice, those poorer women who join the schemes often remain only until they have reached specific goals, but this means that they are then excluded from other benefits channelled to scheme members (e.g. health assistance or legal assistance). The system of shared responsibility for loans among a group of women often provokes conflict, since none will be able to take out further loans if one member defaults. It also leads to an unwillingness to accept women who are seen as poor liabilities for repayment, since all the women will be held responsible. Other problems include the high interest rates, often not fully understood by the villagers, and the pressure from micro-credit providers on villagers to take large loans when smaller loans or other forms of assistance might be more appropriate. Overall, Grameen Bank-style micro-credit provides useful resources for middle to poor rural families, but fails altogether to reach the poorest of the poor, who are regarded as bad risks for lending. NGOs and MFIs are effectively oriented towards expanding their membership and increasing the amount of money they are lending to reliable borrowers, not to ensuring that the poorest and most vulnerable sectors of the village population receive assistance.As a consequence, each organisation builds up a clientele of middle to poor villagers, and directs resources towards them. Despite the positive outcomes for many villagers, the approach is oriented towards the individual entrepreneur rather than collective security, and is often divisive and destructive towards the village community as a whole. Whilst this model suits some parts of the rural Bangladeshi population quite well, Rozario (2000) has argued elsewhere for the consideration of alternative, more cooperative models of micro-credit provision.
IMPACT OF CHANGES ON WOMEN AND ON GENDER RELATIONS

We have mentioned how the way in which Bangladeshi women have responded (or been allowed by their menfolk to respond) to new livelihood opportunities has, throughout, been shaped by the existing structure of gender values.This is not to say, however, that womens access to these new kinds of employment or to micro-credit has no potential to undermine or transform these established gender values. This section discusses the ways in which gender values are beginning to shift, looking at issues of empowerment, dowry and the UP elections. WOMENS EMPOWERMENT The whole idea of empowerment is a difficult and contested one (Rozario 1997a). In this paper the focus is on womens ability to exert influence within and beyond the household. We therefore need to ask in particular whether either womens increased presence in outside paid employment, or womens access to micro-credit funds, has increased such influence. In regard to outside work, it has long been common in rural Bangladesh for poor women to take up paid employment outside the home. This has usually been in highly exploitative situations, e.g. domestic work for wealthier households. Thus earth-cutting work on food for work-style safety net programmes, migrating to the cities to work in tobacco and garment factories and the like are widely experienced as a continuation of this pattern, rather than as gateways to new and more liberated lifestyles. In practice, women who migrate to the cities for factory work, or who take on other kinds of work involving substantial mobility outside the household, are often obliged to take on more responsibility for decisions in their lives than they would have exerted back in the village. To regard this positively as empowerment does not, however, acknowledge how women themselves feel about their situation as part of a new low-paid, vulnerable and unprivileged female employment sector (Rozario 1997b). Their families may appreciate the resources but are almost always concerned about the threat to the womens reputation.Wage-earning women threaten the idealised role of the male as breadwinner, and the danger of male violence, both from unrelated men and from their own menfolk,
can be quite real. At the same time, their contribution to the finances of the family is significant, and inevitably leads to a greater input into household decision-making. Opinions differ on whether micro-credit schemes have led to the empowerment of rural women. Initially, strong claims were made for positive effects of participating women, but researchers have recently argued that the schemes have not empowered women, at least in relation to men. The funds accessed, even if withdrawn by women, are mainly controlled and used by men, and there has been little real change in the structure of gender relations. Thus Montgomery et al. (1996) have argued that,those women who do particularly well as a result of credit are much more likely to be empowered vis--vis other (less well-off) women, rather than vis--vis the menfolk in their household (or in wider society). An intermediate position may be more appropriate. We need to recognise that village women use micro-credit for purposes they themselves value, and in terms of the culturally determined logic of their own situations. In directing micro-credit funds to husbands and sons, they may be making a realistic decision as to the best way of ensuring both their own security and that of their family as a whole. In particular, they may have little reason to use microcredit to contest norms such as parda or izzat (for example, by undertaking income-generating projects involving outside work) unless their respectability has already been seriously compromised. They may on the other hand use micro-credit to withdraw from work outside the home, which they see as dishonourable and humiliating, so as to conform better to gender norms (Kabeer 1998). Such women are nevertheless exercising choice, improving their livelihood options, and raising their status within the local community. Indeed, womens new ability to contribute significant economic resources to the marriage can lead to an increased sense of selfworth and to a greater effective voice in household decision-making processes (Kabeer 1998). Micro-credit as currently delivered in Bangladeshi villages has positive aspects for many women, but it does not necessarily build female solidarity.What micro-credit can do is to strengthen the position of individual women who
have the ability and personal qualities to take advantage of it. Such women can certainly gain influence in relation to other women in the community, through raising their familys overall status. They can also reduce their dependence on patron-client relations with rich villagers and provide for a more secure livelihood for themselves and their children. If the aim is the collective empowerment of women and the creation of village solidarity, then more traditional social mobilisation programmes and forms of microcredit provision other than that pioneered by Grameen Bank will need to be pursued. DOWRY Dowry has long formed a part of Hindu marriage whereas its practice among Muslim and Christian communities in Bangladesh is relatively recent. It has become more common since the early 1970s and dowry amounts have inflated in recent years. The growth of dowry has to be understood from both economic and ideological perspectives. The increasing prevalence of wage labour and the general monetisation of the Bangladeshi village economy have fundamentally transformed the economic basis of marriage. While the values of purity, honour and shame continue to mandate early marriage for women, men marry at a considerably later age, leading to a disparity between the available numbers of men and women.At the same time, the availability of money as a result of labour migration by both men and women and through micro-credit facilities has enabled families with eligible grooms to demand higher and higher sums. Whilst dowry payments are technically illegal in Bangladesh, and have been at least nominally opposed both by the Grameen Bank and most NGOs, the continuing importance of marriage in defining Bangladeshi womens identity is evident in their maintenance and continuing inflation (Rozario 1992). By and large, dowry remains unchallenged and is practised widely by all classes in rural and urban areas. There is a clear sense of helplessness on the part of women when questioned as to why they continue to pay dowry: What can we do? If we do not pay dowry, we cannot marry off our daughters, or Who will take my daughter without a dowry? (Naved 1994).
UP POSITIONS Every UP now has to have at least three women members out of a total of 13, and nearly 13,000 women were elected to UP positions in December 1997 and January 1998, when women first stood for election. The NGOs, who were largely responsible for bringing about this change, have also attempted through voter education to encourage women from relatively vulnerable groups to compete for, and win, positions of authority. Consequently, these positions are frequently occupied by lowermiddle-class women with links to locally operating NGOs. The presence of women in such positions certainly has the potential to enhance these womens voices and, in turn, the voice of village women as a whole. It also provides access to resources for the women members and their families.There is some evidence of female UP members acting assertively to assist their constituents. However, they also face considerable resistance to their using their positions constructively. Their role is not clearly defined in the existing legislation, providing scope for the UP chairman and other male members to restrict their activities. In practice, they are often assigned to look after welfare-oriented works like education and health care, and kept away from major development projects (ASK 2000; Huq and Mohiuddin 2000). In addition, there are numerous stories of elected female members being subjected to sexual harassment.
further. At the same time, we can expect that the presence of women in outside work, and their role as significant providers of cash income both via work and via access to micro-credit, will gradually change village gender values.The basic form of gender relations is unlikely to change rapidly, but the boundaries of acceptable female behaviour are likely to become significantly extended. The impact of the female UP members is harder to predict. Some scholars have suggested that male village elites are likely to begin putting forward their own women to take over these positions and act in elite interests (personal communication: Hussain and Mohiuddin 2001; Huq and Mohiuddin 2000). Whilst this seems a likely enough scenario in many communities, there is already a substantial number of non-elite women with experience as UP members, and attempts at upper-class takeovers may meet significant resistance. If enough female UP members, from whatever class background, manage to overcome male opposition and exercise significant power, this alone may lead to real changes in the perception of women in Bangladeshi society. Resistance from elite men, but particularly from non-elite men, to womens new roles can be expected to continue. The high levels of violence against women, both in public and domestic contexts, appear to be linked to male frustration at their own powerlessness in relation to wider economic forces, with womens new assertiveness as a convenient scapegoat (ASK 2000; Blanchet 2000).There is obvious potential for such resentments to be mobilised for political gain. So far, developments in relation to the provision of environmental protection, quality education, adequate health and contraception and legal protection have been limited. If there is the political will to achieve real changes in these areas, they could lead to significant improvements in womens well-being and in the structure of gender relations.
FUTURE TRENDS
What can we expect the next few years to hold in relation to womens livelihoods in Bangladesh? The monetisation of the village economy will continue, with food provision brought increasingly into the realm of the cash economy. Wealthier villagers and urban entrepreneurs will benefit from these developments, while poorer sectors of the community face increasing impoverishment as landlessness increases and the remaining common land and other shared village resources become unavailable (Westergaard 2000). Women from poorer families will face further pressure to enter into outside work, either locally or through migration to urban centres, and dowries are likely to rise
CONCLUSIONS AND RECOMMENDATIONS

Despite the increased participation of women in paid employment, ingrained gender values are still dominant in Bangladeshi rural society. They contribute to the
perpetuation of dowry, to the overall acceptance of mens authority within the household, to womens perceived dependency on men for their income-generating projects and to the inhibition of womens mobility and actions. All these make it difficult for women to aspire for positions of authority or to effectively exercise the authority that should accompany such positions as they may obtain. Without questioning the genuine interest on the part of NGOs and MFIs in helping poor women to improve their economic position, it is evident that most of these organisations do not perceive the challenging of local oppressive values or local power structure as important. The assumption is that changes will come about gradually once poor women are economically empowered. In fact, it is unlikely that changes in gender relations will come about merely by focussing on economics and poverty. The continuing problems of dowry and violence, at a time when overwhelming numbers of Bangladeshi women are already playing a significant economic role within the family, suggests that there is a need to address gender separately. In particular, it is difficult to see how the oppressive aspects of gender relations among poor people can be improved without also tackling middle-class gender values. It is these middle-class values that poorer people aspire to, and the same values that constantly judge them and find them wanting. If middle-class Bangladeshis themselves were to contest those values more openly, for example by challenging the definition of public space and of most forms of paid employment as male territory, then poorer rural and urban women might begin to see their own participation in the public sphere in less negative terms. RECOMMENDATIONS q Address gender values directly through education on gender issues for both boys and girls starting from primary school level. q Introduce gender sensitisation programmes for mens samities within the village and promote mediation services for domestic and dowry-related disputes. q Take effective action at all levels to counter violence against women and other forms of gender oppression, including the systematic implementation of existing
legislation and the development of new legislation. As far as possible, enlist Islamic and other religious authorities to support such action. q Encourage alternative models of micro-credit to the Grameen Bank approach (less top-down, involving more genuine participation by and social mobilisation of village women). q Emphasise technical training and provision of skills to enable women to take full advantage of micro-credit provision and new employment opportunities. Although NGO and other schemes have provided some opportunities for women to acquire new skills, much more could be done in this area. q Where possible, NGOs should recruit and train village women for NGO positions. q Provide effective safety net measures to maintain livelihoods for the poorest of the poor, and undertake further investigation of how these programmes work in practice, and of their differential impact on women. q Give further emphasis to health care and education in rural areas, with explicit attention to distributional issues between rich and poor, city and country, men and women; investigate ways of addressing these issues effectively (e.g. through universal health insurance).
NOTES
1 Parda, literally meaning curtain, refers to the practice of female seclusion. In the strictest sense parda involves keeping women confined within the home and covering them in veils whenever they venture out. In a wider context, parda refers to womens modesty and restrictions on their interactions with males. In Bangladesh only a small number of women follow strict parda, but from puberty onwards all women are required to observe parda in the wider sense. 2 In other words, descent is traced through the father, and women move in with their husbands kin after marriage, where they are vulnerable outsiders to the local descent group.
REFERENCES
ASK (AIN O SALISH KENDRA) (1998): Human Rights in Bangladesh 1997. University Press Ltd., Dhaka. ASK (AIN O SALISH KENDRA) (1999a): Human Rights in Bangladesh 1998. University Press Ltd., Dhaka. ASK (AIN O SALISH KENDRA) (1999b): Annual Report 1999.Ain O Salish Kendra. ASK (AIN O SALISH KENDRA) (2000): Human Rights in Bangladesh 1999. University Press Ltd., Dhaka. BLANCHET,T. (2000): Constructions of Masculinities and Violence Against Women, Study presented to CARE. Bangladesh, Dhaka. HUQ, M. AND MOHIUDDIN, K.M. (2000): Union Parishadey Nari: Paribartanshil Dhara. Bangladesh Nari Pragati Shangha. Dhaka. JAHAN, R. AND ISLAM, M. (1997): Violence Against Women in Bangladesh:Analysis and Action.Women for Women and South Asian Association For Womens Studies. Dhaka. JORDANS, E. AND ZVARTEVEEN, M. (1997): A Well of Ones Own: Gender Analysis of an Irrigation Program in Bangladesh. International Irrigation Management Institute, Colombo, Sri Lanka; Grameen Krishi Foundation, Rangpur, Bangladesh. (IIMI Country Paper, Bangladesh No. 1) KABEER, N. (1994): Womens Labour in the Bangladesh Garment Industry: Choice and Constraints, in C. Fawzi El-Solh and J. Mabro (eds) Muslim Womens Choices: Religious Belief and Social Reality, pp.164-183. Berg, Oxford. KABEER, N. (1998): Money Cant Buy Me Love?: ReEvaluating Gender, Credit and Empowerment in Rural Bangladesh. IDS Discussion Paper 363. Institute for Development Studies, Brighton. MONTGOMERY, R., BHATTACHARYA, D. AND HULME, D. (1996): Credit for the Poor in Bangladesh: The BRAC Rural Development Programme and the Government Thana Resource Development and Employment Programme, in D. Hulme and P. Mosley (eds) Finance Against Poverty,Vol. 2, pp.94-176. Routledge, London and New York. NAVED, R.T. (1994): Empowerment of Women: Listening to the Voices of Women. Bangladesh Development Studies 22: 155-178. POKRANT, B. AND REEVES, P. (2000): The Imperatives of Globalisation: the Development of Export-Oriented Shrimp Production in Southeast Bangladesh. Paper for Asian Studies Association of Australia conference, Melbourne, July 2000. ROZARIO, S. (1992): Purity and Communal Boundaries: Women and Social Change in a Bangladeshi Village.Allen and Unwin, Sydney.Also Zed Books, London and New York. ROZARIO, S. (1997a): Development and Rural Women in South Asia: the Limits of Empowerment and Conscientization. Bulletin of Concerned Asian Scholars 29(4): 45-53. ROZARIO, S. (1997b): Disasters and Bangladeshi Women, in R. Lentin (ed) Gender and Catastrophe, pp.255-269. Zed Books, London. ROZARIO, S. (1998a): Disjunctions and Continuities: Dowry and the Position of Single Women in Bangladesh, in C. Risseeuw and K. Ganesh (eds) Negotiation and Social Space: A Gendered Analysis of Changing Kin and Security Networks in South Asia and Sub-Saharan Africa, pp.259-275. Sage Publications, New Delhi. ROZARIO, S. (1998b):The Dai and the Doctor: Discourses on Womens Reproductive Health in Rural Bangladesh, in K. Ram and M. Jolly (eds) Modernities and Maternities: Colonial and Postcolonial Experiences in Asia and the Pacific, pp.144-176. Cambridge University Press, Cambridge. ROZARIO, S. (2000:) Women and Poverty: Is Grameen Bank the Answer? E-copy at http://www.uow.edu.au/ research/centres/capstrans/pubs/publications.html (CAPSTRANS Working Paper). WESTERGAARD, K. AND HOSSAIN, A. (2000): Boringram Revisited: How To Live Better On Less Land, in R. Jahan (ed) Bangladesh: Promise and Performance, pp.309-337. University Press Ltd., Dhaka. ZOHIR, S.C. AND PAUL-MAJUMDER, P. (1996): Garment Workers in Bangladesh: Economic, Social and Health Condition. BIDS, Dhaka. (Research Monograph, 18.)
131
The causes of vulnerability in rural livelihoods S.Aminul Islam

Livelihoods are vulnerable when they are unable to cope with and respond to risks, stresses and shocks.The World Development Report of 2001 states that vulnerability measures the likelihood that a shock will result in a decline in well-being. It is primarily a function of a households asset endowment and insurance mechanisms and of the characteristics (severity, frequency) of the shock (World Bank 2001:139).Vulnerability is therefore a function of: the nature of the shock or stress; q the temporal dimensions of the shock or stress, the speed of onset and whether it is sudden or frequent; q its spatial dimensions whether the shock or stress is widespread or localised; q the degree of preparedness within the household, community or state; q a households asset endowment; q a households insurance mechanisms;
q q
a households coping mechanisms.
RISK AND VULNERABILITY IN RURAL BANGLADESH

Table 1 highlights the variety of ongoing stresses that form a backdrop to rural livelihoods. It distinguishes between covariate or common stresses, which operate at the mesoand macro-levels and affect all sections of the community (economic stresses and natural disasters, for example) and individual or idiosyncratic shocks and stresses, which affect particular households (such as death of a family member).
NATURAL DISASTERS AND STRESSES

Among developing countries, Bangladesh suffers from more than its fair share of natural disasters and these have a significant influence on rural livelihoods.
Table 1 Sources of vulnerability

Micro Meso Macro
Natural/environmental
Salinity Aridity Arsenic contamination Pest attack
Waterlogging River erosion Deforestation Cyclone Epidemic Land degradation
Climate change Sea-level rise Flood Drought
Social
Illness Injury Disability Old age Death Crime Domestic violence Unemployment Resettlement Harvest failure Mastanism1
Economic
Political
Political violence
Governance crisis
Source: adapted from World Bank (2001)
CYCLONES The Bay of Bengal has been described as the ideal breeding ground for tropical cyclones (BCAS 1991). Between 1795 and 1991, more than 59 cyclones hit the coastal districts of Bangladesh with varying degrees of loss of life and property (Bramner 1997; Ali 1999). An analysis of cyclone frequency shows no definite trend. In recent years, however, cyclone preparedness has led to a considerable reduction in the loss of lives. For example, in 1970, a catastrophic cyclone with a speed of 224 kilometres per hour killed half a million people. But in 1997, when a cyclone with a velocity of 275 kilometres per hour the highest ever recorded hit the country, only 188 people died (Choudhury 2000; Chowdhury 1998). A study of the 1991 cyclone found that farmers in offshore islands lost: all standing crops; all seeds/seedlings in the field; q stored rice seeds; q agricultural implements; q draught animals; q ponds 90 percent of ponds in affected areas were submerged; q drinking water.
q q
which hit Mankkgonj district in 1969, left a trail of total destruction over 50 square miles and killed 800 people. In 1996, parts of Tangail and Jamalpur were devastated by a tornado with a death toll of over 500 people (Khan 1993). DROUGHT Permanent or prolonged droughts, which make cultivation impossible, are not common in Bangladesh. However, dry spells in the growing season known as crop droughts do affect large numbers of the rural poor, especially those dependent on cultivating rain-fed subsistence crops. It is popular knowledge that seasonal droughts prevail in the northwestern part of the country. However, in reality they affect a much wider area, from Chokoria to Rangpur and cause huge damage to crops (Paramanik 1994; Bramner 1997; Disaster Report 1997; Bramner 1999). During 1973-87, about 2.18 million tonnes of rice was lost as a result of drought.The damage to crops caused by floods during the same period was only slightly more at 2.38 million tonnes (Elahi 2001). RIVER BANK EROSION Bangladesh sits on an active floodplain characterised by annual flooding, strong currents and high sedimentation. These, and other geographic factors, result in severe river bank erosion in large parts of the country. Each year, 53 out of 64 districts suffer from river bank erosion; 17 of them from massive erosion. Akhand and Hossain (1995) reported that, between 1972 and 1990, 748 places were eroded by the action of some 70 rivers in flood.The damage during these 18 years amounted to 6.6 billion taka2 , and more than three million people were affected. In 1997, 51,000 ha of arable land and 4,084 houses were lost to erosion. The damage amounted to ten billion taka (Disaster Report 1997). ARSENIC CONTAMINATION Although cases of arsenicosis have been seen in west Bengal from the mid-1980s, its potential as a serious health issue had not yet been fully recognised. However, when Dhaka Community Hospital set up a health camp at Pakshi in Pabna district in 1996, it found patients with skin lesions similar to those seen in patients diagnosed as having arsenicosis. A quick test of tube wells revealed high levels of arsenic in the water.The latest data indicate that 59 out of 64 districts of the country have wells with arsenic levels above the safe limit set by the World Health
The cyclone caused four major health problems: cuts or injury, diarrhoea, dysentery and fever. The immediate impact of the cyclone cut across various classes and strata in the region. But those who had strong (pukka) houses were more likely to survive death or injury than groups in poorer housing. Fishermen were the most endangered group, as they received no warning while at sea; many perished and many more lost their catches. Poor households were obviously more vulnerable to the effects of the cyclone, which damaged or destroyed their houses and household assets, exposing them to subsequent hazards and setting them on a downward livelihood spiral (BCAS 1991). TORNADOES Bangladesh also suffers from tornadoes, which hit the country during April and May. From the mid-1970s, on average 104 norwesters (low-intensity rotating storms) and tornadoes have affected the country every year (Ahmed and Kabir 2001). Tornadoes mostly occur in greater Dhaka, Faridpur and Jessore districts. A tornado
The causes of vulnerability in rural livelihoods 133
Organization (WHO). About 75 million people are at risk in these districts and 24 million people are directly exposed to the toxic agent. Over 7,000 arsenicosis patients have been identified so far (Anwar 2000). The most common manifestation of the disease caused by arsenic contaminated water is skin lesions. But over the next decade internal cancers may be major health hazards (WHO 2001). Although there has been little study of the socioeconomic impact of arsenic contamination, reports suggest that its impact is alarming. In one village, 30 percent of the households and six percent of the population were affected by arsenicosis. The deteriorating health of these patients was likely to cause severe economic hardships for their families, as many of them were income earners. Further problems arise from the social exclusion of affected households. The disease is viewed as contagious by villagers and in some cases the sick cannot move freely. It can also decrease the marriage prospects of young women (Nizamuddin and Chakraborty 2001). WATERLOGGING Waterlogging occurs when rain or flood water ponds in depressed areas and soft topsoil remains stagnant in the absence of appropriate drainage channels (Adnan 1991). On a large scale, waterlogging inhibits cultivation, damages crops, prevents crop rotation and reduces cropping intensity (Adnan 1994). It is a major problem for agriculture in the southwestern and northwestern districts. In 1997, waterlogging was a serious problem in 16 districts, damaging 20,000 houses and 4,000 ha of crop land.About one million people were affected (Disaster Report 1997).
SALINITY Salinity affects the livelihoods of people in coastal regions. It is caused by sea water encroaching the flood plains at high tide and gradually making the alluvium saline and reducing its fertility (Karim, Hussain and Ahmed 1990). Shrimp cultivation, which requires brackish water, is known to have exacerbated the problem in the coastal belt (GoB 1995). More than half of the net cultivable area in 13 coastal districts are affected by salinity. Soil salinity limits the choice of crops, especially during the dry season, can reduce yields and lead to crop loss. CLIMATE CHANGE AND SEA-LEVEL RISE A recent study by Singh et al. (2000) showed that the mean tidal level at three places (Hiron Point, Char Changa and Coxs Bazaar) is increasing at a rate of 4, 6 and 7.8 mm/year respectively. Other studies have estimated that a one-metre rise in sea-level would affect 17.5 percent of the total area of Bangladesh and eleven percent of its population (Table 2). Three districts would be completely inundated. Over 21 percent of the countrys rice producing land, accounting for 16 percent of the total rice harvest would be lost. Overall it is estimated that losses would be equivalent to 13 percent of GDP or 450 billion taka. FLOOD Flood is inseparably linked with the livelihoods of people in Bangladesh. They are the lifeline of agriculture, but they also bring destruction. Officially, floods are defined as levels of river overflow that, at the very least, damage property. The monsoon rains annually inundate about one-fifth of the country and nearly one-third of its agricultural land.
Table 2 Districts, area and population likely to be affected by a 1m rise in sea-level

Districts Area affected (%) Area affected (ha) Affected population
Barisal Patuakhali Khulna Noakhali Comilla Faridpur

Total
90 100 80 50 15 15
6,600,000 410,000 975,000 275,000 100,000 100,000

8,460,000
4,500,000 2,000,000 3,000,000 1,500,000 5,000,000 1,500,000

17,500,000
Source: Haq et al (1994)
The flood regime is complex with four different flood types: (i) flash floods, which occur in the northeast and northwest of the country and are caused by heavy and sustained rainfall, which leads to rapid run-off in the premonsoon period.The sudden onset of these floods causes massive damage to agriculture; (ii) monsoon or river floods: major rivers rise slowly during the monsoon and inundate extensive areas along their banks. Recession is slow and high flow rates keep the inundated area submerged for weeks; (iii) rainfall floods high-intensity local rainfall and/or prolonged monsoon in the absence of proper drainage can lead to flooding. In natural depressions or semi-arid areas, water may remain stagnant for an extended period of time, threatening livelihoods; and (iv) coastal floods part of the low-gradient coast line is flooded by high water waves resulting from the frequent cyclonic storms in the Bay of Bengal. Floods have caused massive loss of property and life in the country over the years.The floods of 1987, 1988 and 1998 were particularly catastrophic, resulting in large-scale destruction of property and considerable loss of life. The floods of 1987 and 1988 caused an estimated damage of about US$2 billion and reduced GDP substantially (GoB 1987; Flood Plan Coordination Organization 1992b; Siddique 1989). Whether the 1980s saw an increase in flood events is debatable. In a meticulous piece of research, Hofer (1998) showed that the structural measures of flood control (constructed under the Flood Action Plan) have reduced the incidence of normal floods, which are essential for agricultural production, but have not been able to contain catastrophic floods. According to Hofer, ironically, embankments serve to prevent normal floods whilst failing to prevent abnormal floods, thus reducing the beneficial functions of normal flooding (1998: 20). Floods impact on livelihoods by disrupting economic activities, destroying household assets, and causing health problems.Women are especially vulnerable: they have to prepare food under difficult conditions, collect water from a distance and may even reduce or sacrifice their own food for the sake of family.They will often be forced to abandon parda (Begum 1995). Even after the floods have receded, problems continue with the shortage of essential commodities and consequent rises in price.
Assets of poor households are thus further depleted. People adopt several coping strategies: they may sell and mortgage land, or dispose of other assets including agricultural implements and livestock. Some turn to common property resources. Available evidence (Hossain et al. 1992) suggests that floods do not lead to any permanent change in occupation patterns. But normal livelihood activities are disrupted. Members of poor households take whatever work they find: many take up fishing and work as boatmen, others migrate to nearby urban and semiurban areas and work as vendors, porters, cart drivers or secure other informal sector jobs. Class inequality increases as the rich purchase land and lend money at high interest rates. Floods put the livelihoods of char3 dwellers at risk (Schmuck-Widmann 1996;Ashley et al. 2000).They may be at a greater disadvantage because alternative jobs are not available in the chars. Migration is not an option unless a person has a strong kinship network. Relief and other help for the chars are often delayed due to poor communication. River bank erosion is a very great threat and means that many char dwellers have to move every few years. One study showed that, on average, char dwellers had moved eight times by the time they reached 44 years of age (Schmuck-Widmann 1996).
OTHER SOURCES OF VULNERABILITY

Households suffer not only from natural disasters, but also from a broad range of other factors. Rahman (1995) maps out the risk factors that create vulnerability for rural people and lead to downward spirals and trends in livelihoods. From an analysis of poverty trends in 62 villages, he noted the importance of life-cycle factors (such as high dependency rates in the household), structural factors (such as low opportunities and inflation) and crisis factors (such as natural disasters, illness and property loss). One interesting finding of the analysis was that life-cycle factors were responsible for downward mobility in 46 percent of cases. A little more than one-third of cases of downward mobility were due to crisis factors and structural factors were responsible for 17 percent of the cases. According to Siddiqui (2000), the factors that led to deterioration in poor peoples livelihoods were: death/abandonment by male head of the family; decline of access to land and
The causes of vulnerability in rural livelihoods 135
ponds; large family with few income earners; disease/old age; and litigation.
VULNERABLE SOCIAL GROUPS

Vulnerability is inextricably linked to poverty. Traditionally, the most poor and vulnerable occupational category in Bangladesh was thought to be the landless non-farm workers. But the Household Expenditure Survey (199596) suggested that landless agricultural workers were even worse off.They have a 75 percent chance of being poor (World Bank 1998). Over 15 percent of households are female-headed in rural Bangladesh and they are an acutely vulnerable group.Three-quarters of female-headed households are landless and of these households, onethird of female heads are agricultural labourers (Mannan 2000). The same source showed that female-headed households are acutely vulnerable due to the following factors: marginal control over land; q higher dependency on wage earning; q greater involuntary unemployment; q lack of literacy or lower level of education; q greater age of the household head; q less access to household labour.
q
labour power, female-headed households, traditional occupational groups (such as potters, oil pressers, weavers) and some ethnic groups. The characteristics of these households (lack of labour, low access to assets, social discrimination, etc.) make them acutely vulnerable to the multi-dimensional shocks and stresses that affect rural Bangladesh. In many cases, efforts to move out of poverty are undermined by the incidence of shocks and stresses, which wipe out any gains they might have made. Poverty reduction strategies must address the risks, shocks and stresses that the rural poor struggle to overcome.The World Bank states that a future policy for the mitigation of vulnerability needs to address three critical issues: employment generation; insurance; and safety nets (2001). In terms of employment generation, the greatest opportunities in rural Bangladesh lie in the non-farm sector. But the poor and vulnerable need targeted assistance to help them take advantage of the new opportunities. However, more effective broad-based safety net measures will need to be undertaken to target the extreme poor and the most vulnerable people in society.
NOTES
1 Mastans are local strongmen or hoodlums who engage in undesirable criminal activities including murder (mastani).The term mastanism has been used in the context of political culture of Bangladesh to refer to behaviour associated with an illegitimate use of power and or deployment of violence by party activists (Khan et al. 1996). 2 As of May 2002, US$1 is equivalent to 57.25 taka. 3 Chars are small islands formed in the Bay of Bengal through the accumulation of silt and other materials brought by the powerful rivers of Bangladesh.Their size and coastlines are fluid, shifting with the interplay of current and tide.
The elderly are another vulnerable group increasingly so due to the breakdown in the institution of the extended family.A large number of children are also vulnerable as they work to support their families: nearly 29 percent of rural boys (6-17 years of age) and four percent of girls work (Stalker 1996).They receive low pay and are denied basic human rights.
CONCLUSIONS
The development process is always uneven and creates winners and losers. In the 1970s there was serious concern that the Green Revolution would accentuate inequality, displace labour and lead to pauperisation in rural Bangladesh. But this does not seem to have happened (Siddiqui 2000). There has been an overall improvement in the livelihoods of the majority of rural people. But some poor households have not been able to take advantage of the new opportunities those with low
REFERENCES
ADNAN, S. (1991): Floods, People and the Environment. Research and Advisory Services, Dhaka. ADNAN, S. (1994): Perspectives from the Dry Season: Interactions between river activity, sedimentation, waterlogging, floods and water control structures, in A. Atiq Rahman et al. (ed.) Environment and Development in Bangladesh.Vol. 1. University Press Ltd., Dhaka. AHMAD, N. AND KABIR S.A. (2001):Tornadoes in Bangladesh, in K. Nizamuddin (ed.) Disaster in Bangladesh: Selected Readings. Department Geography and Environment, Dhaka University, Dhaka. ALI, A. (1999): Climate Change Impacts and Adaptation Assessment in Bangladesh. Climate Research 12: 109-116. ANWAR, J. (2000): Arsenic Poisoning in Bangladesh. Palash Media and Publisher, Dhaka. ASHLEY, S., KAR, K., HOSSAIN, A. AND SHIBABRATA, N. (2000):The Chars Livelihoods Assistance Scoping Study: Final Report. DFID Bangladesh. BANGLADESH FLOOD ACTION PLAN (1992): FAP-23 Flood Proofing Study. ISPAN, Dhaka. BRAMMER, H. (1997): Agricultural Development Possibilities in Bangladesh. University Press Ltd. Dhaka. BEGUM, M. (1995): Coping with Floods:The Experience of Rural Women in Bangladesh. Ph.D. dissertation, Massey University, New Zealand. CHOWDHURY, Q.I. (ed.) (1998): Bangladesh State of Environment Report 1999. Forum of Environmental Journalists, Bangladesh, Dhaka. CHOWDHURY, A. (2000): Disasters: Issues and Responses. In P. Gain (ed.) Bangladesh Environment: Facing the 21st Century. Society for Environment and Human Development (SEHD), Dhaka. DISASTER REPORT BANGLADESH (1997): Disaster Forum Bangladesh 1998. Dhaka. ELAHI, K.M. (2001): Drought in Bangladesh: A Study of Northwest Bangladesh, in K. Nizamuddin (ed.) Disaster in Bangladesh: Selected Readings. Department Geography and Environment, Dhaka University, Dhaka. GOVERNMENT OF BANGLADESH (1995): National Environment Management Plan.Vol. 1a. Ministry of Environment and Forest, Dhaka. HASSAN, M., AKANDA, A.M. AND HOSAIN, M.I. (1995): Bangladesh Physical Geography and Environment (in Bangla). Bangla Academy, Dhaka. HOFER, T. (1998): Floods in Bangladesh: A HighlandLowland Interaction. University of Berne. HOSSAIN, M., ISLAM, A. AND SAHA, S.K. (1992): Floods in Bangladesh: Recurrent Disasters and Peoples Survival. University Research Centre, Dhaka. HUQ, S., SYED, I.A. AND RAHMAN (1994): Sea Level Rise and Bangladesh, in A. Rahman et al. (ed.) Environment and Development in Bangladesh, Vol. 1. University Press Ltd, Dhaka. KHAN, F.A. (1993): Living with Natural Hazards: Bangladesh. House of Consultant Ltd, Dhaka. MANNAN, M.A. (2000): Female-Headed Households in Rural Bangladesh: Strategies for Well-being and Survival. CPD, Dhaka. NIZAMUDDIN, K. AND CHAKRABORTY, R.K. (2001): Arsenic Pollution in Hatkopa Village, in K. Nizamuddin (ed.) Disaster in Bangladesh: Selected Readings. Department Geography and Environment, Dhaka University, Dhaka. PARAMANIK, M.A.H. (1994): Natural Disaster, in A. Rahman et al. (ed.) Environment and Development in Bangladesh, Vol. 1. University Press Ltd., Dhaka. RAHMAN, H.Z. (1995): Crisis and Insecurity:The Other Face of Poverty, in H.Z. Rahman and M. Hossain (eds.) Rethinking Rural Poverty. University Press Ltd., Dhaka. SCHMUCK-WIDMANN, H. (1996): Living with the Floods: Survival Strategies of Char-Dwellers in Bangladesh. FDCL, Berlin. SIDDIQUI, K. (2000): Jagatpur 1977-97: Poverty and Social Change in Rural Bangladesh. University Press Ltd., Dhaka. SINGH, O.P., KHAN, T.M.A. AND RAHMAN, M.S. (2000): The Vulnerability Assessment of the SAARC Coastal Region due to Sea Level Rise: Bangladesh Case. SMRCNo. 3. SMRC, Dhaka. STALKER, P. (1996): Child Labour in Bangladesh. UNICEF, Dhaka. WORLD BANK (1998): Bangladesh: From Counting the Poor to making the Poor Count.World Bank, Dhaka. WORLD BANK (2001): World Development Report 2000/2001: Attacking Poverty. Oxford University Press, New York.
Lintu, this area is for a barcode

Hands Not Land Book

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Hands Not Land Book

Uploaded by

Copyright:

Available Formats

Hands Not Land:

How Livelihoods are Changing in Rural Bangladesh

Kazi Ali Toufique Cate Turton

Hands Not Land:

How Livelihoods are Changing in Rural Bangladesh

Kazi Ali Toufique Cate Turton

Hands Not Land:

How Livelihoods are Changing in Rural Bangladesh

Part 1: Poverty in a changing rural landscape

16 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

THE STORIES BEHIND THE STATISTICS

Table 1 Trends in landlessness, 1983-96

2.00 19.64 28.21 12.32 37.82 100.00

162,229 5,003,042 5,191,979 2,494,606 4,976,326 17,828,182

0.91 28.06 29.12 13.99 27.91 100.00

-2.93 5.23 2.42 3.23 -0.39 2.15

18 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

A NEW RURAL DEVELOPMENT NARRATIVE

Key features of the approach

STRUCTURE OF THIS BOOK

Chapter 2 What is life like for the poor in rural Bangladesh?

22 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

What is life like for the poor in rural Bangladesh? 23

24 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

What is life like for the poor in rural Bangladesh? 25

26 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

What is life like for the poor in rural Bangladesh? 27

28 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

Table 1 Institutional actors and functions

What is life like for the poor in rural Bangladesh? 29

ONE-OFF HEADLINES OR THE EMERGENCE OF A NEW NARRATIVE?

CHANGE BRINGS NEW WINNERS AND LOSERS

30 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

What is life like for the poor in rural Bangladesh? 31

Barriers to entry impeding poor peoples access to new opportunities

32 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

IMPLICATIONS FOR EXPLORING LIVELIHOODS

QUESTIONS THE NEW RURAL REALITY RAISES

34 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

Understanding livelihood dynamics

HOW COMPLETE IS OUR PICTURE?

36 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

Hands Not Land:

How Livelihoods are Changing in Rural Bangladesh

Livelihoods in contemporary rural Bangladesh: An overview of research papers

40 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

An overview of research papers 41

42 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

An overview of research papers 43

44 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

An overview of research papers 45

TRENDS IN FARM AND NON-FARM LIVELIHOODS

Farm Non-farm All

72.70 27.30 100.00 (13,817,646)

57.08 43.47 100.00 (5,495,300)

66.18 33.82 100.00 (17,828,187)

47.18 52.82 100.00 (6,401,453)

1.35 3.98 2.15

-0.26 2.89 1.28

48 Hands not Land: How Livelihoods are Changing in Rural Bangladesh

CHANGES IN LAND DISTRIBUTION AND TENURE

13,818 3,772 10,045 7,066 2,483 496 1,198 5,495

17,828 6,030 11,798 9,423 2,078 297 1,815 6,401