IAbout Child Poverty - A Bangladesh’s Perspective Syeda Shahanara Begum University of Gothenburg Department of Social Work 2012 II Skriftserien : 2012:3 University of Gothenburg Department of Social Work © Syeda Shahanara Begum Print: Ineko AB, Göteborg 2012 ISBN 91-86796-68-6 ISBN 1401-5781 III Abstract Title: About child poverty – A Bangladesh’s Perspective Author: Syeda Shahanara Begum Key words: Child poverty, Bangladesh, education, spatial difference, multidimensional deprvation, intergenerational transmission of poverty, public services, corruption, economic growth, and income inequality. Distribution: University of Gothenburg, Department of Social Work, P.O. Box 720, 405 30, Göteborg ISBN: 91-86796-68-6 ISSN: 1401-5781 Internet: http://hdl.handle.net/2077/29092 Taking child poverty into account as an enormous concern on the pathway to human development, the pres- ent dissertation aims to examine child poverty’s extent and characteristics, poor children’s views on this issue and their policy recommendation to reduce it. The research questions are: 1) what are the extent and characteristics of child poverty in Bangla- desh? 2) What is child poverty? Why and how do children experience it? 3 (a) To what extent and in what respect does child poverty differ between Bangladesh and China? 3 (b) What are the reasons for the differ- ences in child poverty over time between the two countries? 4 ) What is needed to reduce child poverty in Bangladesh according to its principal victims? The present dissertation is based on research mainly with young children aged up to 14 years. A combination of quantitative and qualitative studies’ inclusion is a strength of this dissertation, which further offers to look at child poverty from different perspectives. For quantitative studies (Chapter 5 and Chapter 7), the empirical materials consist of microdata from: a) Bangladesh HIES for the years 1995 and 2000; b) Chinese data from China Household Income Project (CHIP) of the years 1988, 1995, and 2002. For qualitative studies (Chapter 6 and Chapter 8), data from five focus group discussions––conducted during 2005 and 2006 with 30 participants––are used. The methods for data analysis in quantitative studies are: i) descriptive and multivariate analysis; ii) decomposition framework. A grounded theory approach is applied to analyse the data for qualitative studies. Nine chapters constitute this dissertation. Chapters 1 to 4––based on secondary data––introduce background information, theoretical discussion, methodological issues and overviews of forthcoming chap- ters. Chapters 5-8 contain four empirical studies. The conclusion draws some broad inferences in Chapter 9. The dissertation mainly finds that Bangladesh’s children make up the greater share of the popula- tion where almost half of the poor are children; child poverty rates are––similar to China’s––higher than the adults’ and more extensive than in China (Chapter 5 and Chapter 7). Child poverty plays a vital role in the prolongation of developing, expanding, extending and transmitting poverty on to successive generations; three different interlinked stages––encompassing multidimensionality, spending life in distress, and having intergenerational and gendered dimension––are disclosed in this progression (Chapter 6). Additionally, the effectiveness of growth and income inequality do affect––but not always––child poverty differences across time and countries. Other demographic factors are revealed to play vital roles (Chapter 7). Household head’s education demonstrates strong negative association with child poverty (chapters 5-8). Participants recommended a combination of policies to enhance the capability of poor children and their caregivers (Chapter 8). Policy interventions need to give further attention to: reduce parental pov- erty and income inequality, sustain economic growth, ensure access to education and health care, expose corruption and hidden costs of these services, and eliminate mistrust of the recipients to speed up the extent of child poverty’s reduction in Bangladesh. IV VPreface Identifying and counting children in poverty, knowing and understanding their experience, find- ing explanations for staying in such a situation, locating what steps are taken as remedy and what more are needed to reduce poverty etc. are not straightforward task to do. These are clearly very complicated tasks (dealing with value judgments in particular). My endeavour in this dissertation is to put forward some of these important issues for those who work or plan on working for free- ing poor children in Bangladesh of absolute or extreme poverty. Working on this dissertation was a very exciting, interesting, thought-provoking and passionate endeavour. Although a great deal of this writing is done by my own effort, I would not be able to complete this dissertation without the assistance and support of many others. Therefore, I would like to thank those individuals for their direct and indirect support. Foremost, I would like to thank all the participants of the focus group discussions and those who helped me to organize and make focus group discussions possible and usable for the studies. If they would not share their experiences, perceptions and insights, this dissertation would not come to exist. I am extremely grateful to my supervisor Professor Björn Gustafsson (at the Depart- ment of Social Work at Gothenburg University) for the steady stream of detailed and valuable comments and suggestions that have been so supportive in producing this dissertation. Without his encouragement and steady guidance, I could not have finished this dissertation. He was con- tinuously eager to meet and discuss my ideas, to give me constructive criticism and feedback on different parts of this dissertation, and help me find solutions whenever I was stuck throughout my dissertation. Special thanks to Professor Rafael Lindqvist (presently at the Department of Sociology, Uppsala University, earlier at the Department of Social Work of Gothenburg Univer- sity) for his additional co-supervision, encouragement and guidance and inspiration, which were really helpful to continue to work on this dissertation. Particularly, I am grateful to him for his significant inputs in presenting qualitative studies. I highly appreciate both of my supervisors’ contributions with time, ideas and efforts for funding. I am extremely grateful to Professor Sven-Axel Månsson (presently at Malmö University, VI earleir at the Department of Social Work at Gothenburg University) Professor Ulla Björnberg (at the Department of Sociology) and Nora Machado des Johansson (Associate Professor at the Department of Sociology of Gothenburg University) for their encouragement and support in pur- suing my higher studies. My thanks stay on to Professor Margareta Bäck-Wiklund, and Senior lecturer Maren Bak for their support as well. I would like to thank Professor Peter Dellgran, for his comments on my studies in seminars. Also, thanks are due to Mona Franséhn, Senior lecturer, Head of the Department (of Social Work at Gothenburg University) for her support. I also wish to thank Professor Bjorn Halleröd (at Sociology Department Gothenburg Uni- versity, Sweden) for his precious comments at the seminar held in Gothenburg University. I wish to acknowledge the constructive criticism from all the participants of several seminars held in Gothenburg University since 2004. Further, I am grateful to seminar participants at: International conference of UNICEF and the Graduate Program in International Affairs (GPIA)1, and Interna- tional Conference on Experiences and Challenges in Measuring National Income and Wealth in Transition Economies2. I thank the personnel of Bangladesh Bureau of Statistics (BBS)3 for the support of get- ting accessto the micro data of Bangladesh in a reduced price. For technical assistance to deal with micro data with software program Stata and suggestions, and comments on some part of this dissertation, I would like to thank Doctor Sten Dieden (now at UNEP, Risoe Centre on Energy, Climate and Sustainable Development, Roskilde, Denmark). Including my supervisors, I would like to thank each member of the readers groups, Professor Ingrid Sahlin and Senior lecturer Torun Österberg of the Department of Social Work at Gothenburg Universityfor their help in discussing methodological and conceptual issues, and useful comments on earlier drafts of this dissertation. For useful comments on some portions of this dissertation, I wish to thank specially discussant Professor Peter Saunders (at Social Policy Research Centre at University of New South Wales, Australia) and Professor Alberto Minujin (at 1 Held in 2005 (April 25 to 27) at New School University. New York, USA. 2 Held in 2007 (September 19 to 21) in Beijing, China, jointly organised by the International Association for Research on Income and Wealth (IARIW) and the National Bureau of Statistics (NBS), China. 3 Bangladesh Bureau of Statistics is the National Statistical Organisation of Bangladesh. Its main responsi bilities are: collecting, compiling and disseminating statistics of all the sectors of the country’s economy. VII the Graduate Program in International Affairs [GPIA] at the New School University. New York, USA). I am also grateful for helpful and critical comments and discussions and with Howard White (of Operations Evaluation Department, the World Bank, Washington D.C). I would like to thank my co-research fellow Doctor Deng Quheng (of Institute of Eco- nomics Chinese Academy of Social Sciences Beijing China). Thanks goes to all the staff, Ingeg- erd Franzon in particular, and my colleagues, who are all PhDs at present: Andreas Liljegren, Charlotte Melander, Monica Nordenfors, Margareta Regnér, at the Department of Social Work at Gothenburg University more widely. In addition, I would like to thank many teachers in the past, Linda Lane (Senior lecturer) and Annika Halen in particular (Director of Studies at Gothenburg University). I am, especially, indebted to my late parents, and all my sisters, brothers and relatives. Finally, I owe my loving thanks to my husband, my daughter, and son. Without their support, encouragement and understanding, it would not have been possible for me to come to an end this work. Gothenburg, April, 2012 Syeda Shahanara Begum VIII Contents ChAPter 1: IntroduCtIon .................................................................................... 1 1.2. Overview of chapters ............................................................................................. 5 ChAPter 2: Poverty And ChIld Poverty ....................................................11 2.1. Defining poverty and child poverty .................................................................... 11 2.2. Concepts and terms .............................................................................................. 14 2.3. Approaches to poverty ......................................................................................... 18 2.4. Measuring poverty ............................................................................................... 28 2.5. Staying in poverty ................................................................................................. 32 2.6. Intergenerational transmission of poverty ....................................................... 34 2.7. Child poverty discourse in Bangladesh .............................................................. 35 ChAPter 3: ABout BAnglAdesh ....................................................................... 38 3.1. Bangladesh and some other countries in HDI and HPI in HDR (2011) ............... 38 3.2. Introduction to Bangladesh ............................................................................... 43 3.2.1. A history of backwardness ............................................................................... 45 3.2.2. Population ........................................................................................................ 48 3.2.3. Politics, parliament and administrative units .................................................. 50 3.2.4. Regions ............................................................................................................ 53 3.2.5. Economy ........................................................................................................... 57 3.2.6. Education and employment ............................................................................. 58 3.2.7. Health care ....................................................................................................... 60 3.2.8. Policies and programs ...................................................................................... 62 ChAPter 4: MethodologICAl Issues ........................................................... 65 4.1. Few words about methodological issues ........................................................... 65 4.2. Quantitative Studies ............................................................................................. 66 4.2.1. Data description ............................................................................................... 67 4.2.2. Approaches to data analysis ............................................................................ 68 4.3. Qualitative studies ................................................................................................ 72 4.3.1. Data description ................................................................................................. 72 4.3.2. Approaches to data analysis .............................................................................. 77 ChAPter 5: ChIld Poverty In BAnglAdesh: A study BAsed on sur- vey dAtA of the yeAr 2000 .................................................................................. 83 5.1. Introduction ......................................................................................................... 83 5.2. Description of child poverty ............................................................................... 85 5.3. Probability of a child being poor ........................................................................ 89 5.4. The notion of well-being and capability ............................................................. 92 5.5. Conclusion ............................................................................................................ 94 IX ChAPter 6: ChIld Poverty In BAnglAdesh’s Context: A quAlItA- tIve APProACh ......................................................................................................... 96 6.1. Introduction ......................................................................................................... 96 6.2. Research approach ............................................................................................... 97 6.2.1. The Approaches to Analysis ............................................................................. 98 6.3. A discussion of child poverty ............................................................................100 6.3.1. Encompassing multidimensionality (Category A)........................................... 101 6.3.2. Spending life in distress (Category B) ............................................................. 106 6.3.3. Having intergenerational and gendered dimension (Category C) .................. 113 6.4. Passing poverty from one generation to the next ...........................................120 6.5. Summary and discussion .................................................................................... 125 ChAPter 7: do eConoMIC growth And InequAlIty AlwAys MAt- ter In ChIld Poverty reduCtIon? evIdenCe froM BAnglAdesh And ChInA .................................................................................................................. 129 7.1. Introduction ....................................................................................................... 129 7.2. Context ................................................................................................................ 136 7.3. Data ...................................................................................................................... 139 7.4. Describing child poverty .................................................................................... 143 7.5. A framework for making poverty comparisons ............................................... 147 7.6. Results ................................................................................................................. 150 7.7. Conclusions ........................................................................................................ 154 Appendix .............................................................................................................. 156 ChAPter 8: ChAllenges And APProAChes to ChIld Poverty In BAnglAdesh: A quAlItAtIve study ......................................................... 157 8.1. Introduction ....................................................................................................... 157 8.2. Poverty alleviation programs ........................................................................... 158 8.3. Institutional weaknesses in policy implementation .........................................160 8.4. Research approach ............................................................................................. 161 8.5. Results ................................................................................................................. 163 8.6. Summary and discussion .................................................................................... 174 ChAPter 9: fInAl thoughts ............................................................................ 178 9.1. Few words on the findings from the empirical results .................................. 181 9.2. Policy recommendations .................................................................................... 186 9.3. Summing up ......................................................................................................... 189 References ................................................................................................................ 191 X 1Chapter 1 Introduction Poverty is undesirable and intolerable in anywhere in the world for the reasons of having mul- tidimensional and damaging consequences on individuals regardless of age and sex. Poverty of children in particular leads to unacceptable consequences, which a child must not face. Child poverty as the deprivation of a number of basic needs and social exclusion in terms of that limits children’s participation in society and leads them to severe and complex consequences (UNICEF, 2007). Such a description indicates that the detrimental effects of child poverty reinforce the inter- generational transfer (IGT) of poverty, although some children might move out of poverty when they are adults (Harper, 2004). In a comparative study of rich countries, Corak (2006) found that family economic sta- tus was significantly associated with the labour market success of children in adulthood. In the context of developing countries, studies (Bird, 2007; Lyytikäinen et al., 2006) show that different variables characterize child poverty and are responsible for future labour productivity; i.e. less schooling and lower educational attainment, low income, poor nutrition and inadequate health care and lack of social protection. Particularly, low educational attainment prevents the poor from increasing earnings and income thereby perpetuating poverty in developing countries (Deolalikar, 2005; Ravallion & Wodon, 2000). In addition, the association of parental education and wealth with children’s educational enrolment and achievement is evident (Moore, 2001), for example, in the Sub-Saharan Africa (Quisumbing, 2009) and also in the American context, as parents with low levels of schooling cannot obtain jobs that pay enough to keep their families out of poverty (Corcoran & Chaudry, 1997). Evidence from Peru, Viet Nam, Malaysia, Brazil and Indonesia show that the less resourc- es a household has means lower investment in children (Quisumbing, 2009). Overall evidence from several studies shows that the lower the level of education of children’s caretakers, the less 2the likelihood that they will be able to keep children out of poverty. The issue of child poverty has received less attention than that of poverty of adults among the researchers and alike around the world for a long period, particularly, in the context of de- veloping countries, whereby research on poverty of adults received much attention. In the recent past, however, the need for a child focused perspective in the development and poverty reduction process is brought into notice increasingly by some research, such as literature on global studies as well as country-specific reports (see, for example, Bradbury, Jenkins, & Micklewright, 2001; Corak, 2006; Dieden & Gustafsson, 2003; Gordon, Nandy, Pantazis, Pemberton, & Townsend, 2003; Harper, 2004; Moore, 2001; UNICEF, 2004, 2006). Why is a child-focused approach towards poverty important? Several focal points are stressed in literature in this context. The majority of this work argues that despite prevailing economic growth and substantial fall in poverty, children living in extreme poverty are a great concern for many countries (UNICEF, 2006). This has also been expressed in an empirical study in Gordon et al. (2003), in which the authors use household and individual survey data from 46 developing countries of 1990s and much more available recent data during the time of the study in most cases. The above-mentioned research argues that more than one third of the developing world’s children––over 674 million––are living in absolute poverty. Half of the children of Sub-Saharan Africa are severely deprived of shelter. Children in East Asia are found least likely to be severely deprived. Among the concerned countries, the lowest poverty rate is found in East Asia and the Pacific region, at 7 per cent (43 million children). The highest rate is in Sub-Saharan Africa, at 65 per cent (nearly 207 million children). More than half of the world’s severely food deprived children reside in South Asia. Not surprisingly, a high rate of absolute poverty, with 59 per cent (330 million children) is also found in South Asia. These statistics are alarming in the context that children are at a higher risk of poverty, particularly, in the developing world. Among others, Redmond (2008) and Roelen, Gassmann, and Neubourg (2009), from an extensive literature review, argue for the necessity of child focused poverty measures due to: children, generally, are dependent on the environment to meet up their 3basic needs; by this they mean that required resources for meeting children’s basic needs are held by their parents, household or community members.4 The facts and arguments summarized above tend to suggest that information on this distribution and about poverty at the child-specific level are very essential. Being born, living, and growing up in poor families put doubts on children’s success in their adulthood. This dissertation considers child poverty as a great challenge for any country of the world in terms of human and economic development. More precisely, this certainly appears to be a great challenge to the lessening of poverty in general. The reasons are: firstly, it reinforces intergenerational poverty persistence (Barrientos & DeJong, 2006, p. 2). In other words, growing up in poverty puts the children at risk of being poor in their adulthood and passing poverty from parents to their children and on to their children (see Barrientos & DeJong, 2006; Corak, 2006; Harper, 2004; Moore, 2001). Reducing child-poverty would thereby also reduce adult poverty in the long run (Roelen et al., 2009) need to be developed. Secondly, the children, particularly in the developing countries of Asia and Africa, con- stitute a greater share of the population and many of them are poor (Gordon et al., 2003;Minujin, Delamonica, Davidziuk, & Gonzalez, 2006;UNICEF, 2006). Bangladesh is one of these countries where children make up around forty percent of its total population (BBS5, 2003). Thirdly, chil- dren are the most vulnerable group in poverty and they are not responsible for it (see Bradbury et al., 2001; Harper, 2004; Leisering & Leibfried, 1999; Moore, 2001). Finally, poverty alleviation is not possible by implementing those policies that do not take into account children’s needs. Re- search based knowledge and insight of children’s poverty status can obviously play a vital role in the process of poverty reduction, policy formulation and implementation, by taking into account children’s specific needs. For this reason, more research is needed. This dissertation is a response to this demand and highly motivated by the reasons as mentioned above for studying child poverty in Bangladesh, where it can be assumed that a sig- nificant number of the world’s poor children live. The concept of child poverty needs to be under- 4 Children’s economic dependence varies to a great extent by age across groups of children and countries, particularly in the developing world (Redmond, 2008). 5 Bangladesh Bureau of Statistics. 4stood and addressed from a number of perspectives, since it affects children “through a number of channels” (Barrientos & DeJong, 2006, p. 538). More importantly, children’s experience of poverty and its impacts are unlike that of adults. Particularly, dietary requirements and the role of education during the early stage of life, which are especially crucial for human development (Redmond, 2008). A child-specific approach to poverty can put forward those needs. Specific research questions of this dissertation are: 1) What are the extent and characteristics of child poverty in Bangladesh in terms of World Poverty Line? 2(a) What is child poverty according to poor children and mothers (its principal victims) in Bangladesh? 2(b) How do they experience and account for it? 3(a) How does child poverty (if child poverty estimation is based on 1 US Dollar per person per day in 1985 PPP terms) in Bangladesh relate to child poverty in China, and particularly in Southwest China? 3(b) Based on the experience from Bangladesh and China: is poverty reduction only a question of income growth? 4) What is needed to reduce child poverty in Bangladesh according to previous research, poor children and mothers? This dissertation is based on research mainly with the young children aged up to 14 years, and attempts to get more knowledge on child poverty by using both quantitative and qualitative studies. Quantitative methods are used to answer the first and third research questions. Research question 1 is analyzed by only Bangladesh Household Income and Expenditure Survey (HIES) of the year 2000. Research question 3 (a) and 3 (b) use data of Bangladesh from HES (Household Expenditure Survey) of the year 1995, and data HIES, the same data that are used in analyzing research question 1; and China household income survey of the years 1988, 1995, and 2002 as well. The Chinese data are drawn from CHIP (China Household Income Project) survey. These were conducted in 1989, 1996 and 2003 for the periods 1988, 1995 and 2002 respectively. In answering the second and fourth research questions, I considered qualitative studies to 5be appropriate to examine some factors which survey data has limitations to reach, and, therefore, data from five focus group interviews has been used in these studies. Studying experiences and perspectives of parents and children makes it possible to examine the diversities in children’s experiences in child poverty contexts. While data used for answering questions 1, and 3 (a) and 3(b) were collected by statistical authorities, I myself collected data for addressing questions 2 and 4. To deal with research question 2, I used data from four focus groups composed of poor participants. However, for research question 4, I used data from five focus group interviews that include the same four groups composed of poor participants and one additional group composed of non-poor participants. Each focus group interview lasted around one hour and was conducted in a period from June 2005 to July 2006. The unit of analysis in this dissertation is, mostly, child. The central term in the analysis is child poverty. I also included the notion of wellbeing in the analysis of child poverty. In addi- tion, victim’s perceptions on deprivation, capabilities and functionings, and social exclusion have come out into light in qualitative studies. In the rest of this chapter, I present an overview of the chapters of this dissertation elaborate on some of the major issues in relation to child poverty. 1.2. Overview of chapters This dissertation is composed of nine chapters with Chapter1 to Chapter 4 being descriptive and using data from secondary sources. These chapters introduce the topics, and objectives of differ- ent studies that are included in the dissertation. The dissertation continues by presenting its differ- ent studies in the successive four chapters (5 to 8). Chapter 9 then wraps up the thesis by adding a conclusion. 6Table 1 presents a summary of aims, methods and data in different chapters of the dissertation. Henceforth, the remainder of the dissertation is structured as follows: In Chapter 2, a general discussion on some of the existing literature on poverty and child poverty is offered in order to spot related information from prevailing knowledge. The main focus of this chapter is to provide a review of literature as a useful background of my research topic. Various aspects are discussed, such as how literature deals with defining poverty and child pov- erty with respect to the issues of subjectivity vs. objectivity, the choice of indicators, and absolute vs. relative poverty etc. In this chapter, I also address some of the most frequently used concepts and terms that might have differences in meaning and in context to both poverty and child pov- erty, but are closely linked. A number of approaches are dealt with prior to the section “measuring poverty” in order Table 1. A summary of aims, methods and data in different chapters of the dissertation. Aims Methods Data Introductory Chapters Introduction of different aspects of the dissertation Descriptive Secondary data Chapter 1 Introduction …….. ………. Chapter 2 Appraisal of child poverty and relevant issues Literature review Secondary data Chapter 3 Introduction on Bangladesh Descriptive Secondary data Chapter 4 Discussion on methodology used Descriptive Secondary data and description of my own approaches Studies To study child poverty Quantitative and qualitative Data from national household surveys, secondary sources and focus group interviews Chapter 5 The extent and characteristics of child poverty Quantitative: Descrip-tive and multivariate analysis Bangladesh Household Income and Expenditure Survey—2000 Chapter 6 Definition, reasons and impacts of child poverty, as seen from the perceptions of its principal victims Qualitative: grounded theory approach Data from four focus group interviews Chapter 7 Child poverty’s extent and reasons for the differences and changes between Bangla-desh and China over time Quantitative: Decomposition framework Bangladesh Household Surveys HES 1995/96 and HIES 2000 ,and CHIP (China Household Income Project) Survey—1988, 1995 and 2002 Chapter 8 What needs to be done for child poverty reduction Qualitative: Secondary data analysis and “ground-ed theory approach” Data from: secondary sources and (five) focus group interviews Chapter 9 Wrapping up the dissertation My final thoughts to ending the dissertation My own thoughts based on the material presented in the dissertation. 7to highlight their specific characteristics. In this section, different approaches to identifying and quantifying poverty, including their advantages, disadvantages and implementation are discussed. Thereafter, staying in poverty and intergenerational transmission of poverty, child poverty as a main focus in poverty discourse in Bangladesh, are conversed prior to a discussion on my ap- proach used in the dissertation. In the beginning of Chapter 3, I intend to provide brief information on some indicators for Bangladesh and some neighbouring countries from the latest available human development and human poverty index. Then, the text embarks on contextual information and contains some background information about Bangladesh in association with child poverty. The discussion is based on the data from secondary sources. Furthermore, the discussion in Chapter 3 on some South Asian countries including Ban- gladesh and China has inspired me to proceed on conducting a comparative study to look at child poverty across countries. An explanation for this is that different chapters point to the extent, characteristics and experience of child poverty in a single country (Bangladesh); what remains is to provide a sense of the issues of why child poverty differs across countries and over time. Therefore, I include Chapter 7 to shed further light on child poverty’s reasons for changes over time as well as differences from a cross country perspective in terms of issues related to growth, income inequality and poverty. To see how this works, two countries–China and Ban- gladesh–were selected in the beginning because of convenience. Co-researchers and I have had access to large sample survey from the both countries under study. On top of that, the greater share of population being children is a similarity between both countries. Also, during the period of study these two countries experienced rapid economic growth (although it is higher in China). These were the reasons for choosing these two countries to study the development of child pov- erty. I then turn to Chapter 4 that contains brief description of different techniques that focus on both quantitative and qualitative aspects of research methodology used in this dissertation. This chapter holds an account of data and approaches employed for its analysis. I offer some argu- ments for employing quantitative and qualitative data as complementary. 8The data analysis for the quantitative studies is done by multivariate analysis and decom- position framework. The subsequent sections in Chapter 4 bring in brief discussion on qualitative data that are drawn from focus group discussion (collected during 2005-2006). As will be seen in the later part of this dissertation, I took advantage of using a grounded theory approach in study- ing child poverty qualitatively. Therefore, a concise description is also included in this chapter on the data collection process in qualitative studies and how these were analysed. Chapter 5 describes the extent and characteristics of child poverty in Bangladesh using microdata, which were not previously available from the Household Income and Expenditure Survey of 2000 conducted by the Bangladesh Bureau of Statistics. Additionally, it includes mul- tivariate analysis to see the effect statistically of a number of variables on child poverty and esti- mate the probability of a child being poor. This chapter presents a quantitative study that addresses the first research question: what are the extent and characteristics of child poverty in Bangladesh. The household is often used as income receiving unit, and the individual (child in most cases) as the unit of analysis are being considered. Using an income-based poverty line, a child in poverty is defined as one living in a household with a disposable income lower than $1 PPP in 1985 prices. The poverty line is calculated at Tk. 500.16 per capita per month estimated in terms of PPP in 1985 prices for the Bangladeshi Taka. An upper poverty line, $2 PPP, calculated at Tk. 1000.32 by doubling the $1 per day poverty line is used to look at the robustness of the results. Chapter 6 is based on a qualitative study to provide comprehensive and in-depth knowl- edge about child poverty from its principal victims’ perspectives and experiences. To complement the survey data based fifth chapter, a study of the views of principal victims in Bangladesh on the issue of child poverty is presented in this chapter. The second research question of the disserta- tion—what is child poverty, and why and how do children experience it?—is dealt with in this chapter. The views of victims were collected through focus group interviews. Snowball sampling was used for selecting participants, and a grounded theory approach was used for the data analy- 9sis. In some cases, child poverty—its conceptualization, causes and impacts—is not very different to child poverty in other parts of the world, as is evident in a number of studies. The data form a pattern in which (the concept of) child poverty is characterized by multidimensional deprivation and parental incapability to provide for children. The final Chapter 7, based on a research article (see Begum, Deng and Gustafsson, 2012) co-authored with Björn Gustafsson and Deng Quheng, is a comparative study between Bangla- desh and China based on large survey data, collected at different points in time, to examine child poverty in a global context from Asian cases. A description of child poverty and reasons for the differences of changes across countries over time has been examined in this chapter. It deals with the fourth research question of the dissertation: to what extent does child poverty differ across Bangladesh and China? What are the reasons for the differences and changes over time across the countries? In this chapter, child poverty is portrayed and compared across time and across the two countries. In the context of child poverty, reasons for: a) differences across the two countries, and b) changes over time during periods of rapid economic growth, in both countries are studied. Poverty comparisons are made using a decomposition framework by which poverty differences are attributed to differences in mean child income, demographic differences and differences in the distribution of income. In Chapter 8, I continue presenting focus group studies considering what could be done to alleviate child poverty in Bangladesh. Secondary data on public policy and poverty reduction programs and empirical data are used to analyze the gap between government initiatives and what needs to be done, which conform to the research question 4 “What is needed to reduce child poverty in Bangladesh according to previous research, poor children and mothers”. A grounded theory approach is used for the analysis of empirical data collected from five focus groups held in Dhaka, Bangladesh. Snowball sampling was used to select the participants. Policy issues were conversed in this chapter to draw attention to the plural voices that are in need of being heard for appropriate policy measures to be executed successfully to combat child poverty. In this context, the third research question of this dissertation—what is needed to be done for child poverty reduc- 10 tion in Bangladesh from the perceptions of priciplal victims of child poverty?—was dealt with. Chapter 9 starts with introductory restatement about the overall issue that is being re- searched. In this chapter concluding remarks are made, which discuss the findings more broadly. This chapter also attempts to point out where I find the work exemplary or deficient in its knowl- edge, judgments, or organization. It gives implications of findings, makes recommendations, sug- gests fields of future research etc. This chapter draws attention to using mixed method (both quantitative and qualitative), which made use of large sample survey data from Bangladesh and China and qualitative data from focus group interviews conducted in Bangladesh as well. As an aspect of another move, main findings from empirical studies (in chapters 5-8) are focused until the section “policy recommen- dation” begins. For child poverty monitoring, policy recommendation in this chapter calls for producing and implementing an appropriate official poverty line. In addition to this, the requirement for the more specific and central attention to child poverty reduction policies are suggested to yield to interrupt intergenerational poverty transmission. In this regard, particular attention is drawn to eliminate corruption or mistrust about delivering public service for the poor etc. that are needed to the successful implementation of such child poverty reduction policies. The next aspect dis- cussed is poverty and child poverty through an analytical review on the contributions made in various literatures on poverty. 11 Chapter 2 Poverty and child poverty 2.1. Defining poverty and child poverty In this chapter, an endeavour is made to give a picture of an overview of the contributions in rela- tion with poverty analysis. This is done by including a review of the existing literature on poverty and child poverty from a social science or socio-economic perspective. A long-standing body of literature is available on poverty analysis, in which one needs to construct a poverty profile (about who the poor are). Defining poverty is the starting point in this process, which is found to be very problematic. One encounters a number of aspects while iden- tifying the poor, such as: objective versus subjective (about how to deal with value judgments); the choice of indicators; relative versus absolute; and unit of analysis (Corak, 2006a; Hanmer, Pyatt, & White, 1999; Roelen et al., 2009). A brief discussion is on these issues presented in the following sections. The first conceptual issue in defining poverty or child poverty is selecting an approach either objective or subjective. One approach is to let experts define the poverty line, while another to ask the population on their opinion. The latter is what is made in the SPL (Subjective Poverty Line) method, which I have not used but I nevertheless asked people. For research question 1 and 3 I have used definitions that come from the World Bank’s definitions for poverty lines. For research question 2 and 4 I analysed the opinions of principal victims of child poverty. My ap- proach is a mixed one–which one could claim as strength. The application of a variety of objective methods is a common tradition of setting the poverty line (Pradhan & Ravallion, 2000). Typically, an objectively drawn poverty line for identifying the poor based on income or expenditure is widespread in the literature of developing countries. Subjective approach, in contrast, is used in parts of the developed countries’ literature on poverty. Subjective poverty lines have been based on answers to for example the “Minimum 12 Income Questions” (MIQ). Within this subjective approach, several questions, in association with income satisfaction and adequacy in terms of monetary amount, can be asked. People (respon- dents) consider this to be the minimum necessity for supporting their households (See Gustafsson, Li, & Sato, 2006; Pradhan & Ravallion, 2000; Ravallion, 2010). Based on the responses to this “Minimum Income Questions (MIQ)”, a subjective poverty line can be developed. Such a subjec- tive approach is used in the literature on some developing countries. The poverty line in this approach is extracted from the answers, which is derived by ask- ing the public. In such a case the answers and questions might have different wordings; therefore some argue them as tempting (Gustafsson, 1995, p. 368). Pradhan & Ravallion (2000) articulate also their doubts in getting sensible answers to the usual MIQ in most developing countries. Choosing subjective or objective approach is, as stated in Hanmer et al. (1999): “one of both the weight attached to the various dimensions of poverty and the selection of a threshold”. The second issue is the selection of indicators. There are, as discussed in Hanmer et al. (1999), various dimensions of poverty such as: basic needs, income/consumption, assets (physi- cal capital, common property, private, human capital), human rights (dignity/autonomy, political freedom and security, and equality). Among dimensions of poverty, income-poverty is just one of several. The importance of these dimensions differs across different social groups. The choice of indicators appears to depend mostly on research questions, assumptions and value judgments of researchers or experts in the relevant fields. The third conceptual issue is whether to use an absolute or relative definition of poverty. Literature of poverty analysis on developing countries puts emphasis on absolute or extrem pov- erty. Generally, absolute poverty indicates to not having enough food to eat or not enjoying good health (Deaton, 2006). Bourguignon (2004) defines absolute poverty in reference to a poverty line that has a fixed purchasing power determined so as to cover essential physical and social needs. Here, the influential role of the World Bank and UNDP in defining and measuring absolute pov- erty is evident. The most commonly used definition of poverty is the absolute poverty line set by the World Bank. Since 1990, the World Bank has chosen instead to measure global poverty by the standards of what poverty means in poor countries, which gave the “$1/day” (recently revised by 13 the Bank at $1.25 a day PPP US$) line. From this perspective, poverty and child poverty are being defined and measured in absolute terms. In contrast to developing countries, one very common tradition of literature on developed countries is defining and measuring poverty in relative terms due to its complexities, for instance, related to the living standard of a particular society. As discussed in Ravallion (2010), a typical definition of relative poverty entails that poverty line has higher purchasing power in contrast to the poverty line in absolute poverty that has fixed purchasing power. A relative poverty defini- tion views persons with a household equivalent income less than for example 50 or 60 percent of median household equivalent income in the country where they live. A similar tradition has also been observed among child poverty research on the devel- oped countries, especially in Europe (see for example, Bradbury, 1999; Bradbury et al., 2001). Deaton (2006) argues that the concept of poverty as the inability to participate in society leads to the concept of relative poverty. Relative poverty is considered as having less in a particular dimension compared to others in a society or country under study. Relative measures have also some major drawbacks. Besides, easy applicability to a wide range of datasets, make this measure well established in the EU countries to determine the extent of poverty in individual countries. Despite having the advantages of country-wide measurements, relative measures contain some major shortcomings in global use. For example, they do not consider the general level of income in the country. All these conceptual issues discussed above, at least, are very important prior to defining or identifying who the poor are. The unit of analysis is also very important to consider children poor or not. Some theoretical and practical consideration is necessary to select whether the child or the household would be the unit of analysis (Corak, 2006a; Roelen et al., 2009). If child is taken as the unit of analysis, particularly for a child specific research, then the problem arises in getting child specific information. The availability of this kind of data in household survey is less than the household level, which is a problem in developing countries in particular. These choices rely on theoretical arguments, scientific judgment, and, more importantly, engagement of a level of value judgments as well (Corak, 2006a; Roelen et al., 2009). 14 2.2. Concepts and terms Different terms, such as social exclusion, deprivation, vulnerability, economic development and inequality are increasingly used in relation to poverty. These terms are associated with but have different meanings and contexts from poverty. Constructions of the concept of poverty by re- searchers, policy advisors and policy makers vary due to disciplinary background and ideologi- cal values. They also vary over time and space due to differences in the political, economic and cultural conditions of the contexts in question. For example, vulnerability is for many a forward looking concept and is generally observed at micro level. On the other hand, the distribution of attributes, such as income or consumption, across the whole population is the main focus of inequality. Gini coefficient6 is the most frequently used measure of inequality and measures for the income distribution of a whole region or country. Social exclusion is another concept that has become increasingly important in the litera- ture. Although poverty is one aspect of social exclusion for some authors, by the mid-1990s “so- cial exclusion” had almost taken over from “poverty” as the key term in much European discus- sions (Leisering & Leibfried, 1999). Social exclusion stands for the denial of access to rights and full participation in the society because of a number of factors both for children and adults. This has been echoed by different organizations’ studies and reports that deal with children in poverty (see UN, 2009). They emphasize the fact that the existence of discrimination in different forms in children’s everyday life have crucial impacts. For example, children’s experience of discrimina- tion in everyday life due to their age, gender, class and social group might be very common, which makes them feel excluded. However, the use of the term social exclusion as a substitute for the term “poverty” has been criticized by researchers as being a “fashionable term” in the new discourse of social prob- lems (see Leisering & Leibfried, 1999). T. Atkinson, Cantillon, Marlier, and B. Nolan (2002), and Leisering and Leibfried (1999) articulate in their view on exclusion that the concept does not necessarily, or not only, refer to poverty. For example, long term unemployment is closely linked 6 The coefficient varies between 0 and 1. 0 is a sign of complete equality (resources equally distributed across the population) and 1, indicates complete inequality (resources are concentrated on one area). Gini coefficient can be easily represented by the area between the Lorenz curve and the line of equality in graphs. 15 with social exclusion. T. Atkinson et al. (2002) recommended a range of indicators of social inclu- sion in which financial poverty is one. Thus, exclusion may be related to a number of factors and the common vague use of the term covers a wide range of issues related with poverty, deprivation, unemployment and discrimination. In addition, the concepts of poverty and child poverty have been referred to as depriva- tion of wellbeing in different literatures (Gordon et al., 2003; Minujin et al., 2006; Sen, 1999; Townsend, 1979). Deprivation, in the context of child poverty means, according to Christian Children Fund(CCF) (as cited in Minujin et al., 2006, p. 486): “a lack of material conditions and services generally held to be essential to the development of children’s full potential; the result of unjust processes through which children’s dignity, voice and rights are denied, or their existence threatened”. The notion of deprivation, as described in UNICEF (2004), focuses on the circumstances that surround children, casting poverty as an attribute of the environment they live and grow in. Although poverty and deprivation are highly interrelated, studies on deprivation and pov- erty are in agreement that there are differences in meanings between these two concepts. For instance, Gordon et al. (2003, p. 6) describe that the concept of deprivation covers a wide range of circumstances, irrespective of income, which are experienced by the poor people. According to these authors, the concept of poverty indicates the scarcity of income or other resources which make these circumstances highly likely or bound to occur. The authors, in a discussion on “mea- suring child poverty in developing countries” conceptualized deprivation as “a continuum rang- ing from no deprivation, through mild, moderate, and severe deprivation to extreme deprivation at the end of the scale” (Gordon et al., 2003, p. 6). Vulnerability is another term that is used in poverty discourse. While poverty has asso- ciation with vulnerability, there are differences in meaning of these two concepts. In this context, vulnerability is typically defined as the probability or risk of being in poverty today or to fall into deeper poverty in the future for the individuals/households. A household would be considered as vulnerable to poverty if it is likely for it to be poor in the future. UN (2009, p. 9-10) portrays the concept of vulnerability as: “the likelihood that people will fall into poverty owing to shocks to 16 the economic system or personal mishaps”. Vulnerability is thus a reflection of economic insecu- rity. As discussed by this report, although poor people are usually among the most vulnerable to being pushed into deeper poverty when faced with mishaps, not all vulnerable people are poor. Many people not currently living in poverty face a high risk of becoming poor in a shock situa- tion. In many cases vulnerability has association with the effects of “shocks” such as a drought, a financial crisis, job loss or the major illness of a family member (Haughton & Khandker, 2009; UN, 2009). In most cases, the poor are more vulnerable to the above stated shock situations. Vul- nerability is a key dimension of well-being since it affects individuals’ behaviour in terms of investment, production patterns, and coping strategies, and in terms of the perceptions of their own situations. It is being observed at the micro/meso level (Haughton & Khandker, 2009). For appropriate forward-looking antipoverty interventions, it would be important for the policy mak- ers to identify those who are expected to be poor in the future (Haughton & Khandker, 2009). From this point of view, a household is vulnerable to poverty if it is likely to be poor in the future. Vulnerability of children comes from an inability to cope with existing or probable threats to them in their surroundings. Harper (2004, p. 3) describes that “a gain made due to a period of employment, a good agricultural year, or a family inheritance can quickly be lost due to vulnera- bility”. In such a state of crisis, children are equally, even to a greater extent, vulnerable to poverty than adults. Harper (2004, p. 3) describes more clearly: “Children are equally, if not more, vulner- able to shock than adults. If their education or nutrition is interrupted at a critical time, they may not be able to regain that loss”. Vulnerability to child poverty is approached from many sources such as a drought and a financial crisis (loss of job, or medical expenses of the households), a local flood, indebtedness, widowhood, and other shocks (Davis, 2009; Harper, 2004). These are some of the threats to the surroundings of the poor children. Views on vulner- ability indicate its relation to the effects of shocks. For example, Davis (2009), from a life history interview to investigate poverty dynamics in Bangladesh, found that in addition to being more exposed and less likely to have insurance, poor people are more vulnerable in a crisis situation. Households’ coping strategies in such a situation might be a reduction of food intake or school 17 drop out of one or more children according to this study. In such a situation children are more vulnerable than the adults because of its severe consequences during their adulthood to carry and pass poverty onto the next generations resulting from their own ill health and low education. Economic development, a precondition for welfare, is another term that is used in rela- tion to poverty. There is plenty of literature discussing to what extent economic development reduces poverty, and the relation between economic development and inequality (Dollar & Kraay, 2002).The literature also deals with which aspects of economic development that affect poverty, or which types of economic development that affect poverty. In terms of the results, the research literature is mixed, but primarily finds that economic development is not sufficient for wellbeing but a prerequisite for improving material conditions. In other words, studies argue that economic development does not under all conditions reduce poverty but is essential for the poverty reduc- tion (see Thurlow & Wobst, 2006). While material goods are essential for living, prosperity and social integration, economic growth itself generates certain casualties such as people might be vulnerable to market fluctua- tions and some might lose their earning capacities due to the technological development etc. More importantly, the advantages of economic development must not be equally shared in a society where inequality is a major concern. OECD (2008) in a report on rich countries claims: “how much inequality there is in a society is not determined randomly, nor is it beyond the power of governments to change”. Some authors call for pro-poor growth. That is, the growth process should have a profile that benefits the poor more than others. So the issue is not growth, but how growth is distributed in the population. Chapter 7, in fact, deals with this. On the other hand, inequality is a further important term that has been held responsible by studies for the low impact of economic growth on poverty in very low-income countries (Hulme & Shepherd, 2003, p. 413). Inequality, generally, represents the view that articulates the disadvan- tages of a population group relative to others in a society in terms of the distribution of income or consumption and other attributes such as sex. The difference between the terms poverty and inequality is well established. Poverty and inequality use separate indicators, measurements and conceptual definitions and implications. 18 As these terms are typically defined, Ravallion (2010) infers that poverty is about ab- solute levels of living. In this sense, the concept of poverty provides an impression about how many people cannot attain certain pre-determined consumption needs. In contrast, Ravallion puts inequality, in plain words, as about the disparities in levels of living. By this the author means: “how much more is held by rich people than poor people”. Coudouel, Hentschel, and Wodon (2002) define inequality as a broader concept than poverty in the sense that it is about the entire population, not only those below certain poverty lines. From this perspective, inequality in in- come distribution points to the macro level. While all these concepts stated above are used in relation to poverty and child poverty, separate indicators, measurements and conceptual definitions and implications are in use in this context. Among these terms, economic development and inequality of income distribution points to the macro level or national level whereas vulnerability is used when poverty analysis concerns individual/household, in other words, at micro or meso level. In contrast, the analysis of poverty or child poverty might refer to all levels: micro, meso or macro. 2.3. Approaches to poverty Poverty is a complicated concept that is highly dependent on how it is defined and measured (see Saunders, Bradshaw, & Hirst, 2002) and theories about poverty have become increasingly sophisticated over the last 20 years (Matin & Hulme, 2003, p. 648). There are different questions on poverty, and different theories have relevance accordingly. Thus, different ways of interpreting and defining poverty exist and academics debates centre around the controversies on the issues of how to count and identify the “poor” or differentiate them from the “non- poor” and to find out different levels and causes of poverty. Therefore, it is important to distinguish among theories that are applicable to different levels. Some theories are used to identify the poor, for example, materialistic conceptualization of poverty or income-based approach refers to how to identify the poor but it does not explain pov- erty. Others focus on causes of poverty at the individual level and also, there are different theories 19 that attempt to explain why mass poverty exists and varies between countries or over time. Dif- ferent approaches to poverty analysis are the result of theoretical considerations, research ques- tions and value judgments (Ravallion, 1994). With this in view then the discussion proceeds in the following manner. As discussed in preceding section, the objective and subjective approaches are two dif- ferent ways of dealing with poverty lines. The World Bank chooses the objective approach with income or expenditure, which is labelled as money-metric approach to the determination of living standards and set an income-poverty line to distinguish the poor from the non-poor. However, this approach has critiques (see Hanmer et al., 1999) to have limitations such as adjusting for spatial price variation and for household size and composition, and analyzing the dynamics of poverty. The critics of objective approach put their doubts by questioning the objectivity on the issue of using “basic needs” in poverty measurement. As a result, the subjective poverty line was devel- oped in the 1970s. In the subjective approach, the dimensions of poverty analysis are chosen either by the society under study or the poor themselves. The Dutch research started to work on subjective poverty measures, which extended to the academic circles onward (see Goedhart, Halberstadt, Kapteyn, & Van Praag, 1977; Van Praag, Goedhart, & Kapteyn, 1980). For example, Gustafsson, Li, and Sato (2006) have done work on this approach for China. In addition there are also other attempts. The proponents of this approach argue (see Nunes, 2008): “if value judgments affect measurements, then the methods are not objective and therefore who is making such value judg- ments matter; and, most likely they are the statisticians and researchers”. From this perspective, all poverty lines turn up subjective since these are usually being developed by, and subject to the value judgement of, experts in the relevant fields. The results from this dissertation (on Bangla- desh) have relevance for some issues in the literature on poverty (for example, the concept of child poverty). Materialistic conceptualization of poverty takes the monetary approach in measuring poverty in which poverty is defined by the single variable income. Such an approach compares an individual’s income or consumption with some defined threshold, below which individuals 20 are being considered as poor (Minujin et al., 2006). This traditional one-dimensional approach to poverty is often called monetary approach to poverty. In many developing countries, money is only one out of several income sources and poverty is assessed based on income. From this per- spective, naming this approach as “monetary approach” in literary sense is to a great extent mis- leading and income-based approach seems to be more appropriate among the other alternative ways of addressing it . In this dissertation, an income-based approach has been used in Chapter 5 and Chapter 7 to identify the poor children and their household. Poverty exists, according to this approach, when some people in the society have so little income that they cannot satisfy socially defined basic needs (Kakwani, 2006). Although the income-based approach to identifying and measuring both poverty and child poverty is well established and widely used, it has been challenged by other multi-disciplinary approaches (Hulme & Shepherd, 2003; Townsend, 1979), for instance, the human rights-based approach, the basic needs approach and the capability approach. Kakwani (2006) argues that lack of income is not the only kind of deprivation people may suffer. Indeed, people can suffer acute deprivation in many aspects of life, beyond those defined as basic needs, even if they possess ad- equate command over commodities for example, ill health or lack of education and so on. The political economy approach views poverty “as a product of certain economic and social process that is intrinsic to given social systems” (Silva & Athukorala, 1996, p. 70). This perspective assumes that remaining in poverty is not due to individual or personal quality but because society denies their (poor) access to the legitimate share of benefits that should be made available to them (p. 70). The promoters of this approach pay lots of attention to the historical context arguing that the evolution of current poverty situation in, for example, South Asia was grounded during the colonial period. Unlike the neoclassical approach, the political economy approach views growing class differentiation as a negative outcome of market-led development. A human right-based approach to poverty analysis attempts to integrate the concepts of human rights. For example, Minujin et al. (2006) provide a discussion on this approach in defin- ing the concept of child poverty. They especially emphasize on the endeavour of this approach to integrate human rights concepts, values and language into the poverty reduction dialogue. UNI- 21 CEF uses this approach and the Convention on the Rights of the Child for a definition of child poverty. In its working definition, UNICEF defines child poverty as the deprivation of essential material, spiritual and emotional resources, and social supports and services that are essential to ensure children’s well-being (see Minujin et al., 2006; UNICEF, 2004). For identifying the poor, the human rights-based approach looks at the constitutive rights, which are taken into consideration for measuring and analyzing poverty, and without which a per- son is considered poor (see Minujin et al., 2006, p. 485). One of the limitations of this approach is that constitutive rights might differ from one country to another. However, UNICEF is one of the advocates of this approach for defining and measuring child poverty. The Office of the High Commissioner on Human Rights (OHCHR) developed a common set of rights based on empiri- cal observation that apply to most countries (as cited in Minujin et al., 2006). These are: being adequately nourished; being able to avoid preventable morbidity and premature mortality; being adequately sheltered; having basic education; being able to appear in public without shame; be- ing able to earn a livelihood; and taking part in the life of a community. Sen’s (1981) entitlement approach to famine analysis has drawn decades’ long attention of poverty researchers. In Sen’s words: the entitlement approach provides a particular focus for the analysis of famine. The entitle- ment of a person stands for the set of different alternative commodity bundles that the person can acquire through the use of various legal channels of acquirement open to someone in his position (Sen, 1995, p. 52). The debate on Sen’s approach range from a constructive argument by Osmani (1995) to criticism, for example, by P. Nolan (1993). Most recently, Dowlah (2006) criticizes Sen for not giving a full empirical account of the socio-political situation, mainly the widespread corrup- tion in the analysis of Bangladesh’s famine in 1974. In contrast, Sohlberg (2006, p. 1) counters Dowlah’s criticism by arguing that: “his reading of Sen implies an empirical standpoint that is in general alien to Sen’s entitlement approach”. Although the debate is going on, it is vital for households including their children in a poverty situation to have entitlement to commodities to fulfil at least basic needs. In child poverty context, the parental or household entitlement demands more attention and can be seen in relation to human right based approach. In this regard, UNICEF 22 (2006, p. 10) emphasizes that “the right to an adequate nourishment, health care, housing and quality education must be seen as an entitlement and as a policy priority”. The advocates of capability deprivation approach argue that poverty is more than lack of income. For example, low level of education or poor health is not necessarily related with low income but points to a “capability deprivation”. For example, Sen (1999, p. 87) denotes: “poverty must be seen as the deprivation of basic capabilities rather than merely as lowness of incomes, which is the standard criterion of identification of poverty”. Sen introduced the concept of capa- bilities, by which he means that the basic needs of different kinds must be met for someone to be able to participate fully in society (Sen, 1992). His perspective of capability7 in his words: “does not involve any denial of the sensible view that low income is clearly one of the major causes of poverty, since lack of income can be a principal reason for a person’s capability deprivation”. The literature (for example, Chant, 2006) regards Sen’s capability deprivation approach as a ground breaking approach that has changed the way poverty was previously defined and measured. Thus it is well established that poverty is “not just a matter of money”, although it is most commonly measured in terms of income (see B. Nolan & Whelan, 2007, p. 146). Sen (1987) argues that well-being comes from a capability to function in the society and thus poverty arises when people lack key capabilities, have inadequate income or education, or poor health, or insecurity, or low self-confidence, or a sense of powerlessness, or the absence of rights such as freedom of speech. However, there is no universal agreement on what the key or basic capabilities are. In line with Sen, some argue that if a person is not able to be well-nourished, adequately clothed and sheltered, and not able to avoid preventable morbidity, then he or she can be classified as deprived of basic capabilities (for example, Kakwani, 2006). 7 Sen’s (1999, p. 87) opinions in favour of capability approach are: 1) poverty can be sensibly defined in terms of capability deprivation; the approach concentrates on deprivations that are intrinsically important (unlike low income, which is only instrumentally significant). 2) there are influences on capability deprivation - and thus on real poverty - other than lowness of income (income is not the only instrument in generating capabilities). 3) the instrumental relation between low income and capability is variable between different families and different individuals (the impact of income on capabilities is contingent and conditional). 23 Sen has articulated the broadest approach of capability to well-being (and poverty). He argues that wellbeing is fundamentally multidimensional from the perspective of “capabilities” and “functionings”. As discussed by Sen, functionings deal with what a person can ultimately do and capabilities indicate the freedom that a person enjoys in terms of functionings (see Sen, 1987, 1992). In view of this perspective, “functionings” is about achieved well-being and “capabili- ties” is about the freedom to achieve well-being. Sen (1992) acknowledges that the differences between functionings and capabilities, while it could be useful in an analytical sense, could also be quite muddy, since actual functionings also influence capabilities (Redmond, 2008). The capa- bility approach closely estimates functionings by attributes such as literacy, life expectancy, etc. and not by income. Viewed in this way, poverty is related to the human rights based approach and multi- dimensional approach simultaneously. Gordon et al. (2003) define severe deprivation among children along eight dimensions of wellbeing––food, water, sanitation, health, shelter, educa- tion, information and access to basic services. Those capabilities that relate to health, education, shelter, clothing, nutrition and clean water can be regarded as capabilities that can be called as basic (Kakwani, 2006). Studies indicate that capability deprivation is associated with a number of factors and poverty is lack of resources to achieve basic capabilities. Thus, defining poverty from capability approach is not independent of income because one’s capability to function is also related with income and wealth. What is more, the school of thoughts within the multidimensional approach (especial- ly World Bank, UNDP and UNICEF) argues that poverty has multidimensional effects on both adults and children. The traditional income or consumption-centred conception of poverty has been criticised, among other issues, for its negligence of this fact (see Lyytikäinen, Jones, Huttly, & Abramsky, 2006). In multidimensional approach, poverty is being considered as a multidi- mensional phenomenon and there is significant and growing literature that emphasizes the mul- tidimensional nature of poverty (Bourguignon & Chakravarty, 2003; Gordon et al., 2003; Sen, 1987; UNICEF, 2006). Among others, UN (2009) echoed that poverty extends further than the economic ground to cover factors such as the inability to participate in social and political life. 24 Drawing attention to the complexity and multidimensional nature of poverty is not the result of the recent past but of the centuries long ago. For instance, these were initiated to bring- ing into light in the context of developed countries by Henry Mayhew (1851), Charles Booth (1892), Seebohm Rowntree (1901). They also called for taking into account social conditions, diet and health as well as income in assessing living standards to understand poverty (see also Tomlinson, Walker, & Williams, 2008). Before them, Adam Smith (1776) indicated shame and stigma as being inherent components of poverty, which was motivated and taken further by Sen in a global context. Citing Adam Smith (1776), Sen (1999) argues that the shame that results from poor people’s inability to realize basic capabilities is consistent with the society in which they live. It is universal and absolute and is evident in all societies, regardless of the level of economic development. The UNDP’s human development approach is considered as an alternative to the in- come-based approach to poverty analysis and has its theoretical underpinnings in the work of Sen (1999). This approach used two indices for poverty measurement until 2009. These are: the Human Development Index (HDI) and the Human Poverty Index (HPI). In one hand, the Human Development Index (HDI) is an example of multidimensional measure of well-being in terms of functioning achievements; this index presents the country level functioning achievements in terms of the attributes life expectancy8, educational attainment rate9, and a decent standard of liv- ing10 (Bourguignon & Chakravarty, 2003; HDR, 2009). The index is used to rank countries in terms of “human development” and calculated separately for “very high human development” (e.g., Norway, Australia, Canada and Sweden), “high human development” (e.g., Bahamas, Lithuania, Chile and Argentina) “medium human development” (e.g., China, Sri Lanka, Maldives and India), and “low human development” (e.g., Pakistan, Bangladesh, Myanmar and Nepal) countries. A brief discussion on HDI (and also HPI- 1) presented in Chapter 3 gives an idea about medium human development (e.g., China, Sri Lan- ka, Maldives and India), and low human development (e.g., Pakistan , Bangladesh, Myanmar and 8 Measured by life expectancy at birth. 9 Measured by adult literacy rate and the combined gross enrolment ratio for primary, secondary and tertiary schools. 10 Measured by GDP per capita in purchasing power parity (PPP US$). 25 Nepal) countries. On the other hand, HPI (HPI-1 and HPI-2), is a wider measure, which arrests the many dimensions of human poverty. Human Poverty Index (HPI) is obtained differently—one (HPI- 1) is used for developing and transition countries and another (HPI-2) is applied for a group of selected high-income OECD countries (rich countries). As described by Human Development Report ( HDR (2009)), HPI-1 is used to measure human deprivation in the same aspects of human development as the HDI for developing and transition countries. These aspects are: a long and healthy life, access to knowledge and a decent standard of living. In addition to the similar aspects of HPI-1), HPI-2includes “social exclusion”11 as the fourth dimension (see HDR, 2009).Yet, HDI does not include all important aspects that are associated with human development (such as gen- der or income inequality, respect for human rights and political freedoms). Human Development Report uses other indices such as gender-related Development In- dex (GDI), Gender Empowerment Measure (GEM) etc. that try to do this. An essential factor here is that Human Poverty Index focuses on the most deprived in multiple dimensions of poverty. While applying the multidimensional approach in recent times, the UNDP has brought some ma- jor changes in its indices (see HDR, 2010, 2011). Yet, it would be interesting to look at how the countries are ranked by HDI, and how some countries’ progress appears. These issues are brought into discussion and presented in Tables 2, 3, 4 and Figure 1 in chapter 3. The HDI measures, on the one hand, the average progress of a country in human development. On the other hand, the Human Poverty Index (HPI-1) focuses on the proportion of people below certain threshold levels in each of the dimensions of the HDI. Instead of measuring poverty by income, three measures of deprivation are used to obtain the HPI. Rankings of countries by HDI therefore must not conform to rankings of countries according to HPI-1. Furthermore, when ranking countries derived from the extent of poverty, a ranking based on the World Poverty Line (by the World Bank) must not agree with a ranking anchored in HPI-1 (by the UNDP). Some authors (Hulme & Shepherd, 2003, p. 421) argue that the work of Sen (1981, 1999) has been influential and encouraging among the analysts to conceptualize poverty in a multidi- 11 This is represented by the rate of long term unemployment. 26 mensional way. Despite critiques, particularly, in relation with the difficulties in measuring multi- dimensional poverty, it is argued that the theoretical and conceptual contributions of the human development approach’s are important. Lyytikäinen et al. (2006) argue quoting from several stud- ies that broadening the definition of poverty in this approach changes significantly our thinking about strategies to reduce poverty. Each of the approaches is different in terms of concepts, definitions and methods. No approach can fully replace another approach. For example the political economy approach is not an alternative to the approach of materialistic conceptualization of poverty. This is evident in a number of approaches that exist in poverty analysis literature such as the approach of materialistic conceptualization of poverty, human right based approach, entitlement approach, capability and multidimensional approach. Amongst these, the most conventional and widely accepted one is materialistic conceptualization of poverty, which relies on a single indicator, especially income. Income poverty is only one aspect of poverty and the income-based approach has the limitation in addressing other aspects associated with poverty, such as insufficient outcomes with respect to health, nutrition and literacy, insecurity and powerlessness (see Atkinson et al., 2002; Sen, 1999). The application of an approach requires not only the choice of how and where to set the line, but how to deal with problems of price variation and of family size and composition (see, for example, Ravallion & Bidani, 1994). One important dimension in the poverty assessment lit- erature is who decides on the poverty line. In much writing, it has been experts. But to what extent can one ask people about what is meant by poverty? Therefore, this dissertation combines a well- established method of defining poverty (Chapter 5 and Chapter 7) with attempts to finding out what various groups of people think (Chapter 6 and Chapter 8). Some of the previous studies aim to ask people of their perceptions on poverty. One natural starting point is the work of Townsend (1979). This has thereafter been developed by Mack and Lansley (1985). This study concerns child poverty and intends to add knowledge on the phenomenon in Bangladesh, as a South Asian case. Several terms are used to indicate poverty status of children, for example, child poverty, childhood poverty and children living in poverty. In this dissertation, I use the term “child poverty” to describe the poverty of children. As documented in different 27 studies, all the approaches to identifying and measuring poverty have limitations and advantages. For example, the income based approach is widely used for its simplicity to deal with survey data, even though it addresses only one aspect of poverty, i.e. insufficient income, and not address the other aspects related to poverty such as nutrition, education, lack of health care and so on and in its quantitative analysis, however, the qualitative study has a somewhat broader scope. In Europe, the shift in conceptualization and measuring poverty from the conventional thoughts is evident in literature. These include: the critique of the absolute notion of poverty, ad- vocacy of viewing poverty in relative term, and a number of its dimensions in both monetary and non-monetary terms (Atkinson, 1987; Townsend, 1979, 1985). There is a famous debate with Pe- ter Townsend and Amartya Sen in Oxford Economic Papers (1985) about the extent to which pov- erty is only relative in high income countries. Sen’s controversies (Sen, 1983, 1985; Townsend, 1985) with Townsend were constructive in setting up the legitimate underlying principles for both absolute and relative notions of poverty. The traditional income/ consumption-centred conception of poverty has been criticised, among other issues, for its negligence of the multi-dimensional na- ture of poverty and the importance of public services for wellbeing (see Lyytikäinen et al., 2006). This multidimensional nature of poverty can be measured. Multidimensional poverty or deprivation has mostly been measured at national levels through the Human Development In- dex, Human Poverty index and other indices such as the physical quality of life; measurement at individual or household level is not well developed (Hulme & Shepherd, 2003, p. 407). In this dissertation, I consider some of these aspects in qualitative studies, for example, gender discrimi- nation (Chapter 6) and lack of access to public services (Chapter7). However, this dissertation does not address other important issues such as the dynamics and stability of poverty. The above- mentioned aspects (gender discrimination, lack of access to public services etc.) lead individuals to lower human development resulting from capability deprivation and have bearing upon inter- generational transmission of poverty. Capability variables such as the capability to participate in the society without discrimination may not be easily measured (see Minujin et al., 2006). These complexities of poverty measurements lead poverty analysts to take an income-based approach in defining and measuring poverty. 28 As noted in the preceding sections, an income-based poverty line—International Poverty Line, defined by the World Bank—is used to define poor children in the quantitative study and as an inclusion criterion in the qualitative study of the thesis. My motivation of employing this line is its frequent use in poverty research on developing countries. Yet, it is not without critique (see contributions to Anand, Segal and Stiglitz, 2010). Chapters 5 and Chapter 7 define poverty based on income, and money income is just one component. This line is so far widely used for the developing countries. Applying this poverty line, child poverty has been identified as a state of children’s deprivation of having access to basic human needs and wellbeing while their care givers are incapable of providing for them. The conceptualization of child poverty has been perceived in the qualitative studies of this dissertation as the deprivation of basic capabilities that are related to health, education, shel- ter, clothing, nutrition and clean water. Capability deprivation approach addresses the limitations of income-based approach and take other dimensions of poverty into account that systemati- cally lead to lowering of the development of capability of individuals, particularly, in childhood. This capability deprivation is associated with multidimensional factors and is closely related with human right-based approach. There is no particular approach to child poverty analysis. Child poverty researchers are using different approaches that are commonly used in poverty analysis. Following this, different approaches have been used in this dissertation as complementary ap- proaches for the analysis to child poverty. I attempted to measure income poverty but analyzed the concept of child poverty from different approaches to poverty analysis. 2.4. Measuring poverty Measuring poverty is very important for many reasons: to keep the poor on the agenda, to iden- tify the poor for poverty alleviation interventions (World Bank, 2005), to be able to study the development of poverty and so on. As discussed in Ravallion (1998), three steps are needed to be taken in measuring poverty. These are: to define an indicator of welfare, to establish a minimum acceptable standard of that indicator to identify the poor (the poverty line), and to generate a sum- mary statistics to aggregate the information from the distribution of this indicator relative to the 29 poverty line. The linkage between concept and measurement is important in distinguishing different aspects of research: identifying and counting the poor versus what it means to be poor (Nolan & Whelan, 2007). The standard approach is to establish a poverty line, normally reflecting a mini- mum necessary standard of living to identify who falls below this line (Weiss, 2005). Absolute and relative measures are two basic kinds that are commonly used in the literature on poverty measurement (Bradbury et al., 2001). It should be noted here that “relative” and “absolute” are not only a matter of measurement but they are also different concepts to poverty analysis. Dif- ferent methods have been used in the literature to define absolute poverty lines (see Deaton, 1997, 2006; Ravallion, 1994; Ravallion & Bidani, 1994). Absolute poverty measures are used to define absolute needs irrespective of the society where poverty occurs. These measures utilize the thresholds, making up the poverty lines.12 Some claim that there is no such thing as absolute poverty. The World Bank’s dollar-a-day poverty line13 is motivated by being approximately at the level of absolute poverty line used in poor countries. This line is also known as International Poverty Line. To establish an absolute poverty line, for instance, conversion of a 1985/1993 dollar to local currency is the first step. The next step is to update it to the year of interest using the local Consumer Price Index (CPI) to get the line in terms of prices at the desired survey date (World Bank, 2001; World Bank, 2005). The World Bank plays the most influential role in defining these poverty lines. Using these poverty lines, one attempts to measure the percentage of the population living on less than one US dollar per person per day, or the two dollars per day count. Until recently, International Comparison Program (ICP)14 has provided information on 12 In some cases, the poverty line is a single number. The poverty line for a particular household is derived by multiplying the numbers of household members with this number. Typically, this is done to draw poverty line for low income countries. In other cases, the procedure for deriving poverty line is more complicated, especially for high income countries. In this context, the poverty line for a particular household is defined by a formula or a table. For example, it is typically assumed that a household of two persons need lesser income than two single persons living alone. 13 The dollar-a-day poverty line is rooted in the purchasing power parity (PPP) exchange rates. International Comparison Program (ICP) project produces this. 14 The International Comparison Program (ICP) is a worldwide statistical operation involving some 160 coun tries. It produces internationally comparable price and volume measures for gross domestic product (GDP) and its component expenditures. 30 three rounds of estimates to construct an “average” poverty line: first, in 198515, second in 2000- 200116, and the most recent in 2005 (UN, 2009). In the new and latest round, the dollar-a-day is revised, and raised to $1.25 (per person per day) (see Ravallion, Chen, & Sangraula, 2009; Chen & Ravallion, 2010a). These adjustments are made on the basis of the revision in purchasing power parity (PPP)17 estimates of the ICP’s 2005 round (Himanshu, 2008; UN, 2009). These measures are widely used for the developing countries’ poverty research and are being considered as rep- resentative of the actual poverty line in very poor countries since large shares of the population survive with the bare minimum or less (Chen & Ravallion, 2001, 2004; Deaton, 2006; World Bank, 2005). But, in some countries one can actually survive on a budget that is less than 1 USD, and in other countries it might be higher. The appropriateness of measuring absolute poverty is being questioned and is becoming highly debatable and controversial. Among others, Reddy and Pogge (2002, 2005, 2010), criticized this line in details for be- ing arbitrary and unrelated to a clear conception of poverty and the way purchasing power parity (PPP) is being calculated (see also Gordon et al., 2003). They argue that this measure systemati- cally underestimated the global poverty and suggested alternative methods. Ravallion (2010a), in a reply to Reddy and Pogge, acknowledges the conservativeness of its definition. Nevertheless, he argues that “their claims do not stand up to close inspection and that there are serious concerns about the proposed alternative method of counting the world’s poor”. The most recent round of revised poverty line is also not out of criticism. For instance, Himanshu (2008) and Reddy (2009) pinpoint to the new round with similar criticism. Additionally, it goes after an inconsistent proce- dure in respect to correcting for rural/urban price differences. The debate is continuing similar to the cases of other poverty lines. The arguments sup- port the views that this is not the only, ultimate and perfect poverty line for measuring poverty 15 The poverty line in the first round was set at $1 per person per day when the ICP covered 22 countries (UN, 2009). 16 The revised poverty line in this round was set at $1.08 per person per day. This revision was based on the PPP exchange rates of the ICP’s 1993 round(UN, 2009). 17 PPPs show the ratio of prices in national currencies of the same precisely-defined product in different coun tries. PPPs are both currency converters and spatial price deflators. They are used by the ICP to make the GDPs of different countries comparable. 31 for the poor including child poverty of the developing countries. Despite critiques, in contrast to other poverty lines used for the developing countries, the acceptability of the dollar-a-day poverty line by World Bank among researchers has been well established until an alternative and better poverty line is produced and implemented. A poverty line (absolute) over time is updated only for the changes in overall price levels. It disregards the changes in the composition of baskets of goods or level of the reference income level (Corak, 2006a; Fisher, 1995; Foster, 1998). In contrast, a relative threshold is updated for both. Relative measures are very common in poverty research on developed countries. Unlike the developing countries, the extreme hardship such as starvation is very unusual there. In this context, using absolute poverty line is considered to make very little sense (see Deaton, 2006). Though using an absolute poverty line is very common in the USA. Updating a poverty line in an economy that grows is important. If the poverty line is updated by a consumer price index only, one could, perhaps, say that this is an absolute definition. On the other hand, in the EU, where the discussion is mostly around social exclusion, the poverty lines are usually updated according to the development of the purchasing power of median income of the country where poverty is concerned. Overall distribution of income or consumption of a country is the basis of defining rela- tive poverty lines. These lines could be set at 50 or 60 percent of the country’s mean or median (in much recent work it is actually the median, not the mean) income or consumption (Bradbury et al., 2001; Gustafsson, 1995). These measures have the advantages of conceptual clarity. Nowa- days, EU most often defines the risk of poverty as having income less than 60 per cent of median income. For example, to measure child poverty for monitoring Social Inclusion Program, the European Union uses the relative poverty line as the proportion of children in each country living in households with an equivalent income of less than 60 percent of the national median for that country. Furthermore, in a recent comparison of poverty across rich countries by OECD, such relative measures are used (OECD, 2008). The typical poverty index “the head count ratio” has been used for a longer period by the poverty researchers including Rowntree. Nevertheless, it caught reasonable attraction by Sen 32 (1976) for its limitations, who expresses disapproval of two crude poverty measures, the head count ratio (proportion of persons with incomes less than the poverty line) and the income gap ratio (the gap between the poverty line and average income of the poor, expressed as a proportion of the poverty line). Among the deficiencies of this index are the ideas that the head count ratio expresses the “extent of poverty” by the proportion of poor only. It lacks the considerations of “intensity of pov- erty”—how poor the poor are; and further, it does not take “inequality” into consideration—how poverty is shared among the poor. In his seminal paper, Sen (1976) proposed a more sophisticated index of poverty using an axiomatic approach as alternative; a family of indices introduced by Foster, Greer, and Thorbecke (1984) has been widely used among the poverty researchers. This family of indices (as cited in Gustafsson, 1995, p. 373): uses for each poor unit its normalized poverty gap, which is a number indicating how far below the poverty line income falls on a scale bounded by 0 and (in case of no negative income) 1. Those gaps are raised by a positive parameter before the average is taken and then multiplied with the head count ratio. Higher numbers of the parameter give increas- ing weight to large poverty gaps, and thus indicate greater “poverty aversion”. An advantage of this family of indices is that it is additively decomposable by population and sub- group. The total poverty is then the weighted sum of poverty in mutually exclusive sub-groups. 2.5. Staying in poverty Poverty can be of different kinds, depending on the duration of staying in poverty such as chronic, persistent or temporary. Chronic poverty is different from transitory poverty in terms of moving into and out of poverty over time. Howe and Mckay (2006, p. 198) define chronic poverty as pov- erty that persists over a long period of time, which in different instances may be several years, a generation or several generations. Literature (for example, Hulme & Shepherd, 2003) focuses on at least two distinguished characteristics of chronic poverty that are: its past and perceived future persistence and the likely inability to escape poverty in any reasonable time horizon (Howe & 33 Mckay, 2006, p. 198). Hulme and Shepherd (2003, p. 405) in “conceptualizing chronic poverty” proposed a five-tier categorization for the study of chronic poverty: 1) the “always poor” whose poverty score in each period is below a defined poverty line; 2) the “usually poor” whose mean poverty score over all periods is less than the poverty line but are not poor in every period; 3) the “churning poor” with a mean poverty score around the poverty line but who are poor in some periods but not in others; 4) the “occasionally poor” whose mean poverty score is above the poverty line but have experienced poverty at least one period, and 5) the “never poor” with poverty scores in all periods above the poverty line. The authors categorized further by aggregating the aforementioned into the chronic poor (always poor and usually poor), the transient poor (churning poor and occasion- ally poor) and the non-poor (the never poor). In order to understand chronic poverty, one needs panel data (see Gustafsson, 1995, p. 375) and much analysis of chronic poverty has been based on quantitative data (Baulch & Hod- dinott, 2000; McKay & Lawson, 2003), in particular using panel data sets. Combining qualita- tive and quantitative methods to understand the extent, pattern, and nature of chronic poverty is increasing (Howe & Mckay, 2006). In understanding chronic poverty, multidimensional concep- tualizations are likely to get particular importance (Hulme & Shepherd, 2003, p. 407). As this study indicates, the possibility of escaping poverty is less when individuals are deprived from more dimensions. A possible definition of chronic poverty, given by Moore (2001), is intergenerational transmitted poverty (IGT) and “poverty that spans generations can be seen as both a characteristic and a cause of chronic poverty” (Moore, 2001, p. 4). The empirical evidence (see Hulme & Shep- herd, 2003) shows that staying and growing up in poverty leads to IGT poverty (parents-to-child in most cases); thus it is likely that child poverty and chronic poverty are linked to each other. Therefore, researchers of chronic poverty are increasingly emphasizing on child poverty issues (Hulme & Shepherd, 2003; Moore, 2001) that lead to a transmission of poverty over a life course and between generations. 34 2.6. Intergenerational transmission of poverty In general, child poverty may be defined as the poverty experienced by the children. In such a def- inition of child poverty, emphasis is given on children’s experience of poverty. Nevertheless, one must acknowledge that defining child poverty is not straightforward. It needs more clarification to offer additional insights. The literature, therefore, is shifting its focus towards understanding, defining and measuring child poverty. This is likely due to a mounting concern with its multidi- mensional nature. The need for a multidimensional approach has been recognized increasingly in recent years for poverty analysis. In this perspective, UN General Assembly (UNICEF, 2007, p. 1) provides an insightful definition of child poverty in the following manner: Children living in poverty are deprived of nutrition, water and sanitation facilities, access to basic health-care services, shelter, education, participation and protection, and that while a severe lack of goods and services hurts every human being, it is most threaten- ing and harmful to children, leaving them unable to enjoy their rights, to reach their full potential and to participate as full members of the society. This definition indicates to the deprivation of a number of basic needs and social exclu- sion that limit children’s participation in the society and lead them to severe and complex conse- quences in both the short and long run. Several papers, furthermore, emphasize the consequences of child poverty, which not only encompass the daily uncertainty of basic needs, but also affect all areas of children’s lives—from psychosocial and physical development to educational attain- ment (Bilson & Cox, 2007, p. 37; Hardning & Szukalska, 2000). The majority of these and other related studies conclude that, children living in poor families are more likely in their adulthood to earn less, be less healthy and be unemployed more. This indicates to “Intergenerational Trans- mission (IGT) of Poverty” (from parents to children in most cases), which is increasingly getting importance in the literature (Harper, 2004; Moore, 2001; Sen & Hulme, 2006; B. Smith & Moore, 2006). Such compilation indicates to the possibility of staying poor over life course and transfer- ring poverty between generations. 35 As an aspect of this, the damaging effects of child poverty underpin “intergenerational transmission (IGT) of poverty” process. Being born in and growing up in poverty make threats for the next generation to be poor, while some children might move out of poverty given the op- portunities. However, being born in poverty increases the probability of poverty in adulthood. In a survey of studies among the rich countries, Corak (2006) finds that family economic status is significantly associated with labour market successes of children in adulthood (see also Bradbury et al., 2001). Similarly, the empirical evidence from Peru, Vietnam, Malaysia, Brazil, Sub Saha- ran Africa and India suggests that limited parental resources mean lower investment in children (Quisumbing, 2009). Investigating the causes and consequences of childhood poverty is the central aim in Young Lives Project’s longitudinal study on the developing world: Ethiopia, India, Peru and Viet- nam. Besides, Chronic Poverty Research Centre (CPRC) conducted some research on IGT pov- erty in developing countries (e.g., Bird, 2007; Harper, 2004; Moore, 2001). As documented in these studies, a set of variables characterize child poverty and are responsible for the future labour productivity and intergenerational mobility; for example, less schooling and lower educational attainment, low-income, poor nutrition and inadequate health care and lack of social protection. In a study on Bangladesh, Deolalikar (2005) found child malnutrition to be a leading factor to poor schooling and cognitive outcomes. Increased research awareness has also been directed to the issue of poverty-reduction effectiveness of different policy measures targeting children. For evidence on this point, some authors have illustrated (for example, Bird, 2007) that good qual- ity health and education provisions are important to limit the intergenerational transmission of poverty. 2.7. Child poverty discourse in Bangladesh Although the experience and consequences of child poverty are diverse and widespread (Harper, 2004), less attention has been paid to child poverty than to research on adult poverty in the de- veloping world. Bangladesh is not an exception in this context. In a previous quantitative study 36 on Bangladesh using Household Income and Expenditure—2000, it was found that a significant number of the country’s children live in extreme poverty (Begum, 2005; Begum, Deng and Gus- tafsson, 2012). Although the monetary approach to identifying and measuring both poverty and child poverty is widely used (Bradbury et al., 2001; Dieden & Gustafsson, 2003; Nolan & Whelan, 2007), it has been challenged by other multi-disciplinary approaches (Matin & Hulme, 2003; Minujin et al., 2006; White, Leavy & Masters, 2003). The main argument of these approaches is that the lack of income is not the only kind of deprivation people may suffer. Thus, interest among researchers is increasingly focused on the non-income aspects of poverty. This is evident in the application of qualitative approaches in the analysis and interpretation of poverty, for example, in the Human Development Reports, as well as in several other studies (Narayan et al., 2000; Wint & Frank, 2006). Among child poverty researchers, paying attention to children’s voices and using focus group interviews have become increasingly common (Attree, 2006; Fortier, 2006). Child poverty, as a central and specific topic, has not been very extensively researched in developing countries, which include Bangladesh. It should be noted that the World Bank, United Nations Development Programs (UNDP), Bangladesh Bureau of Statistics (BBS) and Bangladesh Institute of Development Studies’ (BIDS) publications on Bangladesh contain a number of studies that address poverty in general. UNICEF has paid considerable attention to Bangladesh’s poor children in its different reports. There are some analytical studies of poverty, which focus on specific theoretical perspectives, while others review empirical observations (Dowlah, 2006; Matin & Hulme, 2003; Murgai & Zaidi, 2005; Quisumbing, 2007; Sen, 1981). Recently, while addressing chronic poverty, notable writers are increasingly drawing attention to poverty in the early period of life that strongly reinforces IGT of poverty processes (Quisumbing, 2007; Sen & Hulme, 2006). Most studies, however, introduced the poverty of children into discussion in relation to other aspects of poverty and the policy agenda, such as chronic poverty, child labor, malnutrition 37 and low education (Amin et al., 2003; Deolalikar, 2005; Haider, 2008; Matin & Hulme, 2003; Salmon, 2005). For instance, in a study on poverty and malnutrition in Bangladesh, Deolalikar (2005) focused on poor children in relation with malnutrition. He found that child malnutrition is a factor leading to poor schooling and poor cognitive outcomes. Moreover, while focusing on child labor, schooling and school enrolment subsidies, Ra- vallion and Wodon (2000) find the association between low parental earnings and child labour in Bangladesh. They found that the low current household income kept poor children out of school and the income from child labour damages children’s long-term prospects of escaping poverty through education. Thus, some authors have argued that poverty is transmitted into the next gen- eration in Bangladesh. Recently, studies are increasingly drawing attention to intergenerational poverty while addressing chronic poverty (Quisumbing, 2007; Sen & Hulme, 2006). The poor children are most likely brought into discussion in both poverty analysis and policy agenda as one of the major reasons of different problems such as chronic poverty, child labour, malnutrition and low education. However, this concept previously lacked much needed attention and vitality in the policy measures that it deserved in order to tackle serious problems including chronic poverty and inter- generational poverty. One possible reason is that decision makers failed either to understand or identify this as a specific problem and rather counted it in household poverty. In addition, studies argue that appropriate and adequate policy measures in Bangladesh in reducing poverty among children are restricted mainly due to the lack of resources, prioritization of policy and budget al- location (Sen & Hulme, 2006) and good governance. 38 Chapter 3 About Bangladesh 3.1. Bangladesh and some other countries in Human Development and Multi-dimensional Poverty Index in HDR (2011) South Asia, a region of substantial population growth, comprises about a quarter of the world’s population. Bangladesh is one of the countries in this region; other countries are: Sri Lanka, Chi- na, Maldives, India, Bhutan, Pakistan, Myanmar, Nepal and Afghanistan (see Figure 2). Recently, Iran is also considered as a South Asian country (see HDR, 2011). It would be interesting to view selected information from HDR (2011) on these countries that have similarities, to some extent, to Bangladesh with respect to poverty. Presently, the official quantitative data on child poverty in Bangladesh and its neighbouring countries, as far as I know, is not available. I begin this section by presenting a brief discussion from the recent Human Development Report (HDR) of the year 2011. Data are drawn from HDR (2011) for Tables 2, 3 and 4. Figure 1 shows HDI trends respectively among Bangladesh and some other countries around it. All coun- tries presented in Tables 2,3 and 4 are from South Asia, except China and Myanmar. China and Bangladesh are among the countries in the world with the largest numbers of children. Myanmar is included since it shares a common border with Bangladesh. The discussion in the following section is based on Table 2 and the data presented in there are referred to the latest available data for each country. The HDI ranked 187 countries in three dimensions: the standard of living of its people, its population’s access to knowledge and its popu- lation’s chance of living a long and healthy life. HDR (2011) accounts the HDI categories that are defined based on values. The HDI orders countries in the HDI rankings as: very high, high, medium and low human development. With very high human development Norway, Australia and 39 the Netherlands stand on the top, while the Democratic Republic of the Congo, Niger and Burundi are at the bottom of the list with low human development. Bangladesh, Myanmar Nepal and Af- ghanistan have been ranked as countries with low human development. The facts contained in the Human Development Report (2011) make public an alarming situation regarding human develop- ment progress in Bangladesh. According to the Human Development Index (HDI) ranking of this report, Bangladesh is positioned 146 out of 187 countries and regions, with the HDI value 0.500. Only Iran has achieved high human development. Sri Lanka, China, Maldives, India and Bhutan have achieved medium human development except, Pakistan, Bangladesh, Mynmar, Nepal and Afghanistan that have been categorized as low human development. Iran is graded the highest position among the countries presented in Table 2. Only China and Maldives are a bit close to Sri Lanka followed by India and Bhutan, although the latter cases are far behind the prior ones. Bhu- tan at the bottom of the list of medium human development along with Pakistan, Bangladesh and Nepal (with low human development) are found almost at the same level at the bottom of South Asian HDI ranking where Afghanistan ranks the lowest. The life expectancy at birth of Maldives is 76.8 years, which is the longest not only in South Asia, but it also leaves China, behind. Bangladesh’s life expectancy (68.9 years) is com- parable with most South Asian countries (at around 65 to 69 years). Mean years of schooling Table 2. Human Development Index (HDI) in the HDR (2011) elected Asian Human Development Index (HDI) Life expectancy at birth Mean years of schooling Expected years of schooling GNI per capita in PPP terms HDI rank Value Years Years Years (constant 2005 international $) Iran 88 0.707 73.0 7.3 12.7 10164 Sri Lanka 97 0.691 74.9 8.2 12.7 4943 China 101 0.687 73.5 6.5 11.6 7476 Maldives 109 0.661 76.8 5.8 12.4 5276 India 134 0.547 65.4 4.4 10.3 3468 Bhutan 141 0.522 67.2 2.3 11.0 5293 Pakistan 145 0.503 65.4 4.9 6.9 2550 Bangladesh 146 0.500 68.9 4.8 8.1 1529 Myanmar 149 0.483 65.2 4.0 9.2 1535 Nepal 157 0.458 68.8 3.2 8.8 1160 Afghanistan 172 0.398 48.7 3.3 9.1 1416 Source: HDR (2011). Indicator s of countries, including Bangladesh . 40 in regards to access to knowledge Bangladesh is placed at the same level with, Pakistan, India and Myanmar. Bangladesh’s situation is better than India’s in this case. Sri Lanka and Iran come across with the highest numbers of expected years of schooling, with each having average years of 12.7. Maldives and China come next followed closely by Bhutan. India falls next in the table, followed by Myanmar, Afghanistan, Nepal, and Bangladesh, Pakistan’s expected years of school- ing is the lowest. In Table 4, in regards to adult literacy rate of the age 15 years and above, Bangladesh is placed at the same level with Bhutan, Pakistan and Nepal at around fifty five percent. This implies that almost half of the population of these countries are being excluded from the world of reading and communication. India’s situation is better in this case. Combined gross enrolment ratio for primary, secondary and tertiary schools, in the case of Bangladesh, is not better in comparison with other countries, except Pakistan. This ratio is much higher compared to Pakistan, which holds the lowest position, slightly lower than Bhutan and Myanmar, and much poorer than Mal- dives, China and Sri Lanka. The last column in Table 2 is the GNI per capita in PPP terms (constant 2005 international $) where Iran is seen to have the highest GNI per capita in PPP terms among the nations in the table followed by China. Bhutan, Maldives, and Sri Lanka fall next in Table having similar val- ues. Bangladesh falls in the lower part of the table with other nations of Myanmar, and Nepal and Afghanistan. In addition, Bangladesh’s per capita Gross Domestic Product (GDP) in Purchasing Power Parity (PPP) stands much below China, Maldives, Bhutan and Sri Lanka in this respect. Bangladesh is often referred to as one of the poorest countries in the world. Nevertheless, this country has been experiencing higher economic growth and remarkable poverty reduction since the 1990s up to 2006, although income inequality continues to remain a high concern (see Osmani, 2006; World Bank, 2007). Between 1980 and 2011 Bangladesh’s HDI rose progressive- ly, and simultaneously with all regions from 1980 till date (see Figure 1). In view of the indicators used in the HDI, relative performances could be arrested among ranked and compared countries. Yet, the HDI does not facilitate in defining, identifying or measuring poverty (UN, 2009). The discussion in the subsequent section is based on the information provided in Table 3. 45 3.2.1. A history of backwardness The territory, presently known as Bangladesh, was a part of the British Indian province of Ben- gal. The western and eastern parts of Bengal were also known as the West and East Bengal. The Western Bengal now belongs to India and Calcutta is its provincial capital city. The eastern part of Bengal was called East Bengal and was populated mostly by Muslims. It constituted East Pakistan later, a part of Pakistan formed in 1947. After successions of long colonial occupancies of hundreds of years this territory finally came into existence as an independent country in 1971, and was named Bangladesh. One approach to poverty analysis is political economy approach, which pays consider- able attention to the historical context within which poverty in South Asia was evolved (Silva & Athukorala, 1996). From this perspective, colonial occupancy of hundreds of years was an im- portant reason for a lack of economic development in South Asia, of which the territory which is presently called Bangladesh was also a part. To understand the political and economical ground- ing of poverty in Bangladesh, its history can be divided into four periods: before colonization, British period, Pakistan’s period and period after the independence. Before colonial period, the territory that constitutes present Bangladesh was called “Golden Bengal” for being one of the wealthiest parts of India (BBS, 2009). Bengal was recog- nized as one of the wealthiest part of the subcontinent up till the 16th century. During the 1000 BC era, this region belonged to a kingdom called Vanga, or Banga. In 1576, Bengal became a part of the Moghul Empire. This territory has been under the Muslim rule for over five and a half cen- turies from 1201 to 1757. During this period, most of the population of the eastern areas of then Bengal converted to Islam and since then, Islam has played a crucial role in the region’s history, politics and economics (European Commission [EC], 2007). Portuguese traders and missionaries reached Bengal in the latter part of the 15th century and were followed by the Dutch, the French, and the British East India Companies. The British period started when the British took control over Bengal in 1757 from the Nawab (king) of Bengal. The British colonial administration/exploitation in the Indian sub-con- 46 tinent continued for nearly two centuries. Establishments of very few educational institutions including University of Dhaka, railway and transportation, Dhaka Medical College, are some examples of British contributions. Throughout the colonial eras, East Bengal (presently Bangla- desh) received very limited investments in industrial sectors, development of infrastructure and transportation system. It largely relied on the production and export of its agricultural goods. Under the Brit- ish rule, however, industrial development and educational reforms advanced rapidly in western Bengal where most of the people were Hindus. Bengal in the British days had undoubtedly been dominated by Hindus, who formed the upper classes (see Blinkenberg, 1998, p. 256). Many Hin- dus gained economical and political power. In contrast, eastern Bengal, where most of the people were Muslims, remained backward and agricultural (see Blinkenberg, 1998, p. 256) and thus contributed to the fact that its poor population continued to stay poor and thus pass poverty on to their next generations despite those few who managed to move out of poverty. The beginning of Pakistan’s period was in 1947, after the British colonial era, when India was divided into two countries: India and Pakistan. Western Bengal, consisting predominantly of Hindus, became a state of India, while Eastern Bengal became East Pakistan due to their predomi- nant Muslim identity (see EC, 2007; Nilson, 2005). However, the population of East Pakistan (presently Bangladesh) viewed this inclusion into Pakistan as an ultimate new colony for the next 24 years until its independence. Under the Pakistani administration, this region experienced some progress in industrial- ization, urbanization and development of infrastructure etc. than the colonial period and it helped to garner a platform for the economic development for the future. From the Bangladeshi point of view, however, the era of Pakistan’s rule (between 1947 and 1971) repeated the same history of colonial exploitation such as administrative, economic, social and cultural injustice. Inequalities, differences in identity and culture, socio-economic and cultural exploitation of the west explicitly caused a contradiction between the two regions, i.e. East (now Bangladesh) and West Pakistan. Finally, the people of this region revolted and gained their independence in 1971, although at the cost of millions of lives and a broken economic system. Thus continuous colonial oppression led 47 this country to a great economic depression soon after its independence, which led the country to be recognized as one of the poorest country in the world. After the independence, economic recovery was a challenge for this newly formed coun- try. One could expect that the economic prosperity would free its people from poverty after the independence. Instead, the situation was disappointing and can be imagined from evidence that the new country experienced famine in 1973-74. Different studies found weaknesses in gover- nance such as massive corruption in administration (see Dowlah, 2006), lack of entitlement (Sen, 1981) etc. as the main factors in causing the famine. In addition, Bangladesh had faced the multiple challenges like poverty, globalization, and corruption including the lack of democratic environment19 since shortly after its indepen- dence. In 1972, the government led by the Awami league, nationalized all the large and medium scale industries according to its socialist strategy (Soviet based)20. The outcome showed very unimpressive results through low productivity and rapidly turned into loss making industries (Munshi, 2008) and this strategy was abandoned in 1975. Afterwards, it went through various trade and macro-economic structural adjustments initiated by the World Bank and IMF since the mid-1980s (Munshi, 2008). The aim of macroeconomic and trade reforms were to encourage exports by reducing anti-export bias. As a result, rapid industrialization and urbanization with steady economic growth are evident. In addition, the structural adjustment and privatization in Bangladesh’s economy and the related changes in the structure of employment have influenced people’s mobility, and speeded up urbanization and industrialization (see Afsar, 1999). The other policies and reforms in sectors like agriculture, land distribution and education also aided (ADB, 2001) to this developmental stream. A platform for the rapid socio-economic growth as required in this region (present Ban- gladesh) could not be developed for a long period. The country’s economic backwardness due to 19 In a constitutional amendment in 1974, the civil government banned all political parties and formed a new one BAKSAL (Bangladesh Krishak Sramik Awami League), banned all the news newspapers except four, held major constitutional rights of democracy such as freedom of press, rights of speech etc. In 1975, mili tary took control over the administration. 20 According to Munshi (2008), this strategy followed an import substitute industrialization policy, trade poli cies based on high tariffs and quantitative restrictions on import for a decade. 48 low education and lack of infrastructure were deeply rooted during the colonial periods. Although Bangladesh has improved in terms of economic growth and poverty reduction, and other social indicators such as education and health, its bureaucratic administrations during post-colonial pe- riod do not seem to be impressive to tackle the multiple challenges like poverty, globalization, climate change, and corruption. 3.2.2. Population The estimated population of Bangladesh is 144.5 million (BBS, 2009, 2008).The population den- sity of Bangladesh at 979 persons per square kilometre (BBS, 2009) makes it one of the most densely populated countries in the world (Quisumbing, 2007; Sen & Ali, 2009; Sen & Hulme, 2006; UNFPA, 2010). According to the 2005 estimates, around 36 million people, including chil- dren, lived in food poverty or extreme poverty (Quisumbing, 2007). Not surprisingly, the poor economy of this country is not able to deal with the problems related to its high population density and poverty related problems including child labour and malnutrition among others. Its population statistics is presented by sex, age, religion, and region (rural and urban) from BBS (2008) in Table 5, based on the information in this Table, a discussion is presented here. The information in Table 5 suggests that in a population of 144.5 million people in Ban- gladesh, most reside in the rural areas in comparison to the urban areas. In terms of socio-eco- nomic and cultural context, there is significant difference between rural and urban areas. A huge number of dependent populations such as children and women, together with elderly people have great impact on this country’s low income. 49 Unlike developed countries and similar to most developing countries, Bangladesh’s children con- stitute a great share of the total population, which is around forty per cent, and elderly people are very few in comparison to the country’s young population. Male to female ratio in the popula- tion distribution is almost equal, although males are a bit higher. The majority of population is composed of Muslims, followed by Hindus, Buddhists and Christians. Besides religious identity, there are ethnic groups21, who constitute less than 1 percent of the total population (BBS, 2007a). Most of them live in the Chittagong Hill Tracts and in the regions such as Mymensingh, Sylhet, and Rajshahi. They differ in their social organization, languages and culture from the rest of the people of the country. Tribal population’s right to access to jobs and higher education has been ensured by keeping quota for them in the public employment and universities respectively. The minority (on the basis of religion) population has a tendency to concentrate at par- ticular regions at rural level all over the country, although communal harmony persists to a great extent (BBS, 2009). Around 13 million people are Hindus in Bangladesh, even though it is an overwhelming Muslim state. Its constitution guarantees freedom of religion and equal opportuni- ties in all sectors regardless of nationality, religion, race and sex. However, there are controver- 21 Major tribes are the Chakmas, Maghs (or Marmas), Tipras, Murangs, Kukis and Santals. Chakmas and Mar mas generally live in the highland valleys. Table 5. Population distribution by demographic characteristics of Bangladesh Population by demographic characteristics Total population (in millions) Male Female individuals up to 14 years individuals aged 15-59 years individuals aged 60 years and above Muslim Hindu Buddhists Christian Others Rural Urban 144.5 51.2 48.8 33.73 59.76 6.51 89.35 9.64 0.57 0.27 0.17 74.9 25.1 Source: BBS (2008). . Notes: All statistics are in percentage unless otherwise indicated. 50 sies in favour and against in the issue of the limited access to jobs in the government or military, especially at higher levels, of the religious minority groups. 3.2.3. Politics, parliament and administrative units According to the constitution, Bangladesh is supposed to run by parliamentary form of govern- ment. However, it is argued that the history of Bangladesh does not appear according to the con- stitutional commitments in most cases (Haque, 2008): the country has been either under direct or indirect military rule between 1975 and 1991; since its independence, two presidents have been killed in military coups, martial law has been imposed three times and thrice a state of emergency has been declared. The most popular political parties are Awami League and BNP (Bangladesh National Party) among the many. Awami League remained in power (after independence in 1971 to 1975, 1996-2000 and 2009 to date) since independence. BNP won the election in 1991 and 2001, and ruled the country during the periods of 1991-1995 and 2001-2006. There are controversies regarding the ruling government’s democratic character, from the time of its independence until 1991 after which, “democratic governments” have been in power. Arguably, the presence of “somehow” democratic government had positive impact to se- cure economic development in Bangladesh during the above mentioned period. However, the rivalry and intention of capturing and holding power, struck corruption among the major political parties. Anam (2007) has a critical discussion on this issue. She blames the political leadership’s non democratic practices, army rules, repression by the ruling parties, corruption to lay down the groundwork for the events of January 2007: the political landscape in Bangladesh underwent a dramatic shift when military backed caretaker government took control of the nation. Repeated and longer interruption in democratic government is done by the military and other political par- ties that are holding back the country’s expected and projected economic growth. Thus effective long term policy measures for poverty reduction and economic development as required could not be developed or have been interrupted due to political conflict, instability and violence. The Prime minister and ministers are generally chosen from Members of Parliament of 51 the majority party. The national assembly is called “Jatyo Sangsad”; it consists of 330 members of whom 300 members are supposed to be elected directly by the people for a five-year term. To ensure women’s representation in the parliament, 30 women are elected by the parliamentarians. The administrative unit of Bangladesh is divided presently into seven administrative divisions (the highest administrative tier after national level) Dhaka, Rajshahi, Rangpur, Sylhet, Khulna, Barisal and Chittagong. Each is headed by a Divisional Commissioner (DC). During the Pakistan period, there were four divisions. It can be mentioned here that there were five divisions during the survey period of which data has been used for this dissertation in Chapter 5 and Chapter 7. The new divisions, Sylhet and Rangpur have been included in Dhaka and Rajshahi divisions re- spectively in the data used in this dissertation. A division consists of a number of “Zila” (districts), city corporation “Upazila” (subdis- tricts), “Thana” (Police Station), “Union” and “Municipality”22. Every union has several villag- es23 and an elected body of local government at the lowest administrative tier headed by chairman. At district level, “Zilla Parishad” (District Council) acts between the national government and the local government. At the local levels, the municipal corporations are divided into different wards and these are headed by elected ward commissioners. Bangladesh’s government execute their policy and program through this administrative system, which is highly bureaucratic24, and the bureaucrats play the key role in policy formulation and have absolute discretion in implementing them (see also Zafarullah, 2007). Common citizens are more familiar to the government offices than the parliament25. Therefore, the street level bu- 22 There are 64 districts, 6 city corporations, 308 municipalities, 481 Upzilas (sub-districts), 599 Police Sta tions, 4498 Unions (BBS, 2009). 23 There are a total of more than over 68 000 villages. 24 Zafarullah (2007, p. 164) indicated: The public bureaucracy in Bangladesh is constituted of discrete func tional cadres of whom the Administrative Cadre is the premier elite corps. Its roots lie in the Civil Service of Pakistan (CSP), which itself had its origins in the Indian Civil Service (ICS) the ‘steel-frame’ of British colonial rule. Like its forebears, it largely conforms to the structural attributes of the Weberian bureaucratic model, for example, open competitive recruitment system based on academic achievement and in-service training procedures, a promotion process, attractive perquisites, frequent rotation between departments, well designed post-retirement pension packages and so on. 25 Zafarullah (2007, p. 166) describes: Historically, sarkari daftars (government offices) and their incumbents are more familiar to the common citizens than sangshad (parliament). Village folks, for instance, can easily relate themselves to the more visible local council officers than their chosen parliamentary representatives who prefer to live away from their constituencies. 52 reaucrats (public employees such as teachers, police etc.), whom the people face in their everyday lives, have great impact, (Lipsky, 1980/2004; Zafarullah, 2007) on the general public. Evidently, despite the economic limitations for public expenditure for the developmental work and poverty reduction interventions, corruption, especially in the ill-suited bureaucratic system in Bangladesh, works from the opposite direction on developmental activities for its citizens. In developing economies, corruption is a serious issue (see Laffont, 2006). Desta, (2006, p. 42) argued citing from other studies that, “the problem of corruption within or across nations is not a recent phenomenon, nor is it exclusively a Third World problem. Corruption exists both in developed and developing countries in different forms, degrees and has differing consequences”. The author argues that developing countries would be much more prone to corruption than de- veloped countries, since they have other problems such as poverty, ill health, high levels of illit- eracy, low economic growth, and political instability. This is well-evident in Bangladesh’s context where corruption in the administration is being considered as one of the key factors that hinders its economic development. For example, massive government corruption has been blamed for the worst famine of 1974 (Dowlah, 2006). Anam (2007) states that during the military backed caretaker government’s confiscation of power, many of the former governments and political leaders were being held up and have been charged for huge corruption, and Bangladesh has passed through a political crisis. She argued that corruption made room for some of the highly ambitious army personnel to take over power. Consequently, the long term developmental policy implementations are either stopped or delayed. Arguably, these charges were controversial later, and were claimed as being a matter of political harassments in the cases of some political leaders. Laffont’s (2006) essay deals with why corruption is so widespread in developing coun- tries and documents the correlation between development and corruption. The author’s sketched a theory linking corruption and development that suggests corruption per transaction decreases with development, and finds an inverted U-shaped relationship between development (say per capita GDP) and the amount of corruption. The author interprets that this is possibly because of a role for the quality of democracy to strengthen the decrease of corruption. A well-functioning 55 low-lying land. Not only natural disaster but also the immediate effect of the dam called Farakka Barrage on Rajshahi results in flooding almost every year, which has a direct impact on its agri- based economy, which is this division’s main source of income. This is since it has the interna- tional river Ganges in common with India, called the Padma in Bangladesh’s part, on which the controversial dam ‘‘Farakka” has been built during the 1960s. As discussed by Tanzeema and Faisal (2001), the issue turned out to be crucial after 1975. This is because India began to with- draw water from the Ganges in the dry season. Heavy and sudden flood is reinforced by the re- leased surplus water by the dam, and draught caused by limiting the water flow during dry season (see Mirza, 2004; Tanzeema & Faisal, 2001). The damage caused by the Farakka Barrage to the country’s economy and environment is enormous (see Mirza, 2004; Tanzeema & Faisal, 2001). This region experiences the direct and immediate effect of being closest by location. This leads most of its marginal farmers to be made landless during flood, or a desert land during dry seasons. In addition, this division historically lags behind in human and social development over a period and received less importance, and has limited access to public facilities. Besides, other natural disasters (river erosion, draught, heavy rain, storms, cyclones etc.), overpopulation, parental separation, dowry, polygamy, illiteracy have direct and indirect impact on the poverty situation of Rajshahi division. Khulna division is situated in the south-western part of the country. Khulna was called the backbone of the former East Pakistan’s main foreign currency earner, having many large Jute industries, other heavy industries like ship building industry, news print mills etc. The country’s second seaport, Mongla Port and the largest mangrove forest in the world, the Sunderban, situ- ated in this division, also contribute to the division’s economic growth. In spite of having all these advantages of economic growth in Dhaka and Khulna divisions, they still lag behind Barisal and Chittagong. Over population, low education, unemployment, female headed households, inad- equate health care, malnutrition, violence, crime, etc. are considered some common problems to hold back the economic growth of Dhaka and Khulna divisions. Barisal division is located at the south part of the country. The division has the smallest population of the regions (see BBS, 2009). The literacy rate of this division according to BBS 56 (2003) has been the highest (59 percent) and far above the national average (45percent), which might allow a huge number of individuals access to employment. However, the lack of sufficient industrial and business holding is also considered as the root cause of poor financial status of this division. In addition to this other factors including flooding are also responsible for increasing the probability of a child being poor. Chittagong division is the south-eastern part of Bangladesh. Chittagong City Corpora- tion has the second highest population in the country. The largest seaport, Mangla is situated in this division and all large-scale maritime trading (both export and import) of the nation goes through Chittagong. The industrial development of this division is remarkable, which includes large and heavy industries. Chittagong is a very prosperous city. Another important resource of the division is the production of hydroelectric power dams. More importantly, some authors argue that the positive impacts of the “peace process (for the unrest among the ethnic groups) in the Chittagong Hill Tracts” during the late 1990s played an important role on poverty reduction in that region (see Sen & Ali, 2009). There are significant differences in different divisions in terms of the availability of pub- lic facilities, transport and communications, educational institutions, health services, industries, cottage industries, agricultural productions etc. The divisions are very different in terms of their areas, population, and socio economic points of view. These differ accordingly to a large extent in a country with small land. Not surprisingly, budget allocations for the development of different sectors such as infrastructure and communication, education, healthcare services, employment opportunities etc. also vary according to the government priorities on the basis of size of popula- tion and area of divisions. The occurring and type of natural hazards are also very different in dif- ferent administrative divisions. The opportunities of employment, the rate of wage and business facilities also differ depending on administrative divisions and geographical locations of Bangla- desh. These are some factors that are related with variations in poverty in different administrative divisions. 57 3.2.5. Economy Bangladesh is called an agri-based country. Bangladesh’s agriculture is mostly traditional. Its sub-tropical monsoon-type climate favours its agri-based economy which is subject to frequent devastating cyclones and floods. As a result, poor people including their children are vulnerable to natural disasters. Bangladesh’s development in recent times shows that its gross domestic product (GDP) growth rate has averaged 6 percent per year during the last 12 years. This country follows a mixed economy that operates on free market principles. Its economy’s recent trend shows a shift from agriculture to service and industrial sectors, which is very usual for the developing countries with high economic growth. The highest share of its GDP comes from service sectors, and the rest comes from industrial sector and agriculture. The relative contribution of agriculture in GDP has been decreasing over the last few years and the industrial sector has been increasing at the same period (see World Bank, 2007). The fast expanding industrialized sectors like textiles, fisheries, cottage, tea etc. are playing important roles in its rapid economic growth. In spite of this trend, still most of the labour forces in rural Bangladesh are employed in agriculture. Although a larger share of the country’s economy is agri-based, dependency on a tra- ditional cultivation system mainly hinders its potential of high economic growth in a country where people are vulnerable to natural calamities. Different initiatives have been introduced to encourage double and triple cropping, intercropping, and increasing fertilizers use. The country has natural resources like gas and mineral, an estimated huge amount of which is still unutilized and unexplored. This is very typical for a developing country (see Akram-Lodhi, Borras, Kay, & McKinley, 2007). The limitations of resource from unutilized sources are not the only factor that has blocked a higher economic growth. Other factors such as lack of an upgraded trade and trans- port system, governance related problems etc. also play a significant role that hinders the growth of Bangladesh’s economy (Devarajan & Nabi, 2006). Despite these problems, several studies appreciate Bangladesh’s average economic growth, especially since the 1990s, that has been crucial in raising living standards and reducing poverty (for example, Osmani, 2006; Sen & Hulme, 2006; World Bank, 2007). A significant num- 58 ber of households in which children live, regardless of its economic growth, continue to live in poverty. Consequently, poverty continues to be a major social problem and a significant number of children are victims of this situation. Inequality in income distribution among the households is another big concern for both adult and child poverty in Bangladesh. Continued high income inequality, in both urban and rural areas, remains a major concern for the potential impact of Bangladesh’s recent progress in reduc- ing poverty and improving economic growth on the poorest (Sen & Ali, 2009). This has been reflected in Rodas-Martini (2001) in which the author describes that Bangladesh has the greatest and most volatile inequality among the countries of South Asia. The adverse effect of high degree income inequality blocks the poorest of the poor population, including children at the bottom level, in attaining benefits of its high economic growth. 3.2.6. Education and employment Illiteracy is another key problem and the country has presently achieved almost half of its goal of attaining hundred per cent literacy rates (HDR, 2009). Bangladesh’s recent improvement in girls’ education and increasing school enrolment are remarkable. These results have been brought about by different initiatives. Despite having free education (primary [1-5 grades] and secondary [6-10 grades] for children of both sexes and rural girls respectively) and at times associated with certain incentives such as Female Stipend Program (FSP), education is not guaranteed for all poor chil- dren and does not prevent many poor children from working (Campaign for Popular Education [CAMPE], 2007; Haider, 2008; Kabeer, 2003, 2009; Mahmud, 2003; Ravallion & Wodon, 2000; Salmon, 2005). Other child poverty related problems prevent many children from going to school. The national education system is divided mainly into Primary (from grades 1 to 5), Sec- ondary (from grades 6 to 10), Higher Secondary (from grades 11 to 12) and Madrasa (Islamic education, 12/14 years schooling). Undergraduate, graduate and postgraduate studies are offered by different universities (autonomous and private, public) in the country. Primary schooling con- tinues to remain the first and foremost priority for developing countries. Likewise, Bangladesh’s priority was to make primary schooling accessible to all of its children. 59 Several studies (Devarajan & Nabi, 2006; GRM International, 2010; Osmani, 2006; UNI- CEF, 2006; World Bank, 2007) appreciated Bangladesh’s “positive changes”, and “tremendous achievement” in primary school enrolment in recent years, gender parity in school as a result of large-scale demand-side initiatives for the girls and the poor (Ravallion, 2006). These steps are: i) universal and compulsory free primary education, ii) free secondary education for the rural girls, iii) and the scholarship program for secondary schooling for rural girls, iv) increase in budget allocation on education etc. It can be mentioned here that these welfare provisions are only avail- able in the public schools and does not cover all school aged children. Despite these initiatives, a huge number of children drop school before completing their primary education due to poverty, and these children remain out of school due to insufficient free primary school. The private schools are expensive and sending the children to these schools is beyond the capability of poor parents. In developing countries, Case (2006, p. 270) argues that wealthier families can afford to educate their children, and aid them in finding superior jobs. However, Bangladesh’s poor children are far behind all of these. In addition, the adult literacy rate shows that nearly half of its population does not have any education (see Table 4). This is one of Bangladesh’s major problems, perpetuating a number of other problems such as unemployment, child labour, low earnings and poverty. In addition to illiteracy problem different studies raise the question of quality of given education in Bangladesh (see Ilon, 2000). Education in public institutions at all levels is very cheap where primary education and education for girls up to 12th grade is completely free of cost. However, public institutions are not as adequate as required for all children. On the other hand, education expenditure for second- ary, higher secondary and university education at private institutions are comparatively very high, which causes exclusion of the poor. Thus, high education expenditure and lack of access to public educational institution play negative roles in children’s education. Generally, better educated workers earn higher wages on average. There are casual ef- fects of schooling on earnings in both developing and developed countries (See Case, 2006, p. 269). It is evident in Bangladesh’s case by the fact that it is a low income country with low literacy rates. Although the unemployment rate has dropped recently with its stable economic growth 60 (see World Bank, 2007) household unemployment and underemployment remain an important social problem for this country. Recent contributions (UNDP, 2011) have documented that over thirty eight percent of its population is unemployed or underemployed, given that the country has a labour force over sixty million. UNDP (2011) further points to the fact that these numbers are increased every year by two million job seekers. In addition, a huge number of women continue to stay outside the job market, although women’s labour market participation has increased sig- nificantly in low paid jobs. A sizeable portion of the manpower is employed in different countries of the world, mainly in the Middle East. International migrants contribute significantly as remittances to the household disposable income, in addition to the earnings of larger proportion of the labour force domestically employed in agriculture, forestry, fisheries followed by service sectors and manufac- turing (BBS, 2007). As many developing countries, a significant number of child labours work in Bangladesh, indicating the probability of their growing up to a similar number of poor adults in the future. Most of these children work in agricultural sector, especially in rural areas. They also work in informal domestic work, especially in urban areas. Turning to rural urban interplay, rural and urban economies are highly interdependent and complementary in Bangladesh as in many other developing countries (Bhuyan, Khan, & Ahmed, 2001). However, the difference between urban and rural areas is high in terms of employment, access to education, health care services, and transportation system and so on. Although both the manufacturing and service sectors have expanded rapidly in rural areas that provides the oppor- tunity of new employments, migrants concentrate on the city in search of jobs and better lives. 3.2.7. Health care From a health point of view, child health care plays a crucial part in child development and has a great impact on educational attainment in early stages of lives and thereby to increase the likeli- hood of getting access to better jobs and higher earnings (Deolalikar, 2005). Bangladesh’s recent tremendous improvement in health indicators such as malnutrition, immunization, infant mortal- ity rate, life expectancy etc. proved to move hand in hand with fertility decline and increase in 61 literacy rates, especially improvement in women’s literacy rates. However, ill-health continues to remain as one of the major social problems. Bangladesh’s recent remarkable progress in expand- ing child immunization and lowering child malnutrition, declining infant and child mortality can- not conceal the fact that malnutrition still remains among the highest levels in the world (Amin & Basu, 2004; Deolalikar, 2005). This exacerbates a situation, which already creates concerns over child health care. In Bangladesh, the low percentage of GDP spending on the health sector (Osmani, 2006; HDR, 2011) indicates the likelihood of having inadequate health care services, particularly, for poor children in most cases. World Health Organization’s (see WHO, 2010, Table 7, p. 131) latest available data (of the year 2007) on health expenditure ratio shows: firstly,social security expenditure on health (as percentage of general government expenditure), and private prepaid plans (as percentage of private expenditure) on health, are zero. Secondly, out-of-pocket expenditure (as percentage of private expenditure) on health is 97.4. When it comes to per capita government expenditure on health, it is 14 PPP international dollars26, which is very low. As noted in Table 4. These reflects that the state spending on health care system is very limited as in many developing countries. Universal health care (free of cost for all) has been committed in the constitution, which has a long way to go to become a reality. The existing health care system is not well organized and composed of mainly public, private and NGO services. The public sector is largely used for in-patient and preventive care while the pri- vate sector is mainly for out-patient curative care. In the absence of public insurance system and required quantity and quality of health care system, and addition of corruption, spending on health mostly is met by patients’ own or their households in Bangladesh (see Ahmed, 2005, p. 12). Another concern is that ill health in Bangladesh increases economic inequality and holds back growth. Ravallion (2006, p. 207) presents empirical support from a vast set of literature for the link between inequality and growth, via the incidence of under nutrition. According to his findings, ill health lowers aggregated productivity. This is more acute in poor communities where high income inequality, less economic growth and the highest degree of malnutrition persists. Ac- 26 PPP international dollars are calculated using purchasing power parities (PPP). These are rates of currency conversion constructed to give an explanation for differences in price level between countries. 62 cording to WDI (2007, p. 37), poor communities—rural, remote, and in urban slums—typically face multiple health risks related to gaps in infrastructure, services, and trained personnel. In such a situation, as stated above, children living in poor households are assumed to be more exposed to ill health. 3.2.8. Policies and programs Bangladesh is largely following World Bank’s policy for its anti-poverty strategy, which consists mainly of: pro-poor economic growth 27policy to increase income and employment opportunity; human development to enhance the capabilities of the poor by providing them with more educa- tion, better health services etc.; gender-sensitive development to reduce gender discrimination; provision and expansion of safety nets; expansion of participatory governance to empower the poor; and reforming the service (see Islam, 2004). The poverty alleviation programs in Bangladesh include means tested cash transfers for children for school enrolment, for the poor elderly, the working age (work based), and pension for the public employees. As discussed by Ravallion (2006), Bangladesh’s Food for Education Pro- gram (FFP) relies on community-based targeting of food transfers that aim to create an incentive for reducing the cost for the poor of market failures. He states that this program was one of the earliest of many school-enrolment subsidy programs now found in both developed and develop- ing countries. Another program is food/in-kind transfers, typically aimed at the poor and often com- bined with micro credit programs (see Islam, 2004; Ravallion, 2006). These safety net programs are administered through a wide variety of ministries concerned and NGOs for example, Grameen Bank, BRAC, Proshika etc., and the UN’s organizations. The pension program is an effective program to prevent poverty at an older age, although this program keeps a huge number of non- government employees outside. However, in comparison, the other programs are in reality inad- equate in meeting the needs of most of their target groups. In rural areas, most children are reported to lack access to public services including higher 27 Growth that reduces poverty. 63 education and healthcare because most opportunities are available in urban areas (Siddique, 1998) and for the better off. Differences are also apparent across the seven administrative divisions in Bangladesh. As mentioned earlier (in Chapter 3), there were five administrative divisions during the survey period: Dhaka, Chittagong, Rajshahi, Barisal and Khulna. Although most of Ban- gladesh’s poor live in rural areas, recent statistics show that urban poverty rates are increasing, particularly for the households in absolute poverty (Afsar, 1999; Murgai & Zaidi, 2005). Murgai and Zaidi (2005, pp. 14-15) noted that during 1990s rural growth “was broad-based while urban growth mainly benefited the relatively affluent”. Due to poverty even within the urban areas, few families can afford the expense of the desired level of education (Siddique, 1998) or health care for their children. As stated in the preceding section, Bangladesh’s achievement is regarded as tremendous, particularly, in expanding child immunization and lowering child malnutrition (Deolalikar, 2005). However, malnutrition still remains among the highest in the world and more severe than that of many other developing countries (Deolalikar, 2005; Dowlah, 2002). Additionally, many poor children remain untreated due to the lack of access to limited facilities in public hospitals. Similar to interventions for education, the initiatives taken by the government and NGOs are not adequate for the poor household and their children. This is likely to lead the nation to severe consequences. This is because, in most cases poor children, who are poorly nourished, have poor education and low aspirations, as noted in Harper (2004):“may be unable to reverse these accumulated prob- lems later in life”. Osmani (2006) expresses his concern that the low level of public expenditure, particularly in education and health sectors in Bangladesh, which is even lower than developing countries’ standards, bears the risk of not improving enough the capabilities of the poor. In other words, low public expenditure per capita is resulted in low levels of human development. Thus, limited resource for public or private investment is an obstacle in initiating child poverty reduction strategies in Bangladesh, which is very similar to the situation in many devel- oping countries. The lack of well-organized redistribution of wealth and taxation system, proper utilization of natural resources etc., nature of prioritization of policies and resources are some of the reasons why child poverty receives less attention in policy agenda. Moreover, studies high- 64 light that together with limited resources, bad governance leads to bad policies (Hulme & Sheph- ard, 2003), which is very crucial for creating an environment of long term policy implementation. The political instability, corruption, conflict and violence also have been identified to be associ- ated with bad governance in many countries including Bangladesh. Together with the above-mentioned problems, natural hazards, floods in particular, ex- acerbate household poverty almost every year while abnormal hazards come as an unexpected shock and cause unanticipated decline in income and consumption (Banerjee, 2007). Food scar- city due to natural hazards ultimately causes price increases in the most affected divisions. Ac- cording to Sen’s “entitlement approach” (Sen, 1981), this goes beyond the purchasing power of poor families. It hurts the poor more than the non-poor (Banerjee, 2007). Obviously, under all these adverse circumstances, children in poor households are more affected. The underprivileged are unable to move out of poverty through own initiatives and tend to grow up with less education, which increases the likelihood of passing poverty on to the next generation. 65 Chapter 4 Methodological issues 4.1. Few words about methodological issues This dissertation uses both quantitative and qualitative methods to study child poverty from a number of standpoints. My intention is to treat quantitative and qualitative data as complementa- ry. My motivations are: firstly, some aspects of poverty can be well researched using quantitative data. For example, in this dissertation the quantitative studies aim to provide with statistical in- formation using large scale survey data, which is representative at national level based on specific and limited indicators. This provides the opportunity to examine the extent and characteristics of child poverty in Bangladesh (see Chapter 5) and to compare the extent and development of child poverty in Bangladesh and China (Chapter 7). In addition, an attempt is made to examine the incidence and factors associated with child poverty over time and across countries. For these kinds of studies, quantitative method is better suited as statistical or numerical representation is the main concern. Secondly, for some other research questions qualitative methods are better suited. Al- though household survey is being regarded as the primary data collection tool, its limitations for grasping the complete picture of poverty due to the use of closed-ended questions (see Narayan, Patel, Schafft, Rademacher, & Koch-Schulte, 2000, p. 14). For example, the gender issue, an important aspect of poverty, cannot be well captured by conventional survey data, since consump- tion is assumed to be equal for household members. In reality, intra-household allocations (food distribution in particular), are not typically equal for females and males. This gender discrimination causes girls and women to be more vul- nerable to poverty (Chant, 2006; UNICEF, 2006). Similarly, one can argue that children do not experience poverty in the same way as adults do. Another example of survey data not providing 66 the complete picture of child poverty is the lack of data covering the violence, negligence, and discrimination experienced by the child. These issues are especially important for child poverty research to explore. Thus child poverty needs to be addressed using other forms of data (qualita- tive) as well. For in-depth knowledge and better understanding of the principal victims’ percep- tions and experience of child poverty, qualitative study is needed (see Narayan, et al., 2000). The World Bank Poverty Assessments are beginning to include qualitative and partici- patory methods to complement household information. Recently, focus group interviews are in- creasingly in use among child poverty researchers. Its advantage of producing a wealth of detailed data from a very small number of people provided the motivation to expand on my previous quantitative study of Bangladesh’s child poverty. More importantly, through the voices of the poor children and women themselves, the present study first attempts also to heighten awareness of policy makers so that appropriate policy can be formulated and implemented to alleviate child poverty in Bangladesh. Quantitative and qualitative observations provide researchers with different ways of op- erationalizing and measuring theoretical constructs and practical concepts. Therefore, different methods are used in different chapters. I use quantitative method, which provides a high level of measurement precision and statistical power, regarding child poverty in Bangladesh and China. On the other hand, qualitative data produces a wide range of in-depth knowledge from the partici- pants’ view in Bangladesh’s context. My intention in the latter case is to raise issues relevant with child poverty but not to generalize. Turning next, I aim to add more on methodological issues in the following sections by addressing the procedure of data description and data analysis used in this dissertation. 4.2. Quantitative Studies I worked with other researchers in the quantitative study on Bangladesh and China. Chapter 7 is based on a research paper co-authored with Björn Gustafsson and Deng Quheng. The collaborat- ing researchers were experts on China (while I had the role as expert of Bangladesh) so the chap- ter could build on expertise on both countries. Besides, I conducted the other quantitative study 67 (Chapter 5) myself in which large sample surveys are used. 4.2.1. Data description The National micro data of Bangladesh Household Income and Expenditure Survey (HIES) of the year 1999/2000 are used for the Chapter 5 whereas HIES1995/96 and 1999/2000, and Chinese survey data (China Household Income Project, CHIP) of the years 1988, 1995 and 2002 have been used for Chapter 7. HIES—2000 used almost the same sample design as in HES (Household Expenditure Survey)—1995 with slight modification. The definitions for rural and urban were the same in both surveys. The modifications were made in four Statistical Metropolitan Areas. In ad- dition, some changes were also made in the questionnaires in 2000. However, they did not affect the comparability of the facts of this dissertation. The number of PSUs (Primary Sample Units) in the Statistical Metropolitan Areas (SMAs) was doubled in HIES 2000 making it 140 instead of 70 in HES 1995-96. But at the same time the number of households of each of these PSUs was halved in 2000 i.e. only 10 households in 2000; in HES 1995-96 it was 20 households (BBS, 2003). Thus the number of sample households remained the same in both surveys. Two-stage stratified random sampling (under the framework of Integrated Multi-Purpose Sampling) has been used to draw samples for 1995-96 and 2000 from Population and Housing Census 1991. For the survey, each enumeration team28 were trained for interviews and data col- lection. Each interviewer and female facilitator pair visited and interviewed 10 households (in one PSU) each per month and collected data by interviewing (BBS, 2003). The total, rural and urban sample households’ and individuals’ numbers for both survey years are presented in Table 6. It can be mentioned here that I had access to the microdata, which made it possible for me to do all computations myself. The Chinese data comes from the three waves of the CHIP (China Household Income Project) survey conducted in 1989, 1996 and 2003 for the reference periods 1988, 1995 and 2002. Both rural and urban areas are included in the surveys. The rural survey of 1988 covers 28 prov- inces, while nine were not retained in the 1995 survey. These same provinces and also Guangxi and Xinjiang come into view in the 2002 survey. The 1988 urban survey was drawn from 10 28 For HIES-2000, there was one supervisor, 2 interviewers and 2 female facilitators. 68 provinces to represent the Coastal, Central and Western provinces (for more details on the Chinese data, see Eichen & Zhang, 1993; Li, Chuliang, Zhong, & Ximing, 2008). The two samples for Bangladesh and in the Chinese sample for 1995 contain around 16 000 chil- dren (individuals aged 14 years) in total. The larger size for the Chinese sample of 1988 is due to: a) another sampling scheme; and b) higher earlier birth rates. In addition, the decline in birth rate in China is the reason for the Chinese sample of 2002 not having more than 9000 children. The two later sub-samples for southwest China comprise about 2000 children. Income-based child poverty in Bangladesh is measured using monthly income from wages and salaries, pensions, contributions and professional fees earned by the members of the household. To estimate monthly household per capita income, all members’ incomes from both agricultural (such as crop production, livestock and poultry, fish farming and fish capture, farm forestry, agricultural assets etc.) and non-agricultural (such as daily wage, salaried wage, non- agricultural enterprise etc.) sources have been considered. Non-wage income from interest, divi- dends, earnings from agricultural and non-agricultural activities, land and property, rent, gifts and assistance, insurance benefits, and other special types of receipts are also included in household income. Income-based child poverty has been measured using monthly income by taking into ac- count the above-mentioned income. 4.2.2. Approaches to data analysis For quantitative data analysis, multivariate analysis and decomposition framework have been Table 6. The numbers of total, rural and urban sample households, and individuals Observations HIES 2000 HES 1995 All Rural Urban All Rural Urban Sample Households Individuals Adult: Children 7440 38518 23447 15071 5040 26231 15548 10683 2400 12287 7899 4388 7420 39501 22482 16569 5040 26453 14794 11659 2380 12598 7688 4910 Source: . Bangladesh Household Expenditure Survey (HES)-1995 and Bangladesh Household Income nad Expenditure Survey (HIES) 2000. Note: All statistics are in numbers. 69 used. The quantitative data was used to identify the poor and also to analyse the poverty profile (Chapter 5); and how economic growth has affected the development of poverty (in a comparative framework in Chapter 7). As mentioned earlier, the measurement of poverty used in quantitative studies is usually income poverty. For measuring and defining child (income) poverty, “International Poverty Line” (one dollar per person per day) is employed throughout the dissertation unless otherwise indicated, which has been defined by the World Bank. This measure is widely used for the developing world (see for example, Dieden & Gustafsson, 2003; Gordon et al., 2003). Deaton (2006, p. 12) states: “The World Bank uses this poverty line, a common international poverty line, designed to be appropriate for extreme poverty, defined as poverty in the poorest countries”. I brought into notice briefly about some of the controversies (see for example contributions in Anand, Segal and Stiglitz, 2010) with regard to the limitations of this line in section 2.4. The most recent update on this is found in Ravallion, Chen, and Sangraula (2009). This paper presents major update of the international “$1 a day” poverty line, which has been raised to$1.25 PPP in 2005 prices. A serious limitation this dissertation shares with all poverty measurement is the issue of intra-household allocation, which has not been taken into consideration. That is the process (Haddad, Hoddinott, & Alderman, 1997) by which time, money, and other resources are allocated within households. In this dissertation, it has been measured, firstly, if a household as an entity is poor,then it deems that all children living in the household are poor. In reality, some children in non-poor households have low consumption and in some poor households some children have high consumption due to the differences of age, sex, household composition, etc. However, the appropriate estimation of intra-household allocation from the conventional income data used for this study is not possible. Considering limitations of the alternatives, I prefer an income based poverty line for measuring child poverty in Bangladesh, because, poverty measurement solely in terms of in- come is widely used in child poverty research. As mentioned earlier, the poverty line has been calculated at Tk. 500.16 per capita per month by estimating in terms of PPP in 1985 prices for Bangladeshi currency, the Taka. The main reason for selecting it for child poverty measurement 70 is its acceptance for comparability at the international level. An alternative poverty line, USD 2 PPP, has also been used here in Chapter 5 to examine the robustness of the results obtained from the International Poverty Line. This line has been calculated at Tk. 1000.32 by doubling the USD 1 per day poverty line (International Poverty Line). It should be mentioned here that there is no official measure for assessing child poverty in Bangladesh. Using this poverty line, child poverty has been measured for the population of Bangla- desh in which children are considered as poor if they live in households with disposable per capita income lower than 1 USD PPP (Purchasing Power Parity).The same poverty line is employed and updated only with the changes in consumer price index (see Chapter 5 and Chapter 7). To investi- gate how child poverty varies across a number of variables, such as children’s age, sex, residence, religion have been used. In addition, other variables such as household heads’ educational level, employment status, and marital status, access to land, age, sex, household size, and religion have been examined as well. However, due to data limitations variables such as health indicators and natural calamities and others have not been studied. The unit of analysis plays a vital role in research, particularly in child poverty research. There are different units: households, family and individual (Corak, 2006a). The analyst’s choice of appropriate income receiving unit often considered the household; and, the unit of analysis is the individual; household income per capita is considered to each member of the same household. From these considerations, work is done in two steps. At first, income of all members of a house- hold has been considered to estimate per capita income. Next, based on the per capita income, households and all of their members are classified as poor or not. From this classification de- scriptive statistics are made, for example computing poverty rates for all children in Bangladesh, belonging to a particular category etc. As mentioned earlier, I assumed that incomes are equally distributed within household regardless of sex, age, birth order etc. In reality, however, this is not probably the case. In Chapter 5 multivariate analyses show how different variables affect the probability of a child being poor. I consider a child’s probability of being poor in terms of given socio-economic and demographic characteristics of the child or his/her household head. At this stage the main 71 unit of analysis is the child, although characteristics of the household head are used as explana- tory variables in some cases. Children under study of this dissertation are those individuals that are aged up to 14. It is difficult in the Bangladeshi cultural context to specify the age when an individual is considered a child because the word in Bengali language for “child” has different meaning than what is meant in the bench mark of UNCRC (UN Convention on the Rights of the Child)29. Concerning this, quoting from other studies, Haider (2008, p. 52) describes: There is no Bengali word to describe the life stage from birth to the age of 18, the bench- mark embraced by the UNCRC…“shishu”, the word in the “Bangla” language for “child”, has a meaning quite different from that spelt out in the UNCRC. Considering this cultural complexity, the individuals aged up to 14, minimum age re- quired up to junior (class 8) schooling, have been considered as children. For counting the poor, the most commonly used measure head count ratio has been employed. Child poverty rate has been calculated by dividing the total number of poor children by the total number of children. Multivariate analysis involves observation and analysis of more than one statistical vari- able at a time. Since child poverty has been assumed to be associated with a number of variables, I made use of this method for quantitative studies. Firstly, logistic regression has been used (see Chapter 5). I examined the influence of a number of independent variables such as location (both regions and rural and urban) of residence of a child, age of child and household head’s charac- teristics (education, employment, marital status etc.) on the variations of the dependent variable, child poverty. Secondly, in Chapter 7, a decomposition framework (see Danziger & Gottschalk, 1995) has been used for poverty comparisons between two countries (Bangladesh and China) over a time. Applying this method, the differences in poverty that have been observed between situations A and B are due to three components: differences in average income, differences in demographic composition and differences in the distribution of income. Finally, in Chapter 5, I also looked at the notion of wellbeing by examining the poor chil- dren’s living condition in terms of access to some variables, such as safe shelter, safe water and electricity. This provides an idea of how many of poor children are deprived of their rights to basic 29 In the year 1990, Bangladesh became a signatory of UNCRC. 72 human needs. Quantitative researchers use similar approaches to measure deprivation poverty for developing countries (see Gordon et al., 2003). Additionally, for a better understanding of depri- vation and suffering regarding fulfilling basic needs, the reflection from the qualitative study has been used in Chapter 6 and Chapter 8. 4.3. Qualitative studies I planned, designed, and conducted the qualitative studies (presented in Chapter 6 and Chapter 8). I collected and analyzed data myself from focus group interviews for qualitative studies to under- stand child poverty and its remedies as perceived or experienced mainly by some of its principal victims. Possessing the highest degree of vulnerability among the poor population, poor children and women are being considered the principal victims in this dissertation. I employed a grounded theory approach (see Glaser and Strauss, 1967) to listen to the voices of the poor and non-poor from their own point of views about what steps are needed to be taken. Section 4.3.1 describes data, and Section 4.3.2 details the approaches to data analysis that are used in qualitative studies as mentioned above. 4.3.1. Data description For the data collection, I used focus group discussion. Denzin and Lincoln (2000) suggested that focus groups offer a way for researchers to attend to multiple voices and are especially important for making audible the voices of oppressed people who are demanding that policy makers, politi- cians, academics and others in position of power listen to them. This approach addresses three important issues that I took into account when I chose to use data from focus group interviews for child poverty research. These are to listen to the multiple voices, to amplify the voices of the underprivileged and to bring them to the attention of the policy makers (Denzin & Lincoln, 2000; Krueger, 1997). Focus group interviews are a way of collecting qualitative data by engaging a number of people in an informal group discussion, which focuses on a particular topic or a set of issues (Wilkinson, 2004, p. 177). Compared to other qualitative methods, the moderator has less power 73 over a group than over a single individual (Denzin & Lincoln, 2000). The method was deemed useful for generating ideas, though there are some limitations such as group influence on indi- vidual members and level of participants’ involvement (Morgan, 1997). As an aspect of this, a pilot study was conducted in 2005 with slum dwellers in Dhaka, the capital city of Bangladesh. The pilot study was deemed very useful to avoid these limitations in the real setting as far as possible. Also, the other advantages from such a study were to examine the effectiveness in generating ideas on a new issue like child poverty and probable limitations in conducting focus group interviews. Mainly, the aim of conducting a pilot study was to learn from the field practically (see Patton, 2002) so that limitations can be tackled in the real setting. The group for the pilot study was composed of 1 poor man and 5 poor women. On the basis of the pilot study, I took into account different limitations as much as possible while designing the methodol- ogy of this research. I conducted five focus group interviews (the size of the groups varied). The composition of the focus group interviews (of which data has been analyzed in this disserta- tion) is presented in Table 7. The results from five focus group interviews are presented in Chapter 6 and Chapter 8. Chapter 6 used four focus groups (Group 1, Group 2, Group 3 and Group 4) com- posed of poor participants. However, Chapter 8 used data from five focus group interviews that include the same four groups composed of poor participants and one additional group composed of non-poor participants (Group 5). selection of participants: A total of 30 participants were recruited for the focus group in- terviews. For selecting the samples, snowball sampling was employed. This method is suggested Table 7. The composition of focus groups. Group 1 Group 2 Group 3 Total Group 4 Group 5 Total Children Children Children Children Poor women Non-poor women Participants Age (in years) 6-14 6-14 6-14 6-14 20-45 25-45 6-45 Sex Male Female Total 5 1 6 3 2 5 4 4 8 12 7 19 -- 4 4 -- 7 7 12 18 30 Note: Data were collected from focus group discussions he ld in Dhaka, Bangladesh over a period around one year from June 2005 to July 2006. 74 to be appropriate “for grounded theory interviews, whereby the researcher has limited exposure, familiarity and connectedness to diverse communities” (Green, Creswell, Shope, & Clark, 2007, pp. 480-81). In addition, I was guided by and proceeded to theoretical sampling (Backman & Kyngäs, 1999; Hjalmarson, Strandmark & Klässbo, 2007) to focus data collection as the analysis continued. In addition to snowball sampling, theoretical sampling was used as much as possible in order to “focus the data collection, enrich the categories that emerged and guide ‘where to go’” for the coming data collection” (Hjalmarson et al., 2007, p. 237). For example, the participants (all children) of the second group (Table 7) were chosen and asked to “facilitate the analysis by adding supplement information to the areas that were found thin” in the first group discussion, in order to eliminate uncertainty that I had about the emerging concepts and theories (Morse, 2007, p. 241). I followed a similar pattern for the rest of the focus group discussions until terminating data collection. My intention was, as stated by Morse (2007, p. 241), to get the “final missing pieces of the puzzle, polish data collection, and complete process of saturation”, in other words, information that I required. Three of these groups consisted of altogether 19 poor children aged 6-14 years, irrespec- tive of sex, religion or region. These groups were homogeneous in terms of age, educational level and poverty status. To capture the voices of women, I included in the other two groups 1) four poor women with low or no education (Group 4), and 2) seven working women who were not poor and had postgraduate education (Group 5). Poor women’s selection criteria were regardless of age (around 20-45 years) and religion. The poor participants (children and parents/household heads) were selected from three different slums of Dhaka, the capital city. However, they origi- nally migrated from different rural places and have been living in urban areas for more than five years. The non-poor group was composed of working women, who have postgraduate education, aged around 25-45. The lower limit of the age group was a bit higher since they were highly quali- fied, and got married and gave birth to children at a relatively higher age than the poor women’s group. The women participants in Group 5 belong to rich families in contrast to the poor women participants in Group 4. The inequality between members of the poor and non-poor groups is huge in terms of wealth and education. All women participants have their own children; however, none 75 of them are mothers of any children participants. Composition of groups: Besides children, women are in this dissertation regarded as principal victims of child poverty. Women and children were selected for focus group discussions because they possessed the highest degree of vulnerability among the poor population. As noted earlier, three of the groups were composed of poor children aged 6-14 (but heterogeneous in terms of sex, religion and region). The remaining group was composed of poor women, mixed in terms of age, region and religion. Listening to women, in particular, was crucial since they are consid- ered to be an underprivileged group among the poor population and they place a higher premium on welfare-related goals for children (Moore 2001, p. 9, UNICEF, 2006). For this reason, I was interested in listening to women’s voices and women were also more interested in participating when they were contacted, and I myself recruited all women participants. However, children groups consist of both sexes. I maintained a certain level of homogeneity within the groups and heterogeneity among the groups. What was common to all groups was that all the participants were rural migrants. Time and financial limitations prevented inclusions of participants living in all divisions of Bangladesh. Two locations were used for the focus group discussions as suggested by the key infor- mants and the group members. One setting was a kindergarten school and the other was a living room of a group member (non-poor) where the participants felt comfortable. Getting access is problematic, especially, to poor children. For selecting the sample of focus group interview, snowball30 sampling was used. To recruit samples, three key informants who had access to and special knowledge about the possible informants were selected and played an important role for getting access to the first participant. Then, this participant found another one and so on. In this way, I continued until the desired sample size was obtained. For screening31 eligibility, I used questionnaires containing few questions. I used a poverty line, which is similar 30 Cournoyer and Klein (2000) define snowball sampling as a technique when applied to studying network by asking earlier subjects to nominate candidates to be included in the sample. They add that it refers to a methodology that starts with a single element and then grows and grows like a snowball getting bigger and bigger as it is rolled through the snow. 31 This kind of screening is essential for matching individuals with specific groups when groups are composed in different ways (Stewart et al., 2007, p. 54). For screening, two separate questionnaires for background information, demographic characteristics and other variables in relation with the purpose of the study were used - one for children, and another for children’s parents and for women participants. 76 to the International Poverty Line as defined by the World Bank. The selected participants were considered as poor if their household income fell below this line. I formulated the focus group questions, designed and implemented the selection of the participants and moderated the discussions. My role was in most cases as unintrusive as possible, as described by Denscombe (1998), such that the researcher introduces a theme or topic and then lets the interviewees develop their ideas and thoughts. The initial question that I posed in each of the focus group was: “Whom do you consider to be poor children?” In each focus group the discussion on this question was followed by the questions: “Why do children become poor?”; “In what ways do the poor children suffer in poverty?” and “What can be done to alleviate poverty?” These questions32 based on my research questions guided me to explore the experience and knowledge of child poverty and related issues. I also prepared an abridged transcript based on relevant and useful portions of the discussions by watching and listening to videotape record- ings of each focus group. To illustrate the findings and interpretation of participants’ statements, translations of different quotes from the original transcripts are reported in the above-mentioned chapters (See Chapter 6 and Chapter 8). Each focus group interview lasted around one hour and was conducted from June 2005 to July 2006, in local language (Bengali) and videotaped. Later I transcribed the discussion and translated into English. Two assistants who have postgraduate degrees in Social Work helped me throughout the sessions, taking field notes etc. The participants have been informed that partici- pation in discussion is voluntary. To get informed consent the parents/ household heads/children have been informed about the nature and purpose of the study verbally and in written. An in- formed consent form to each of the participants, including children and their parents, was given in advance to review and sign. For reliability and validity for focus group interviews, several ac- cepted protocols were followed (Krueger, 1997, p. 68). The discussions were monitored carefully for ambiguity and participants were asked to verify the moderator’s summary and interpretation. 32 Since my intention was to allow discussants to develop their own thoughts by using their own words, I con sidered thematic interviews more appropriate in focus group discussions. In this way, emphasis was given on participants’ thoughts, experience, knowledge and perception of child poverty and its relevant issues in Bangladesh. 77 4.3.2. Approaches to data analysis To address research questions in Chapter 6 and Chapter 8, data from focus group interviews have been utilized by identifying the main themes and concepts. Grounded theory approach has been used for qualitative data analysis. It was originally developed by Glaser and Strauss in “The Dis- covery of Grounded Theory” (1967) in which the authors argue that qualitative analysis could systematically generate concepts and theories based on observational and interview data (see also Marvasti, 2004, p. 83). The advantages of using this approach (see Charmaz, 2005, p. 519) are: firstly, it provides emphasis on theorizing close to the data and secondly, it is committed to the development of theo- ries. These advantages were very useful to contextualize the phenomenon of child poverty, and for making sense of data in our study. Especially, this approach’s closeness to the data, commitments to theory development, and flexible guidelines were the motivations. This method provided me with the possibilities of conceptualizing, and discovering reasons and impacts of child poverty in some of those fields in which survey data has limitations, in other words, to contextualize the phenomenon of child poverty from the victims’ experience. Coding is the first step in taking an analytical stance in grounded theory approach. As dis- cussed by (Charmaz, 2006), codes are emerging theory; and it was the coding process that started the chain of theory development. I made use of substantive coding, as stated in Holton, (2007, p. 265), in which the researcher: “works with the data, directly, fracturing and analyzing it, initially through open coding for the emergence of a core category and related concepts”. open coding: Coding of data in this study is embarked on writing expressive labels. Typ- ically, this procedure of coding is known as “open coding”. I used open codes aiming to put data in meaningful categories or themes. I put down a series of concepts that came to my mind during reading through the data (see Charmaz, 2005, pp. 518-19). My intention of utilizing open coding was to identify topics of interest in my study. As I continued, certain codes (labels) commenced to reappear. Thus, open coding entails a range of codes encompassing the reflections, thoughts and denotations of the transcripts. The “Open code” columns of Figure 5 and Figure 7 in Chapter 78 6, and Figure 8 in Chapter 8 provide examples of a summary of open coding. The process started with line-by-line coding of extracted transcripts of FGDs. For selecting this method, I considered, firstly, the time factor. For example, the analysis in word-by-word coding takes a bit longer time in comparison to line-by-line coding. Secondly, using word-by-word coding can lead to losing of pattern to an extensive amount. An example of line-by-line coding is presented in Figure 4 (Chapter 6). The remark, quoted in this Figure, was delivered by a participant in the “poor women group”. By coding her statement line-by-line, it gives us an idea that low parental earnings were insufficient to spend on educational expenses. Also, this statement illustrates a clearer expression of a non-educated mother’s missed opportunity and deprivation in one hand, and her eagerness to education in the other. Moreover, poverty transmission can be understood between generations (from parents to children and onwards). This statement and similar statements from other participants of the dis- cussions were considered and taken together for generating the axial code of “encompassing multidimensionality”. Axial codes are sub-categories. Axial coding or sub-category: As open coding ended up, axial coding was initiated. Axial coding is the next coding phase in the analysis process of relating categories to their sub- categories at the level of properties and dimensions. In axial coding, relationships among the open codes are found to commence, as discussed by Charmaz (2006), and Corbin and Strauss (2008). The coding is called axial because its occurrence takes place around the axis of a category. Figure 6 (Chapter 6) provides an example of axial coding: from FGD’s statements to first level (open) and thereafter second level (axial) codes (or sub-categories). At this stage of coding, I noticed that some of the relationships around certain groups of open codes, which showed that some of the characteristics of these codes having commonalities. As I attempted to organize multiple open codes, some connections among certain sub-categories were revealed. Thus, a more abstract conceptual level of axial codes emerged (see Figure 5, Figure 6 and Figure 7 in Chapter 6). For example, open codes (non-fulfilment of basic needs, parental low earnings, high ambition, de- prived of education) in the columns “Open codes” deliver the axial code of “focussing suffering of deprivation” (Figure 6 and Figure 7 in Chapter 6; see also Figure 8 in Chapter 8 for examples 79 of open codes). Category and selective coding: The final stage of coding process is “selective coding”. Some of the principal concerns in this coding are: the discovery of a core category, relating all other categories to this core category and refinement of a theory. The conceptual ideas were stim- ulated during the process of coding of the transcribed texts of focus group discussions (FGDs). Familiar characteristics were shown by a number of axial codes. At this point, I grouped them in higher level groups, each of which was called “category” (Figure 5 in Chapter 6; Figure 8 in Chapter 8). Additionally, some new events were placed into previously emerged categories. As I con- tinued, new categories came to appear in this process. Thus, I coded for as many categories as fit successively. Relevant concepts of interest were selectively chosen from axial codes as categories (e.g., categories A, B and C in Figure 7 in Chapter 6). I continued this process to draw up the boundaries of data collection and analysis. Holton (2007, p. 275) argued that this could be done for “selective coding to theoretically saturate the core category and related categories”. I followed this process from selective coding to theoretical saturation of categories and core category. Thus, my assortment of any of these categories or axial codes is not prearranged. I uti- lized axial coding in this study as a process to build up sub-categories. In this process, I looked for links for discovering sub-categories; and, for the most part, axial coding was employed to identify categories. At the advanced level of this stage, categories began to come in to view. Later, at a more advanced phase, selective coding was used to recognize the core category. Core category: Identifying a core category was the primary focus while I started “selec- tive coding”. It took place at the final phase of the coding process. During this phase, emerging categories were organized, processed and compared. Also, they were merged and put together through integration, refinement and unification. My intention in doing so was, as stated in Hjal- marson et al. (2007), to examine their characteristics, whether these were sufficient to explain the concepts under study. Several categories started to come into view. At this step, reviewing the data and re-coding them afterwards were required for the emergent categories. 80 The process of such constant return to the data aided me to develop sensitivity of emerg- ing insights and helped further to theorize on child poverty. Finally, a relationship was in view between the emerging theory and all other categories. At this final stage of analysis, the core category was developed (see also Figure 7). The selective coding and analysis continued until I believed that “the core variable, its properties, and its theoretical connection to other relevant categories were sufficiently elaborated and integrated” (Holton, 2007, p. 280). From open codes, axial codes, categories selective coding and core categories emerged from the analysis of the interview texts. The data has been analyzed by including many catego- ries, interactions, incidents etc.; these were coded during open coding. I frequently interrupted the coding as suggested by Strauss (1987) to write a theoretical note (e.g. to record comments, ideas, similar concepts and categories etc.) which was very useful later. This directs me as a researcher closer to grounded theory (see Berg, 2007, p. 318). I carefully kept in mind, as Strauss (1987, p. 32) indicates that variables must “earn their way in to the grounded theory”. Therefore, I never assumed the analytic relevance of any traditional variables such as age, sex etc. in this study of child poverty. Instead, if the data failed to support any variable of interest as analytically relevant, I accepted the result (see Berg, 2007, p. 318). Memos were written in a free way during the whole process of the data collection and the analysis (Hjalmarson et al., 2007). Furthermore, I frequently interrupted the coding or transcrib- ing to write theoretical notes (e.g. to record comments, ideas, similar concepts and categories etc.). These were very useful in the later part of the analysis to aid the theory building process. I also employed “constant comparison” method in which the researcher compares data with data, data with category, and category with category (Charmaz, 2005, pp. 518-19). I kept open the options to review and reframe the codes and emerging themes at a later stage of the analysis. Constant comparison method was followed during the whole process from open coding to selective coding. Mainly, three types of comparison were made in this process. First, data were compared with data for similarities and differences between “substantive codes and their proper- ties, later the categories with their contents” (Backman & Kyngäs, 1999, p. 151). This process aided me to code a category and thus the properties of that category began to develop. 81 Although categories were initially developed from the close contact with the data, a high- er level of abstraction was achieved through the process of constant comparison (Dey, 2007, p. 168). Glaser and Strauss (1967, p. 36) explain categories as “conceptual elements of theory” (see also Dey, 2007, p. 168). Second, for the theoretical elaboration and saturation of concepts, I made comparison between “emerging concepts and more data to generate new theoretical properties”. Finally, developing concepts were “compared to each other to set up the best fit between potential concepts and a set of indicators” (Holton, 2007, p. 278). As I proceeded with constant compari- son, a core category was surfaced. My approach was, as Glaser (2007, p. 99) pointed out: The general implications of a core category lead to the need for generating a formal theory of the core by looking at data and other studies within the substantive area and in other substantive areas using the conceptualizing constant comparison method. The reanalysis of analyzed data are done to make constant comparisons for saturating the gaps among categories that appeared (see Hjalmarson et al., 2007, p. 237). All the data was reanalysed after the last group discussion. My intention in doing so was to reach at a theoretical saturation by constant comparisons. For theoretical saturation of the emerged categories and core category, as Glaser and Strauss (1967) discussed, I considered them saturated when gathering fresh data no longer sparked new theoretical insights nor revealed new properties of core theoreti- cal category. In other words, saturation was reached when “adding further to the data makes no difference” (Dey, 2007, p. 185). However, there is disagreement about the meaning of saturation (Charmaz, 2006, pp. 113-14). Theoretical saturation in this study refers to my, the researcher’s, “judgment that there is no need to collect further data” at the time when I terminated data collec- tion (Wiener, 2007, p. 306). The data resulted from a unique interaction between me as the moderator and the groups and the participants in each group. This interaction provides a sound basis for the interpretation of focus groups. My involvement in the data collection process helped me to think about the analysis during analyzing my own data. In this study, I myself worked as a researcher and moderator. I also analyzed the emerging data of focus group discussions. This provided me with more insight and knowledge about the research overall and enabled me to establish a variety of important links 82 between the research questions and the collected data (see Stewart, Shamdasani, & Rook, 2007). As a researcher and analyst, my role ranges from collection of raw data to the formulation of interpretative comments. Being the researcher and analyst of this study, I prepared an abridged transcript33 for data analysis based on relevant and useful portions of the discussion that developed on watching and listening to video recording of each focus group. Translations of the quotations from the original transcripts are being reported to illustrate the findings. Selecting illustrative ex- amples is needed for the interpretation of participants’ statements. I aimed to use different quota- tions corresponding to the research questions and to the categories that emerged from the coding of data. For reliability, separate questionnaires have been used that include interviewees checking relevance to the purpose and research questions. Also, focus group interviews have the advantage of spontaneous cross checking for reliability among the group members during the discussion. There are limitations of focus group discussion in the question of generalizability of the findings, since the numbers of participants are small and not statistically representative of any group (see Sheldin & Shulman, 2004). However, generalizability of findings is not my purpose. Instead, my purpose is to bring into discussion a range of relevant issues and to obtain in-depth and detailed understanding of child poverty in Bangladesh, as is done in most qualitative research methods. In this study, our results have been complemented by quantitative study, which is rep- resentative at national level. 33 Kruger and Casey (2000, p. 131) suggest that abridged transcript should be prepared by someone who pos sesses thorough understanding of the purpose of the study. 83 Chapter 5 Child poverty in Bangladesh: A study based on survey data of the year 2000 5.1. Introduction The severe consequences of growing up in a poverty stricken environment have drawn widespread attention. Evidence confirms that children are much more vulnerable to poverty than adults since its multidimensional impact on children has long-term effects and can be transmitted to the next generations (Barrientos & DeJong, 2006; Boyden & Cooper, 2009; Harper, 2004; Moore, 2001). However, compared to poverty for adults, child poverty has received less attention than it deserves, particularly in developing countries (Minujin et al., 2006) such as Bangladesh. In Ban- gladesh, a significant number of the poor are, as shown in this chapter, children who live in poor households, a fact that casts doubt upon the recent well-documented improvement in poverty of Bangladesh (Matin & Hulme, 2003; Sen & Hulme, 2006). Unfortunately to the author’s knowl- edge no study has attempted to count the poor children in Bangladesh. This chapter aims to create a poverty profile of Bangladesh’s children to draw attention to the extent and characteristics of child poverty. That information is needed to develop plans for remedying the situation. The research questions that are addressed in this chapter are: In the Ban- gladeshi context, what are the extent and characteristics of child poverty? What factors determine the probability of a child being poor? This study uses the national microdata of the Bangladesh Household Income and Expenditure Survey (HIES)––2000 (see Chapter 4 for data description). As child poverty variations are found on several dimensions, multivariate analysis has been ap- plied. There is no official measure for assessing child poverty in Bangladesh. Hence, the follow- 84 ing method is adopted. Children are defined as individuals aged 14 years or less (for a definition of “child” details, see Chapter 4). As a measurement of child poverty, I applied the International Poverty Line defined by the World Bank. This line is: equal to $1.08 a day in 1993 PPP terms (referred to as “$1 a day”). This line has a similar purchasing power to the $1 a day line in 1985 PPP prices, in terms of the command over domestic goods (World Bank, 2001, p. 17). Thus, a child in poverty is defined as one living in a household with a disposable income per per- son per day lower than $1 PPP in 1985 prices. As discussed earlier, the poverty line is calculated at Tk. 500.16 per capita per month estimated in terms of PPP in 1985 prices for the Bangladeshi Taka. An alternative poverty line, $2 PPP, calculated at Tk. 1000.32 by doubling the $1 per person and day poverty line is used to examine the robustness of the results. However, the international poverty line is increasingly becoming a centre of controversy for being arbitrary and unrelated to a clear conception of poverty and for the way PPP is presently being calculated (Reddy & Pogge, 2002, 2005, 2010; Reddy, 2009). In response, although Raval- lion (2010a) acknowledges the conservativeness of the definition, he rejected Reddy and Pogge’s critique on the ground that they oversimplified the problem of measuring poverty and that they exaggerated faults in the Bank’s methods. In addition, he strongly condemned their proposed alternative method for creating an international poverty line ( see Chapter 2 and Chapter 4). Turning to the issue of intra-household allocation, all poverty measurements based on conventional survey data share a limitation that there are problems in knowing about the process by which time, money and other resources are allocated within households (Haddad et al., 1997). The assumption of this study is: if income of the household is below the poverty line, all children living in the household are deemed poor. In reality, some children in non-poor households have low consumption and in some poor households some children have high consumption due to dif- ferences in age, sex and household composition (Gustafsson, 1995). However, the appropriate estimation of intra-household allocation from conventional household income data is regarded as not achievable. 85 Since all poverty lines are to some extent arbitrary, an income based International poverty line is preferred for measuring child poverty in Bangladesh, taking into consideration that: data has been used of the year 2000 when this line was widely used and xxxxx it was designed as a poverty measurement for the poorest and developing countries (Deaton, 2006; Ravallion, 2010a). In addition, even though this specific line is currently controversial, it is used in the absence of a better alternative. 5.2. Description of child poverty The following section presents the relationship between child poverty and socio-demographic variables such as geographic location and the characteristics of the head of household in Bangla- desh using bivariate analysis. At the national level almost thirty per cent of all children live under the International Poverty Line, compared with just over twenty percent of adults (Table 8). Almost forty percent of the total population of Bangladesh is composed of children, and nearly half of its poor are children. Not surprisingly, Bangladesh has similar trends in high child poverty rates to other countries that have a large proportion of children in their population (Gor- don et al., 2003). This indicates that a large number of the country’s young population is at risk of the consequences of poverty. The child poverty rates more than doubled in all cases when the poverty line was changed from $1 a day to $2 a day (Table 8). The results presented in the rest of the chapter are based on a $1 a day poverty line, with the exceptions indicated. Table 8. Child poverty rates, the compositon of the poor and the entire population of Bangladesh in the year 2000 at national, rural and urban levels (percent). Poverty rates Composition of poverty Composition of the population in Bangladesh National Rural Urban National Rural Urban National Rural Urban Using $1/day poverty line: Children Adults Total 28.0 20.4 23.4 33.9 26.5 29.5 13.8 8.3 10.3 46.9 53.1 100.0 46.8 53.2 100.0 47.8 52.2 100.0 39.1 60.9 100.0 40.7 59.3 100.0 35.7 64.3 100.0 Using $2/day poverty line: Children Adults Total 72.8 61.2 65.8 80.2 71.4 75.0 54.9 41.2 46.1 43.3 56.7 100.0 43.6 56.5 100.0 42.5 57.5 100.0 39.1 60.9 100.0 40.7 59.3 100.0 35.7 64.3 100.0 Source: Author’s estimation from Household Income and Expenditure Survey—2000, conducted by Bangladesh Bu-reau of Statistics. 86 The difference between rural and urban poverty rates is significant in Bangladesh. The poverty rate in the whole rural population is about triple the rate for the whole urban population, while the rural child poverty rate is more than double the rate for urban children (Table 8). These variations are probably mainly the outcome of the country’s traditional agri-based economy that offers the rural population fewer alternative sources of income and lower earnings than the urban population (CPRC, 2004-05). As expected, a comparison of child poverty rate among divisions shows spatial variation, and this has been reported in Table 9. The great variation between the divisions of Bangladesh in terms of area, population, and socio-economic conditions are possibly some of the main reasons for spatial variations in child poverty. The highest child poverty rates are found in Rajshahi, fol- lowed by Dhaka, Khulna and Barishal, with the lowest in Chittagong division. The high rates for Rajshahi can easily be explained. Frequent flooding, an underprivileged tribal population, along with the other social problems such as overpopulation, the dowry system, polygamy, illiteracy, and so on, aggravate the problem. In this region, natural disasters are compounded by the effect Table 9 . Child poverty (using $1/day poverty line) in divisions of Bangladesh in 2000). Poverty rates (percent) Population of different divisions All Rural Urban Chittagong: Children Adults Total 19.6 13.9 16.4 24.3 18.1 20.8 7.1 4.3 5.4 Dhaka: Children Adults Total 28.2 19.1 22.5 35.9 26.4 30.2 14.0 9.2 10.8 Khulna: Children Adults Total Rajshahi: Children Adults Total Barisal: Children Adults Total 26.7 19.7 22.2 42.6 32.3 36.3 21.6 14.6 17.3 32.5 26.2 28.5 48.1 39.0 42.6 25.1 18.5 21.1 16.6 9.3 11.8 23.0 12.7 16.4 10.9 4.8 7.0 Source: Author’s estimation from Household Income and Expenditure Survey—2000, conducted by Bangladesh Bu-reau of Statistics. 87 of the “Farakka Barrage” that has a direct and devastating impact on its agri-based economy. This controversial dam was built during the 1960s on the Ganges River in India. Heavy, flash flooding is caused when the dam’s surplus water is released and drought when water flow is limited during dry seasons. Under these conditions marginal farmers become landless during the floods or have desert land during dry seasons and more households tend to move into poverty, including their children, as a consequence. Dhaka and Khulna have the second and third highest child poverty rates. Both divisions have more economic advantages than Rajshahi. For example, Dhaka, the capital of Bangladesh, is the centre of industrial, commercial, cultural, educational and political activities in Bangladesh. Khulna possesses industries such as shipbuilding and it has the country’s second seaport, which contributes to the division’s economic strength. Chittagong and Barisal show similar child pov- erty rates. Chittagong has the advantage of the country’s largest seaport and remarkable industrial development, including large and heavy industries. However, overpopulation and unemployment continue to stifle economic growth in all divisions. While there is considerable geographic variation in child poverty, Table 10 shows no no- ticeable variations according to religion and sex. The higher rates were found for the age groups up to 10. These rates are higher for the children in households with 5 to 8 members. Above this level, child poverty decreases at all levels, possibly because of the increased number of earners in larger households. 88 Table 10 presents child poverty by different characteristics of the heads of households. Child poverty has been observed to rise as the level of the household head’s education falls. The poverty rate is higher in rural areas than in urban areas when the household head has no education. This re- lationship is reversed when household heads have the highest level of education. Evidently, little or no education limits access to permanent, salaried employment with permanent, higher wages; it only allows access to temporary jobs (informal and casual) with lower earnings (Corcoran & Chaudry, 1997). The data shows that variations in the child poverty rate are related to the household heads’ occupation. Around half of the households which suffer child poverty have a head who is wage employed and this is similar to the proportion for the whole population. The child poverty rate is the highest for rural children whose household heads are wage employed, followed by the agri- cultural self-employed. This is expected since work is not always available in the traditional agri- culture-based economy; poor households with low education are more likely to lack the required Table 10. Child poverty (using $1/day poverty line) according to household head’s characteristics in Bangladesh in 2000 at national, rural and urban level (percent). Household heads’ characteristics Poverty Rate Composition of poverty Composition of population National Rural Urban National Rural Urban National Rural Urban Gender Male Female Total 28.0 29.2 28.0 33.8 35.9 33.9 13.6 16.3 13.8 93.3 6.7 100.0 93.7 6.4 100.0 91.1 8.9 100.0 93.5 6.5 100.0 94.0 6.0 100.0 92.4 7.6 100.0 Employment type Wage Self (non-agriculture) Self (agriculture) Nonworking Total 31.3 18.5 31.3 24.8 28.0 40.8 24.2 32.2 29.2 33.9 14.2 11.1 18.8 16.2 13.8 46.8 13.9 31.9 7.4 100.0 45.8 12.0 35.6 6.7 100.0 53.3 25.5 9.6 11.6 100.0 42.0 21.1 28.6 8.4 100.0 38.0 16.8 37.4 7.8 100.0 51.6 31.5 7.0 9.8 100.0 Marital status Married Unmarried Divorced, widow, etc. Total 27.6 29.8 36.0 28.0 33.4 40.7 43.5 33.9 13.5 4.0 21.3 13.8 93.0 1.2 5.8 100.0 93.3 1.3 5.4 100.0 91.6 0.3 8.1 100.0 94.4 1.1 4.5 100.0 94.7 1.1 4.2 100.0 93.6 1.1 5.2 100.0 Education No literacy Class 1-5 Class 6-10 Class10+ Total 35.8 21.5 18.5 3.4 28.0 39.6 27.5 25.1 5.9 33.9 22.8 7.1 8.4 1.4 13.8 74.4 11.9 12.9 0.8 100.0 74.4 12.6 12.3 0.7 100.0 74.3 8.1 16.2 1.3 100.0 58.3 15.6 19.5 6.7 100.0 63.8 15.5 16.6 4.1 100.0 44.8 15.8 26.6 12.8 100.0 Access to land No access Have access Total 28.1 27.9 28.0 38.6 29.4 33.9 13.7 14.4 13.8 60.6 39.5 100.0 56.3 43.7 100.0 86.3 13.7 100.0 60.4 39.6 100.0 49.5 50.5 100.0 86.9 13.1 100.0 Source: Author’s estimation from Household Income and Expenditure Survey—2000, conducted by Bangladesh Bu-reau of Statistics. Unit of analysis is children. 89 skills for most non-farm activities and only have access to low paid wage employment (Rahman, 2004). The poverty rates tend to be the lowest at all levels for children in households headed by non-agricultural, self-employed workers. Generally, land has traditionally been the most important source of income in rural Bangladesh. Rural child poverty rates are observed higher among those who do not possess any kind of land holdings, and more than half of poor children live in these households. Families headed by divorced/widowed/separated people have the highest child poverty (rural and urban), and the rates for female-headed households show that the variation only slightly depends on the gender of the head of household. Most children live in male-headed and married families according to population and poverty composition. 5.3. Probability of a child being poor The following discussion describes the characteristics are associated with the probability of a child being poor in Bangladesh. Multivariate analysis was used in this study because it is possible to control for several variables that may be associated with poverty simultaneously. It was consid- ered useful to make use of logistic regression analysis to analyze a child’s dichotomous poverty status. To reach a preferred model, different models were tested using different specifications. Here an estimate of a “full” regression model, containing all the variables of interest, is reported. The model contains a dependent variable, a child’s dichotomous poverty status having two values (1 for “poor” and 0 otherwise) and a number of independent variables that include location of residence, division, household size, age group, household head’s education, employ- ment type, gender, marital status, religion, and land access. Table 10 reports the results of calcu- lating the odds ratio: the first column refers to all children, rural and urban, while the second and third columns refer to rural and urban children separately, applying the model to the two groups separately. 90 The variable “Land access” is a continuous variable and the unit of land is hectare. The other variables are categorical. The odds ratios in the first column shows that the variables that increase the probabil- ity of a child being poor are those in which the household head has no education, and lives in a rural area in the Rajshahi division. The coefficients are highly significant ((p-value .000). The other variables that increase the odds ratios are where the head of household is wage employed, divorced or unmarried and Muslim and where the household has 5 to 10 members and the child Table 11. Odds ratio for a child being poor. Source: Author’s estimation from Household Income and Expenditure Survey—2000, conducted by Bangladesh Bu-reau of Statistics. Notes: Statistically significant at 0.000 Variables All children [95% Confidence interval] Rural children [95% Confidence interval] Urban children [95% Confidence interval] Location of residence: Rural 3.07 2.76 3.41 n/a n/a Urban 1 Divisions: Dhaka 2.01 1.81 2.24 1.93 1.71 2.17 3.00 2.26 3.96 Khulna 1.99 1.73 2.29 1.75 1.50 2.06 4.00 2.88 5.54 Rajshahi 3.71 3.32 4.15 3.59 3.18 4.06 4.49 3.32 6.07 Barisal 1.31 1.12 1.52 1.19 1.01 1.41 2.32 1.53 3.52 Chittagong 1 1 1 Age group: Age group 9-14 .80 .74 .87 .78 .72 .85 .88 .73 1.06 Age group <=8 1 1 1 Household heads’ educational level: No education 11.90 8.38 16.89 9.97 6.65 14.96 22.74 11.17 46.28 Primary education 6.76 4.71 9.71 6.54 4.31 9.92 5.78 2.70 12.36 6 to 10 years schooling 5.94 4.14 8.51 5.73 3.77 8.70 6.58 3.17 13.67 >10 years schooling/others 1 1 1 Land access: Have land access .69 .63 .76 .66 .60 .73 1.00 .73 1.38 Do not have land access 1 1 1 Households’ gender: Female .96 .77 1.19 1.05 .83 1.33 .61 .37 1.01 Male 1 1 1 Household heads’ marital status: Unmarried 1.75 1.20 2.55 2.05 1.37 3.08 .62 .14 2.68 Divorced/widowed/separated 1.74 1.39 2.18 1.74 1.35 2.24 1.90 1.15 3.19 Married 1 1 1 Religion: Non-Muslim .72 .61 .84 .76 .64 .89 .48 .28 .81 Muslim 1 1 1 Household size <=4 .66 .54 .82 .63 .50 .79 .82 .47 1.41 5-10 1.35 1.12 1.63 1.35 1.10 1.65 1.33 .80 2.22 >10 1 1 1 Household heads’ employment type: Wage employment 1.77 1.58 1.98 1.99 1.75 2.27 1.21 .97 1.51 Self-employment-agriculture 1.53 1.35 1.74 1.62 1.41 1.86 1.63 1.09 2.44 Non-working 1.39 1.16 1.67 1.33 1.08 1.65 1.78 1.22 2.58 Self-employment non-agriculture 1 1 1 Number of observations 15071 10683 4388 Prob> chi2 0.0000 0.0000 0.0000 Pseudo R2 0.1250 0.0876 0.1328 Poverty rates 28.04 33.90 13.76 91 is less than 8 years old. The results indicate that an additional hectare of land in the household decreases the probability of a child being poor. The base category for division is Chittagong; the poverty probability for an otherwise identical child living in Rajshahi is nearly four times higher. Dhaka, Khulna and Barisal divisions follow this. The first column shows that rural location significantly affects the odds ratio. This means that other variables included in the model do not fully explain why rural children have a higher probability of being poor than urban children do. To illustrate the magnitude of this effect, the pre- dicted probability of being poor was calculated from the estimated equation for typical children in different locations (see Table 12). In this Table, each number shows the probability of a child being poor when division, rural / urban and education of household varies. The “default” child lives in an urban household of Chittagong division where the household head has with 6-10 years or more or other schooling, is wage employed, male, married and Muslim, and the household size is of 10 or more members and the child is in the age group 9-14. The predicted probability of be- ing poor for this child is only 2.6 per cent. Table 12 . Probability of a child being poor in rural and urban location of different divisions with different levels of household head’s education. Probability of a child being poor (percent) Divisions Chittagong Khulna Rajshahi Barisal Dhaka Households’ Characteristics Urban Rural Urban Rural Urban Rural Urban Rural urban Default Default but with no education Default but with 1-5 years schooling 2.6 8.4 2.3 41.0 31.3 26.8 8.5 58.8 48.3 29.1 9.4 32.1 23.7 17.5 5.1 43.3 33.4 21.5 6.5 Source: Author’s estimation (specification reported in Table 10) from Household Income and Expenditure Survey— 2000, conducted by Bangladesh Bureau of Statistics. Notes: Default: household head—with 6-10 years or abov e or other education, wage employed, male headed, married and Muslim; and the household size is of 10 and above members and the child is in the age group 9-14. 92 In the second row of Table 12, the household head’s education level is the lowest (no education), while in the third row the education level is up to five years of schooling. The impact of education can then be clearly seen. Comparing the results, it is found that the probability of child poverty increases to 8.4 per cent if the household head has no education in urban Chittagong. The child’s probability of being poor based on living in a particular division (the results presented in different columns) in rural and urban areas was calculated (as described above). By changing a child’s residence from urban Chittagong to rural Rajshahi, the probability of a child being poor is 58.8 percent. In such a case, the probability of a child being poor in- creases more than 22 times (58.8/2.6=22.63). Similarly, If the child’s residence is changed to rural Khulna the probability is 41 percent, which is more than 15 times higher (41.0/2.6=15.77) if the household head has no education. Thus, child poverty clearly depends on where the child lives. In particular, in a household where the head has no education, the probability of a rural child being poor is higher in comparison to an urban child. 5.4. The notion of well-being and capability Although income/consumption-based poverty research in developing countries is well-estab- lished, an alternative criterion, “the notion of well-being and capability”, is increasingly being used by social scientists (Lyytikäinen et al., 2006; Sen, 1993; Sen, 1999). This approach refers to a broader set of circumstances for assessing poverty, including sanitation, healthcare, access to safe water and electricity, and so on. Some of these variables were examined; the results are presented in Table 13 and discussed in the following section. 93 Very few children live in dwellings with brick walls and concrete roofs. Non-poor children fare far better regarding shelter. The risk of ill health is worsened by poor access to water and sanita- tion.Children are at high risk as they or their care takers are rarely covered by insurance against risks of sickness, floods, crop damage and fluctuations in prices (Amin, Rai, &Topa, 2003). In Bangladesh, a low proportion of poor children have access to sanitary toilets, particularly in ru- ral areas. NGOs and UN organizations, together with Bangladesh’s government, have made re- markable efforts to ensure safe drinking water is available through tube-well, as reported in dif- ferent World Bank reports (World Bank, 2007). The result also shows that most children have access to tube-well water. However, the results are worse for poor children’s access to Tap water at all levels. Unlike many developed countries, households’ access to Tap water (e.g. municipal water supplies, which is organized and protected) in the country is very limited. This indicates that children’s lack of access to Tap water, which is much worse for poor children, is not favour- able for good sanitary practices in Bangladesh. Table 13. Selected indicators of living condition for poor and non-poor children in Bangladesh at national, rural and urban levels (percent with access) Children Acc e s s to dwell in g s Sanitary access Access to drinking water Access to Electricity with brick walls and concrete roofs Tap watera Tube well Other sources Poor Rural Urban All 0.3 6.9 1.1 4.9 12.1 5.9 0.3 8.3 1.4 95.3 89.9 94.5 4.5 1.8 4.1 6.5 46.9 12.2 Non--poor Rural Urban All 2.0 29.8 10.7 14.3 33.0 20.8 0.3 26.0 9.3 95.9 72.9 87.9 3.8 1.1 2.9 24.1 77.5 42.7 Source: Author’s estimation (using $1/day poverty line) from Household Income and Expenditure Survey—2000, conducted by Bangladesh Bureau of Statistics. Note: a: Organized and protected water supply through taps for a community, e.g., municipal water supplies. 94 Very few poor Bangladeshi children have access to electricity. Children need adequate and nu- tritious food, appropriate clothing, a proper shelter to live in, healthcare and education for their well-being. Benefits and allowances directed to children have been shown to fulfil these sorts of needs, and to keep child poverty rates low, in developed countries (Bradbury et al., 2001). As in many other developing countries, poor children lack this kind of support in Bangladesh, which is very crucial for their development and well-being. There have been some policy responses in order to increase education and improve child health- care in Bangladesh. Despite limited success, they have been criticised for reaching only a frac- tion of the poor population, and for excluding those most in need of assistance, the vulnerable poor (Amin et al., 2003). Limited resources, the nature of policymaking and budgetary priori- tisation of the state, including lack of good governance, might be some of the possible reasons why state initiatives do not achieve the required well-being of children in Bangladesh (Sen & Hulme, 2006). 5.5. Conclusion The contribution of this chapter is that it has attempted to quantify child poverty in Bangladesh and analyze how it varies with different variables. No previous study has attempted to do this as comprehensively. The study was based on national microdata of HIES—2000. It is important for future research to study the development of child poverty in Bangladesh over time by including data from the 1995, 2000 and 2005 surveys. The main findings are that the country has a large number of young people under the age of 14, who constitute a large proportion of the total population. The poverty rates for children are higher than for adults in all cases. Secondly, significant spatial variations are found in child poverty across different divisions of the country as well as between rural and urban areas. Ra- jshahi has the highest child poverty rates. The probability of poverty for many children depends significantly on where their household is situated. Thirdly, child poverty varies significantly ac- cording to the household head’s educational level, particularly in rural Bangladesh. With little 95 or no education, people only have access to temporary jobs with low earnings in both wage employment and self-employment in agriculture, which makes the entire household, including the children, fall below the poverty line. Lastly, this study found that very few poor Bangladeshi children, especially in rural areas, have access to electricity. Not surprisingly, most poor children live in poorly constructed dwellings without proper sanitation, in conditions that do not promote healthy child development. The present study indicates serious concerns about a significant number of children grow- ing up in poverty. In addition, different factors that have been found to be closely associated with child poverty may not be the only explanation. This statistic does not provide any obvious addi- tional insight into initially posed second, third and fourth- questions in Chapter 1. The results found from the present study indicate to the risk that there will be a similar number of poor adults in the future, and similar consequences can be expected for children living in poverty in Bangladesh. These children will more likely thereafter to pass poverty on to their own children and their children’s children and so on. There need to be interventions to interrupt this process. However, drawing a conclusion on the existing policy interventions and the benefit received by the poverty affected children are beyond the scope of the data used in this chapter. Furthermore, the present study has limitations to suggest what could or should be done to inter- rupt processes of IGT of poverty. In addition, the study presented in this chapter does not grasp or include child poverty’s victims’ viewpoint from their own experience. For many analysts, policymakers and social scien- tists, it would be very useful to listen to the voices of the victims of child poverty. A qualitative study––based on children’s experiences told in their own voices–– offer additional insight into the matter for details. Furthermore, the questions that were posed in the beginning of the series of studies presented in the dissertation remained unresolved in this chapter. The next chapter will, therefore, proceed by presenting results from focus group discussions with poor children and women. In particular their understanding in relation to the above mentioned questions: 2(a) what is child poverty according to poor children and mothers (its principal victims) in Bangladesh? 2(b) How do the poor children and mothers (principal victims of child poverty) experi- ence and account for it? 96 Chapter 6: Child poverty in Bangladesh’s context: A qualitative approach 6.1. Introduction Quantitative research using survey data has limitations in comprehending the complete picture of poverty, even though it is suitable for the study of certain issues and qualitative study is better for some issues (Hayati, Karami & Slee, 2006; Narayan et al., 2000). Consequently, this chapter applies a qualitative perspective. The research questions addressed are: 2(a) what is child poverty according to poor children and mothers (its principal victims) in Bangladesh? 2(b) How do the poor children and mothers (principal victims of child poverty) experience and account for it? To carry out this research, I used focus group interviews. These have the advantage of producing a wealth of detailed data from a smaller number of people, and are increasingly used by child poverty researchers (e.g., Fortier, 2006). Applying a “grounded theory approach” for data analysis makes it possible to interpret and understand child poverty across various dimensions. Issues relevant to child poverty, as conceived in focus group discussions of this study are not very different from child poverty in other parts of the world, as is evident from a number of studies (Chant, 2006; Kalaycioğlu, 2006). What is distinctive, however, as illustrated in this study, is that whereas “very extreme hardship such as starvation, miserable living condition, etc., are very uncommon in advanced industrialized societies” (Iceland, 2006, p. 26), the opposite is true for Bangladesh and many other developing countries. Parental poverty, insufficient income to meet children’s basic requirements, and gender 97 discrimination are some of the factors that are found to contribute, directly or indirectly, to child poverty. Poor child rearing capabilities constitute obstacles for children’s development, for ex- ample, in relation to health care and educational opportunities. Distinctively, this chapter is a step in the direction of qualitative approach inquiry, and seeks to detail some of the contributory vari- ables to child poverty, such as corruption, gender discrimination, and child abuse, which survey data has difficulty reaching. These factors are found to further aggravate an already dire situation by limiting poor children’s access to existing public services and facilities. The children’s suffer- ing from hunger and poor living conditions in their lives carries numerous severe consequences in the present and future. This indicates to an increased risk of continued poverty and a subsequent passing of poverty to the next generation. The structure of this chapter is as follows: in the next section the discourse on child pov- erty in the literature is discussed. This is followed by discussion of methodological issues and approaches to data analysis. Next, the chapter presents its results in the section entitled “A discus- sion of child poverty”. Finally, a summary and discussion of the results are presented. 6.2. Research approach As mentioned earlier, qualitative analysis has the advantages to inquire and offer additional in- sight into some of the issues. By comparing advantages and disadvantages of different qualitative methods, I chose focus group discussions, which I found as a better alternative in producing rich data on child poverty. The advantages (such as closeness to data, systematic concepts genera- tion etc.) of grounded theory approach34 inspired me to use it for analysing the data produced in the focus group interviews. Snow ball sampling was used to get access to informants. The study reported here used four focus group interviews to assess child poverty from the viewpoint of its poor children and poor women. As noted earlier in Chapter 4, I had time and budget limitations that resulted in that I kept my interest restricted only to listen to these voices (see Chapter 4 for details on methodological choices). By including men’s groups and other age groups, future 34 See Glaser and Strauss, 1967; Charmaz, 2005; Charmaz, 2006. 98 research would probably contribute more insight into the knowledge on child poverty literature. As mentioned in Chapter 4, 23 participants (19 children and 4 women) were recruited from three different slums of Dhaka for the four groups (Group 1-Group 4) (see Table 6). Children were chosen from the age group 6-14. The reasons for this age group are: firstly, children aged below six could not participate spontaneously in the discussion as I attempted to do so. Secondly, the upper age limit to be considered as children that has been fixed for all the studies in this dis- sertation is 14 years. The ages of women were not restricted; however, having a child or children was one of the main criteria for selecting women participants. Also, highly interested potential participants were given preference to consider in selecting women participants in Group 4.Thus, the age group for the selected women remained around 20 to 45 for the women participants for Group 4. 6.2.1. The Approaches to Analysis The data from focus group discussions (FGDs) were analysed by a grounded theory ap- proach, which proved to be very useful for contextualizing child poverty in this study. Differ- ent analytical tools such as coding, constant comparison method and memos were used for the analysis of data. Using diagrams (e.g., figures) provided a visual representation of categories and their relationships while constructing the analysis (see Charmaz, 2006, p. 117). My intention with utilizing open coding was to identify topics of interest in my study. The “Open code” columns of Figures 5 and Figure 7 provide examples of a summary of open coding. The process started with line-by-line coding of extracted transcripts of FGDs. An example of line-by-line coding is presented in Figure 4. Figure 4. Line-by-line coding of a participant’s statement. Source: drawn from focus group discussions of women’s groups. Interview statement Line-by-line coding Our financial state is bad and cannot educate our children. We are not even educated ourselves. Had our parents educated us then we could’ve had a better job. Describing parental low earnings Articulating educational expenses Noting not being educated/lacking education Expressing deprivation of and eagerness to education Stating missing opportunity 99 I created certain labels as open codes for different concepts of FGD transcript. An example of this is presented in Figure 7. As I continued, certain codes (labels) commenced to reappear. From the RSHQFRGHVD[LDOFRGHVDQGFDWHJRULHVZHUHGHYHORSHGDQG¿QDOO\WKHFDWHJRULHVWKDWHPHUJHG were integrated to core categories (see Figure 6 and Figure 7). Figure 6 provides an example of D[LDOFRGLQJIURP)*'VVWDWHPHQWVWR¿UVWOHYHO RSHQ DQGWKHUHDIWHUVHFRQGOHYHO D[LDO FRGHV (or sub-categories). )LJXUH7KHFDWHJRU\³(QFRPSDVVLQJPXOWLGLPHQVLRQDOLW\´                    Source: drawn from focus group discussions of children’s and women’s groups. Axial codes (Sub- categories) “If our parents earn more, then we can go to school and get jobs, become rich (in future) (laughs)”. Parental low earnings High ambition, deprived of education “If the income is more than we won’t have as many problems getting by”. Emphasizing Parental incapability Expressing aspirations and the likeli- hood of them being realized Selective codes  …‘’ƒ••‹‰ —Ž–‹Ǧ †‹‡•‹‘ƒŽ‹–› “When we are sick we can buy some medicines. But not the expensive ones. We cannot eat what we want to eat. Like fruits”. “If my father could have a proper work, he would earn more money, shop groceries, we could eat. If my father can earn we can otherwise not”. “They (parents) took me there to get me admit- ted but it cost too much money!” “Children have a wish to grow up, educate themselves and make it big for themselves. We couldn’t even say such things (meant in their “My daughter was only 1 years old. She had many diseases but I couldn’t get treatment for her”. “‘We have no land. Can’t give us food or clothes. Can’t take a proper shower”. “We have to sometimes go to school without having eaten anything”. Focusing suffering of deprivation Open codes 1RQIXO¿OPHQW of basic needs'  Statements from focus group discussions 100 The next section is about the result from the data analysis of FGDs’ transcripts. I begin with pre- senting the discussion on the data analysis of the categories (which includes discussion on open codes as well), followed by the core category. My approach here, as Glaser and Holton (2004) explain, has several fundamental analytic rules, where they claim that sorting for writing up can start anywhere. They describe: It will force its own beginning, middle, and end for writing. Once started, analyst soon learns where ideas are likely to integrate best. ……… This rule forces focus, selectivity and delimiting of the analysis. [p. 72] In the two qualitative studies presented here (in Chapter 6 and Chapter 8), I started by presenting categories or core categories and vice versa (that are emerged from open codes and axial codes) as they fit best to focus, select and to set the limit of the analysis. 6.3. A discussion of child poverty Three categories emerged from focus group data analysis using a grounded theory ap- proach. These are: “encompassing multidimensionality” (Category A), “spending life in distress” (Category B), and “having intergenerational and gendered dimension” (Category C) (see Figure 5 and Figure 7).The description of empirical findings is presented in the following sections accord- ing to the above mentioned categories. Figure 6. Summeries of categories and axial codes. Category A: Encompassing multidimensionality ● Axial code 1:Focusing suffering of d eprivation ● Axial code 2: Emphasizing parental incapability ● Axial code 3: Expressing aspirations and the likelihood of them being realized Category B: Spending life in distress ● Axial code 1: Voicing hunger ● Axial code 2: Stressing poor dwellings ● Axial code 3: Articulating poor access to public services ● Axial code 4: Describing child abuse Category C: Having inter generational and gendered dimension ● Axial code1: Giving emphasis to parental poverty ● Axial code2: Highlighting insufficient income to meet basic requirements ●Axial code 3: Calling attention to gender discrimination Notes: Axial codes are sub-categories. 101 6.3.1. Encompassing multidimensionality (Category A) This section covers a description of the category “Encompassing multidimensionality”, which stands for an account of child poverty as conceived by the participants. The starting point for the discussion of each focus group was the question “Whom do you consider to be poor children?” Child poverty was conceived to be multidimensional as perceived, experienced and understood by the participants. From line-by-line coding of extracted transcripts of FGDs, certain open codes started to reappear. These were non-fulfilment of basic needs, parental low earnings, high ambi- tion, deprived of education and so on (see Figure 7). At this stage, based on the relationships around certain groups of open codes, axial codes commenced to emerge. Thereafter, the concept of child poverty was surfaced from the three major emerging axial codes: 1) focusing suffering of deprivation; 2) emphasizing parental incapability; and 3) express- ing aspirations and the likelihood of them being realized (see Figure 5, Figure 6 and Figure 7).The suffering of “deprivation” of basic necessities was the main focus in the children groups. The women’s group emphasized on “parental incapability” related to low income or lack of resources for child development; and, all four groups discussed their “aspirations and the likelihood of them being realized”. focusing suffering of deprivation: In their attempt to visualize the characteristics of poor children, all participants paid attention to the limited access to food, shelter and other very basic human needs such as clothing, health care and education. As understood from their discus- sions, such deprivations were severe on the account of children’s suffering. The principal concern in this conversation was the anguish of living in such absolute or extreme level of poverty. For instance, one said: “When it rains the water leaks into our house”. Another said: “We wear torn cloths. There is water on the footpath when it rains, then there is no place to lie down on”. These are some examples of the frequently expressed concerns about suffering of severe deprivation for the children who grow up in absolute poverty. Unlike children who do not grow up in poverty, children living in absolute poverty, like the participants of this study, suffer in many aspects of lives. A clear agreement was not made on 102 who are absolute poor. Nevertheless, according to the participants’ description, child poverty in absolute term means: not having a satisfactory quantity of foodstuff to eat, having parental less capabilities to provide for their children’s basic requirements such as paying for accommodations, education, shelter, clothing, and health care. The following excerpt is an illustration of how a child (who lives in absolute poverty) defines their poverty status: I am poor because I don’t have money. We don’t have money. ‘Poor’ people are people like us. We live on the footpath, have nothing to eat, and lack everything…we cannot pay rent on time; the house-owners yell at us. Child poverty, as sketched by the aforementioned child, is “lack of everything”. This quote con- veys a definition of child poverty as deprivation that entails the household’s inability to pay for food and rent, as well as an absence of everything needed to fulfil basic human needs. This kind of deprivation offers nothing but huge physical and emotional suffering that includes frustra- tion, sorrow, humiliation and powerlessness. Thus, children’s accounts repeatedly expressed their limited access, or no access at all, to basic needs. However, what is striking here is that none of the participants claimed getting access to basic needs as their right to “be adequately nourished; be adequately sheltered; have basic education; be able to appear in public without shame” as de- clared by the Office of the High Commissioner on Human Rights (as cited in Minujin et al., 2006). The participants in particular were not even aware of children’s rights. The findings that indicate deprivation from a number of directions in poverty are consistent with the discussion among the poor women. One woman said: “We have problems with providing food and shelter, clothing, educating the children”. Turning to healthcare, the participants did not think that spending on healthcare for the poor children is compensated for by the public healthcare services. Responding to a discussion on healthcare, one woman asserted: “Treatment is something I have to pay for myself”. In this circumstance, reliance on market based healthcare services can cause health deprivation in the cases of poor families that already struggle to provide subsistence for their children. Thus, health deprivation (Gordon et al., 2003) increases the risk of causing the emergence of long-term pov- erty (Harper, 2004). As outlined above, child poverty is not only a matter of living in low-income 103 households. Rather, children can suffer severe deprivation in many aspects. emphasizing parental incapability: The discussions in the children’s groups primarily concentrated on describing deprivation experienced by the poor children in relation to parental incapability to meet their basic human requirements. Deprivation was found to be associated with monetary aspects of parental inability to provide for their children, including water and electricity. Also other factors such as parental access to or command over resources, education, earnings etc. act as “capability inputs” that a play vital role in producing wellbeing of children. For instance, the incapability of parents and other care givers were carried found to be resulted from their poor command over resources. The inadequacy in command over resources is produced from a range of channels, for instance, market (earnings and property income) and public provision or other non-market channels. Ill health and low education were noted from the discussion as impaired personal conversion factors to contribute to produce parental incapability. The participants were fully aware of this. This is clearly expressed in FGDs that parental incapability forces children to go through extreme hardship. For instance, when one child participant attempted to define poor children as: “The poor (children) are those that (whose parents) do not have money, wear torn cloths, have no supply water, and electricity”. Another child said: “Those that can’t eat twice a day. (Their) parents have a difficult time running the household on their money”. Not surprisingly, the state- ments outlined above viewed very usual characteristics to identify poor children. These include: not having an adequate amount of food the children are required to eat, a safe shelter to live in, appropriate clothing they need to cover themselves with, clean water for drinking and sanitation they demand, and access to electricity they necessitate. These statements indicated that child pov- erty and parental incapability are closely related. Surely, this is an issue, particularly for a child who resides in a society where public provision of welfare or other non-market sources does not work in the shortfall of market income. In such a case, due to inadequate market income, parents or care givers pass very tough time to manage sustainable livelihood to afford for their children. Women’s views on parental incapability were also associated with income. All the par- ticipants in the women’s group focused on parental incapability. This is indicated to the lack of 104 income or resources that deprives their children. These views show that one’s ability to provide for one’s children is also related to income and wealth. One woman participant defined child poverty as having low parental income: “Those (parents) who have a low income are called poor”. Although many of these parents do work, their earnings are very low. As in most cases, the lower parental capability (in terms of income and resources) the lesser is their command over commodi- ties that the children require in particular. In fact, this is particularly true for those who are in short access to financial or in-kind support from any other sources. Besides the children participants, women participants stressed parents’ frustration at be- ing unable to fulfil the basic need of their children, particularly to provide food, healthcare and to educate their children on account of poverty. In this context, while poor parents struggle to feed their children and pay rent for shelter, for obvious reasons, the women’s group expressed the parental priority for providing food rather than sending children to school. This is evident in the following statement: “I still cannot educate my younger daughter… I have to pay rent and pay for food so I cannot afford it”. In such a situation, acute suffering of deprivation in many aspects of life is one of the major concerns, in respect to child poverty, particularly in a society having poor econoomy. Although the participants from all poor groups viewed child poverty primarily as parental incapability resulting from a low income, they also viewed child poverty as an outcome of the low level of parental education. expressing aspirations and the likelihood of them being realized: Child poverty was found to be the barrier between children’s hopes or aspirations for the future on one hand, and the likelihood that they would be realized on the other. All the participants, particularly in the chil- dren’s groups, considered their social and economic potential, and agreed that education was the most important channel to realize their dreams. Despite parental incapability, data from the women’s group showed their strong desire to provide educational opportunities for their children, and even to provide a high quality education. For example, one woman said: “My baby is growing up. If we earned a little more we could edu- cate him... You understand the expense of raising children. They should be sent to good schools and given everything of high quality”. In the conversation another woman articulated children’s 105 wishes, saying: “Children have a wish to grow up, educate themselves and make it big for them- selves. ‘Mom, I will make all your sorrows go away when I have completed my education’, my daughter says to me”. Children’s ambition is also reflected in most of their high expectations regarding education and their stated dreams of becoming a doctor, an engineer, a teacher or even a pilot. Becoming a doctor was the most common ambition in the children’s groups, as one child said: “I want to be a doctor when I grow up”. Another girl expressed her aspiration to be a teacher in her adulthood, saying: “When I grow up I want to be a teacher”. Another girl said: “I want to be a pilot”. “If I could be the owner of Japan City Garden then I would be rich”, said one child par- ticipant, and another child, in a similar vein, said: “And saving money in the bank. And then find big connections and then own Japan City Garden. I will go straight from the bottom floor to the 16th floor”. Japan City Garden is a huge and very attractive housing complex in the fast growing mega city of Dhaka, comprising around 15/16 storied apartment buildings in Dhaka. The par- ticipants in this study live close to this housing complex in different slums. By being rich, they meant to move out of poverty, which seems to be only a dream to them. However, their dreams know no boundaries. One child said: “I want to buy the world!” (laughs). Unfortunately, the gap between the desire, hope, aspiration or ambition and the lack of realistic possibilities of achieving these goals are extreme. As a result, the educational potential of children remains unexplored and unrealized. From the participants’ perspectives, it is clear that child poverty is multidimensional and affects children in a number of ways. Child poverty, from the participants’ point of view, can be defined as a state of multidimensional deprivation regarding the children’s basic human needs, such as proper food, safe and secure shelter, clothing, and a minimum of healthcare and educa- tion. Most importantly, the definition of child poverty includes deprivation from a number of dimensions. As a result, children’s dreams cannot be realized, in most cases, such as one of the participants of this study who passes their lives in child poverty. The data analysis also depicts an understanding of poverty as a condition in which children’s potential and dreams cannot be realized, due to a mismatch between what parents want for the children, what children want for 106 themselves and what the parents are able to offer. 6.3.2. Spending life in distress (Category B) The final question that was discussed in the focus groups was: In what ways do the poor children suffer in poverty? I created certain labels as open codes for different concepts of FGD transcript such as starving, unsafe shelter, limited public health care and primary education, discrimination, early marriage and so on (see Figure 7). From open codes different axial codes (sub-categories) started to appear. For instance, four subcategories were developed: 1) voicing hunger, 2) stress- ing poor dwellings, 3) describing child abuse, and, 4) articulating poor access to public services (Figure 7). I grouped these sub-categories into a category in higher level group. At this point, the second category “spending life in distress” emerged from the above mentioned axial codes (see Chapter 4 for details on methodology). Thus, from the open codes, axial codes, and categories were developed. In this section, I present a discussion on the emerged Category B that includes three sub-categories as mentioned above. voicing hunger: Having an adequate quantity of foodstuff to consume or lacking capa- bility to pay for them is a common concern for many poor families, particularly in the developing world. In the focus group discussions, the poor children repeatedly referred to their helplessness in the face of hunger, which forced them to even eat inedible rotten food since starvation was the only other option. Some of their comments exemplify this: “We do not feel like eating bad food but there is no choice when we get hungry”, “We eat food from the street”, “We beg for food in the houses and we get old rotten food”, “They (the donors—from whom children collect food) only give us rotten food”. Such struggle for getting edible and sufficient food is very common in many underdeveloped countries among the poorest segment of society. Alternatively, residing in absolute or extreme poverty decreases food security. In such cases, the capability of parents or caretakers is very low to provide for their children’s basic survival. In essence, the powerlessness and vulnerability to hunger of both the poor children and their parents were reflected as in the explanation of one child : “When we don’t want to eat then our parents say ‘There is nothing else to eat. What can we do? Eat whatever is there’. If we don’t 107 want to eat even then, they scold us and make us eat”. Thus, the limitation of the parental ability to get their hands on acceptable food has been put across the discussions. Also, the participants’ food insecurity with hunger or their less accessibility to nutritionally adequate and safe food has been voiced. Being born and growing up with such limited access to right food puts children’s proper physical development in doubt. Their suffering from persistent hunger leads to malnutri- tion. This certainly appears to be the case for many poor children and might affect their growth and development. The more a child is deprived of basic needs, food in particular, the greater the risks of not getting a healthy and fully nourished child. Children’s multiple experiences of painful sensation of hunger due to persistent invol- untary access to inadequate food by no means could be desired. Like the children, the women participants also talked about children’s starvation when discussing economic hardship for chil- dren and even for their caregivers. One of them said: My husband had jaundice. We treated him. We did not know then that it would be so ex- pensive. Then we sold his house and land in the village to collect the money. My children starved in the house, there was no money for food, but the medicine had to be bought. My children starved. My money is gone, and so is my husband (crying). Several poor children and women recounted similar tragic tales. They spoke of their hardship in lacking access to food, healthcare and adequate housing. Thus, long-term illness or death of their care-takers put the children at the risk of falling into poverty. As I mentioned ear- lier, starvation or hunger restrict the poor children to take full advantage of child development and human growth tactics. This is because unfulfilled hunger harmfully affects health and labour productivity in the long run. Thus the poverty is carried on in their adulthood to the next genera- tions. Child participants expressed their experience of hunger as early as the morning, stating there was not even breakfast. This was even the case for the ones who attend school. One child participant at one point stated: “I go to school without having eaten anything. There usually isn’t any food then”. This unsatisfied hunger at this early stage of their lives carries the risk of severe consequences in children’s physical, psychological and human development at the present time, and most probably, if they manage to survive, in the future as well. This sort of deprivation among 108 children anywhere in the world is dreadful and totally unacceptable. stressing poor dwellings: According to the most participants’ description, living con- ditions (regarding access to shelter, water, sanitation etc. in particular) are found to be beyond imagination for many residents who are living in rich countries. It was evident from the discus- sions that some children’s lives are at stake because of unsafe shelter (for example, houses are made of polythene and have unsafe roofs etc.). One child, providing an example, said: “Our house got blown away in a strong wind. We make houses out of (polythene) paper. You know the blue polythene? If it rains the water drips, and storms can make the roof fly away”. This child’s description demonstrates that they live in inadequate and temporary built housing35. In view of the analytical framework developed in the statements outlined above, it would imply that the residences available for them are not safe and children do not have any choice but to live on such badly built house. Living in such poorly constructed house put the households at risk of injury, and health problems such as bronchitis. Unquestionably, the issue (of poor living condition) raised in the discussions offer an additional touchable sense of the lived experience of children living in poverty. In comparison with adults, children are more vulnerable in such cases in the underdeveloped communities of the world. In such poor living conditions, children even experience disturbed sleep. One child elaborated on their miseries and compared the standard of living of the poor and non-poor, saying: “We live in the water…the roof leaks when it rains. Then we have no place to sleep. You have a house and a car. If it rains it does not affect you. You can still live normally. We cannot”. In addition to physical health problems, children who are raised in such unhealthy condi- tions are at an increased risk of having emotionally damaging (such as anxiety and depression) behavioural problems (e.g., hyperactivity and aggression). Insecure residence is not only a threat and unfavourable to children’s wellbeing, but it unsettles children with their families, which is hardly encouraging for achieving formal education. It is reasonable to suppose that poor children 35 These were constructed with impermanent and basic available materials. 109 are more vulnerable compared to the other better-off children who live in safe shelter. As noted in the earlier chapter (see Chapter 5, Table 13), households’ access to Tap water (e.g. municipal water supplies, which is organized and protected) in the country is very limited. Therefore, a considerable amount of time is being spent for water collection from a long distance for daily use, which includes drinking water. In many underdeveloped countries, including Ban- gladesh, a substantial number of children in poor families are used for this purpose. This kind of household work particularly prevents many children from going to school. Moreover, the col- lected untreated water (except tube-well water) also risks children’s health problems, including life threatening diseases such as diarrhoea. Although Bangladesh’s recent high achievement in extending access to safe drinking water has been highly appreciated in many studies (World Bank, 2007), the children discussed their access to safe water still being limited. The lack of access to improved water (such as Tap water) means also lack of access to sanitation (toilets, water, sewage, drains and rubbish disposal) in many developing countries like Bangladesh. In Bangladesh, the water supply or infrastructure (e.g. sewage pipes and treatment works), needed to make sanitation services available, hardly reach the poorest segment of the so- ciety. Consequently, lacking access to piped water or in-residence pipes is not favourable for good sanitary practices in Bangladesh. Unhygienic sewerage and bad sanitation (e.g. overflowing and backed-up drains and toilets) are not only grotesque to children, these are also life-threatening36. An illustration of unhygienic sanitation is concealed in the following statement: “We have trou- bles in the bathrooms and the toilet”. Another child said: “When it rains and the drains overflow their bathrooms also overflow. Yac! (expression of disgust)”. As noted above, the adequate housing and access to improved water source37 for poor children are at greater risk. For instance, children living in badly built houses lack supportive facilities required for better educational accomplishment and good health. Children who grow up 36 Many “communicable” diseases, such as dysentery, diarrhea and cholera spread out due to unhygienic sew erage and bad sanitation. 37 Access to an improved water source refers to the percentage of the population with reasonable access to an adequate amount of water from an improved source, such as a household connection, public standpipe, borehole, protected well or spring, and rainwater collection. Unimproved sources include vendors, tanker trucks, and unprotected wells and springs. Reasonable access is defined as the availability of at least 20 li ters a person a day from a source within one kilometer of the dwelling. 110 in such living condition are at increased risk of a wide range of negative outcomes including ill health and less educational achievements. Consequently, they cannot come up to scratch to get hold of required proficiencies for the potential employment in future. This would imply that the likelihood of being unemployed, or of being employed in casual jobs in future increases. In both ways, the chances of building earning capability in the adulthood decline. Describing child abuse: Not only do poor children lack proper food and healthcare and live in miserable conditions, their lives are made even more distressing when they become the victims of mistreatment from the other children and adults. Even though the occurrences of child abuse are of a great concern in all socio-economic groups, children in particular are being vic- timized of different forms of abuse because of their poor socio-economic status in many cases. Among others, one child provided an example of how they are mistreated by the r better-off people: We suffer due to food and cannot live an easy life. If we try to do something, people scold us and say GO! GO! And drive us away. They say, “You are slum children. We do not want to play with you”. They call us bad names. The children’s descriptions revealed that discrimination due to their poverty is similar to racism, when people become the victims of discrimination due to the pigment of their skin, the state of their living conditions or their despised national identity. In some instances, discrimina- tion due to poverty is worse than racism. This was evident when a child spoke of his experience of segregation in his everyday life: “We cannot sit in the same place as rich people”. Child abuse due to poverty can take several forms. For example, poor children experience physical assault, violence, negligence, discrimination and humiliation in several cases at the hands of the non-poor, primarily classmates and teachers. Some examples were given by the child participants: “If we go near rich people they say ‘Go away’. They (non-poor) misbehave”. “They push the poor people and make them fall down, “They push us holding our necks”. Child abuse also leads to the feelings of frustration and unfairness, and ultimately to the poor children’s negative attitudes towards the non-poor. The poor children feel misunderstood and unfairly treated by non-poor children, something that could change if the non-poor should one 111 day become poor: “They (non-poor) don’t understand that we are equal. If they were also poor then they would understand. They (non-poor) will also become poor eventually”. Some children recounted their experiences of being beaten by teachers and neglected by non-poor peers due to their poor attributes–not wearing clean clothing, for example. One child described an experience: “When they (poor children) go to the school the children don’t let them sit…The teachers beat them…If you wear dirty clothes to school then they beat you, because they cannot keep their clothes clean”. Others in the same group supported this view describing similar experiences. In all the children’s groups the participants provided examples of how child abuse made children reluctant to go to school, promoting them to drop out. Unfortunately, being beaten by the teach- ers, guardians and elders is a social norm and acceptable in most parts of Bangladeshi society (Conticini & Hulme, 2007). Although the women’s group did not mention any physical violence towards their chil- dren, negligence and discrimination on the part of teachers and peers were discussed. As dis- cussed in Conticini & Hulme (2007), beating students in school is taken for granted. It is viewed as a means to manage the children who are unwilling to study. The mothers reported that the children become reluctant to go to school and often drop out as a result of being humiliated at school. One woman said: “My daughter says, ‘No, I cannot study like this. You cannot pay my fees in time. People taunt me in school’. And so I have stopped her education completely”. Thus, the poor children drop out of school and this results in an inadequate education. Furthermore, some parents send their children to work, which can be seen as another form of child abuse that results in children’s low education. Thus, child abuse for the poor children acts as an important contributor to low education. Articulating poor access to public services: The poor are much less likely in having benefit of accessing to public services than others. The public services include safe drinking wa- ter, sanitation systems, health care and education. Their vulnerability in the face of the shortfall in household income could be lessened if they had access to the existing public services (e.g., health and education sectors). However, children lack access to public services such as primary and sec- ondary schooling in Bangladesh even though the former is free of cost and the state policy is to 112 make it “universal”. A recent study by Cameron (2011) on choosing a primary school also points out that the delivery of services to slum populations including education are very limited. This has been reflected in this study as well. There are very limited opportunities–– in terms of either “financing education” in private schools or accommodating children in government schools with education free of cost”––for the poor children to be enrolled in schools that provide primary edu- cation in urban Dhaka. The reflections from the focus group discussion also support this fact as a result of which the participants claimed that some of them did have access to primary education while others did not.38 In contrast, all of them claimed to lack access to secondary education. This was reflected in our group discussions. To highlight disadvantage in accessing to public services such as education, one woman said: In the government schools, where poor people can send their children, they only teach un- til class 5 and then no more. We can’t send the children to other (private) schools either… They have high admission fees, expensive books. This woman’s frustration poses the question of whether, and to what extent the capability of children’s parents could afford their educational spending. From the discussants’ perspectives, two major challenges out of many are: 1) high prices of essential basic commodities (such as foodstuff) and available private services, and 2) less access to the current public services. In many developed countries (e.g., Sweden, Norway and other Scandinavian countries), state provides support to some extent to reinforce parental capabilities in such cases. Although all children’s entitlement to obtain education free of cost is their constitutional right in Bangladesh, to date, this fundamental right has not been executed for all children. As a result of which children––from lower socio-economic groups––suffer from a number of dimensions due to education depriva- tion throughout their adulthood. On top of that education deprivation not only has diverse impacts on individuals but also weakens the chances of escaping poverty in the long run. Despite this limited access stated above, some participants talked about corruption in public services. An example given was that school allowances for girls’ education (according to our participants) are allocated to rural girls only; however they are rarely available in reality. A 38 Also, there are many rural areas that lack required education institutions to provide education for all children. 113 woman participant said: “They say, ‘There is no money (allowances for girls’ education)’. They (personnel in charge) keep the money themselves in the name of admission fees”. Similar alle- gations were made against healthcare services. Most participants claimed that public hospitals, which are supposed to be free, are not free in practice. Several participants recounted examples of having to pay for hospital costs that should have been free. One child said: “It costs money to buy medicine. Also to pay the (personnel in charge)”. According to another boy: “If you go inside (to be admitted) Dhaka Medical (Hospital) it costs money”. Thus, claims of corruption were made in the discussions, which might be true, and if not true then show a level of mistrust that arises from a lack of transparency and accountability of public service providers. Either ways, poor children are denied access to a healthy development and higher educational attainment indicating to the risk of IGT of poverty, a topic that will be elaborated in section 6.4. 6.3.3. Having intergenerational and gendered dimension (Category C) Three sub-categories––giving emphasis to the parental poverty, highlighting on the insufficient income to meet basic requirements, and calling attention to gender discrimination––constitute Category C “having intergenerational and gendered dimension”. These emerged as important contributing factors to child poverty from the participants’ responses to the interview question: “Why do children become poor?” giving emphasis to the parental poverty: Among all the groups, the participants dis- cussed parental poverty as one of the main contributory factors in child poverty. The participants meant specifically that parental insufficient income and fewer resources can be two major rea- sons among others for parental poverty. By and large, parental income and access to resources (capability inputs) are considered as a means of well-being production. Unlike citizens of some welfare states (e.g., Scandinavian countries), this is applicable for the most citizens of develop- ing countries where the only source of investing in children is parental income. In such a case, an individual’s income from: 1) wages (e.g., if employed) or other sources (e.g., if self-employed in small business), and 2) properties are very important. In a wider sense from an economical point of view, resources are concerned with a range 114 of commodities, services and other aspects. However, from the poor child participants’ perspec- tives, owning land has been regarded as a resource that could enable them to have other sources of income. According to the child participants, it appears that parental access to resources such as land could bridge the gap between spending and income with extra money. One child participant noted: My parents are poor since, if my mother’s parents had left any land for us, then we could have run that and become rich ...If we had some land then we could have earned some extra money from that. As noted by this child, I find two important issues that came into view: first, parental poverty has been passed on from the preceding generation to the next i.e. from grandparents, to parents to grandchildren; and the second, due to parental poverty, it is less likely for the next generations to have inherited properties such as land. Similarly, participants in the poor women’s group identi- fied parental poverty as one of the major explanations for child poverty. One woman remarked: “Our financial state is bad and we cannot educate our children”. In an effort to keep the expenses in line with their low income, parents prioritized spending on food rather than education, as is evident by the discussion with this woman. As expressed in this statement, it seems unlikely that children (who are raised in parental poverty) obtain education, since their parents cannot manage to pay for it. The impact of parental poverty on children’s economic status as adults is not straightforward. Yet, it is less likely that poor children with less education would be competent to take full advantage of getting well-paid jobs in their adulthood. Therefore, the likelihood of being poor is greater in the next generation if the present generation of children stays in parental poverty for a longer period. As another woman explained succinctly: My mom isn’t in a very good state in the village. She gets by somehow. If my father was educated and had a government job then my siblings would be in a better position and my sister wouldn’t need to come and stay with me. In all the groups, participants agreed that low or no education put their parents at risk of unem- 115 ployment or only gave them access to low-paid jobs. These jobs include rickshaw puller, brick breaker, maid, etc. For example, one child participant said: “Our parents have not studied…there- fore have no jobs and we are poor”. Thus, the participants in all the groups also affirmed the positive relationship between education and earnings. One of the members in the children’s group said: “Their parents didn’t have any education so they couldn’t earn any money”. Moreover, the lack of women’s participation in the labour market or having only a single earner in the family was discussed as contributing factors in the children’s group. One said: “My mother doesn’t work outside. She stays at home. My father works sometimes. If he gets money then he brings food and my mother cooks”. Another child added: “We live on whatever my father earns”. As noted earlier, the participants placed great emphasis on the factors such as limited resources, parental employment status and, low education and low income as being the main con- tributors to parental poverty. The principal concern that is frequently expressed in the discussions is: children who are born and grow up in parental poverty are more likely to be less educated. In such a case, this restricts them to have access to potential well-paid jobs. Thus, their employabil- ity only to casual jobs in later life limits their possibilities to earn enough. Consequently, the risks of a variety of negative effects on current generation of children for living in parental poverty are highly likely to create intergenerational cycle of poverty. highlighting on the insufficient income and resources to meet basic requirements: In this sub-category, the concept of child poverty has been grounded on the contemplations of deprivation of basic needs. This deprivation seems to have an association with having not as much of parental access to earning and properties such as money and land. Parental insufficient income and less access to resources make poor children more vulnerable now and in the future. In all the groups participants discussed “insufficient income to meet basic requirements” of the children in poor families as another crucial contributory factor in child poverty. One woman participant said: If we had a higher income then they could get better things. They (children) can only get the things that we can afford. It’s not like we have a very high income. Whatever we earn we spend and live on. 116 It was apparent in the discussion that households spend almost all of their income to satisfy their basic needs, which is not enough. One child commented: “My parents are poor… because they have no land”. Another child added: “That is why they cannot give us food and clothes”. “The poor children cannot eat because they (parents) don’t have money for that”, one more child com- mented in a different group. Another child said: “They can’t educate us due to the lack of money”. Thus, the children participants talked about their poor access to food, clothing, education and other basic needs resulting from parental inadequacy in relation to income and resources to pro- vide for the families including the children. In particular, children were also conscious of the additional spending of their parents dur- ing sudden shocks. For instance, one child participant spoke: “All that we earn now gets spent (on food and others). There are always sudden accidents, right. So my parents don’t have any money left in their hands”. This statement affirmed that whatever the poor earn get spent on purchasing primary requirements. Even more worryingly, requirements of extra expenses at the occurrences of sudden shocks (such as illness, accidents, etc.) make matters worse. The poor are less capable of dealing with such sudden shocks with extra costs for health care, social service etc. Similarly, one woman echoed her vulnerability and worries in the face of increased ex- penditure on housing (e.g., rents) and children’s schooling: “They (house owners) have increased my rent (of her residence) and fees of the children’s schools. I do not know what to do”. In such a situation, any parent similar to this woman would be puzzled concerning how to solve the problem. The poverty of income and price rises on daily necessities act simultaneously on the poor parents, the impact of which at present is highly likely to carry on its influence on children’s economic status in their adulthood. As noted earlier, the consequences could not be expected greater for the poor children. In such cases, compared to the better-off, the present generation of poor children is more likely to be tomorrow’s poor adult and poor parents. The groundwork of intergenerational transmission of poverty may, therefore, be created in the current generation and the next. Other factors, such as being a member of household headed by a female were discussed as contributing to insufficient income, especially where she is the sole earner and the family is 117 large. A participant in the women’s group was especially insistent on this point. She illustrated her case as follows: “I am the sole earner and have children, and with my salary, my household manages somehow…even the food is alright, though light. Then I have to pay the rent. Nowadays the prices are so high in the markets”. As reported by this woman, children having only one low-paid earner in the family are more likely to be poor in comparison with children having either both parents or more than one adult care taker as earners. As noted in the above statement, evidence on sole and low-paid household heads indicates that they need to spend a greater proportion of their earning on purchasing food. Additionally, they are more likely to face difficulties to pay for other basic requirements, such as house rent. On top of that, the likelihood of mounting distresses increases for the reason of increasing prices of daily necessities. A common fact that the poor participants of this study in Bangladesh share with the poor in many developing countries is that their earnings do not cover the basic necessities. There is always a shortfall between income and spending (Harper, 2004; Hulme & Shepherd, 2003). In other words, expenses are far above in regard to earnings. Unhappily, these matters deprive the children of the poor of meeting their basic needs. Calling attention to gender discrimination: In this study, the participants of different groups pointed on the fact that boys and girls were treated differently in relation to intra-house- hold allocation of distributing foodstuff and spending on education. The subjective visions on the imbalances of sex-related bias at household levels, with reference to their own experiences, were debated among the participants. Girls’ experiences of being neglected and discriminated against by their parents were evident in the discussions among the children’s groups. Children stated that they were well aware and understood gender discrimination at household level. For example, one girl said: “They (par- ents) give each of us one handful. But to the boys they give as much as they want...I say, ‘Give me some rice too’, and then she (mother) gives it to me”. The extent to which gender biases exist in 118 intra-household food distribution may be roughly guessed while reading through this statement. Boys are the parents’ first choice among the children in food distribution, as is documented here. Undeniably, the boys get food without asking while the girls get them by asking. Cultural or eco- nomic preference for boys is noted as socio-cultural norms and traditions in many societies, particularly, in developing countries. Although gender biases persist more or less all over the world, discrimination against women with low-income, as is seen in traditional societies is strik- ingly different than that in relatively affluent societies. The boys also reflected their parents’ attitude of “no return on girls’ education” saying: “Parents love sons more. Daughters do less and parents also have to spend more money (for dowry) on daughters”. Another participant added: “They (sons) get money from the bride’s side too”. This discussion referred to the dowry system, which is widely seen as the cause of negative attitudes towards investing in girls. In essence, instead of investing in girls’ education, preferences are given to invest in boys’ education. As outlined in the above account, the reasons for this might be 1) social and structural discrimination on the basis of sex, and 2) cultural norms and values of the conventional society are promoting that the boys in their later life take liabilities of their parents’ source of revenue. As a result, preferences are more likely given to boys than girls in spending money on education, and possibly even in distributing food. Girls are denied access to the required level of education which restricts their potential opportunities to obtain a well-paid job. Thus, it is more likely that today’s poor girls would remain poor as adults. Furthermore, it is documented that boys’ education is acquired, in many cases, at the ex- penses of girls’ education where there is a shortfall in parental income. For instance, one woman described her preference to educate her boy instead of her girl. The reasons are her helplessness, as a sole earner, in the face of required high spending in one hand and less income on the other. She said: “I am alone and it’s not possible for me. I have a son who’s studying. I haven’t been able to enrol my younger daughter yet. Wherever we go they ask for a lot of money”. Therefore, as described by the victims of gender discrimination in the focus groups, gen- der discrimination contributes to future child poverty by keeping girls under-privileged, under- nourished and under-educated. These findings support the results of the large literature on Ban- 119 gladesh and on other South Asian countries on discrimination against girls and women in the intra-household allocation of food, medical care and schooling (Salmon, 2005). Studies (Moore, 2001) emphasize the gender differential in parental investment of time and capital in education and training, health and nutrition, and general care. Undernourished girls face higher risks of maternal and child mortality. During adulthood their own children are more likely to be under weight, unhealthy and unnourished if they con- tinue to live in poverty. As described above, gender discrimination and the low status of women can have irreversible effects in inhibiting children’s health and educational attainment (Harper, 2004). The families of the participants are institutions of a patriarchal order, and maintain gender segregation and discrimination (similar to the perspective of the radical school of thought on the conceptualization sex/gender described in Calás & Smircich, (2006, p. 289). This is a cultural or social norm in childhood that is often experienced throughout life and between generations (Harper, 2004). The views on gender discrimination for the women’s group differ from those discussed in the children’s groups. Women participants reflected on how early marriage and a husband’s greater age ultimately place girls and women at risk of poverty in the future. One woman said: “My mother married me off when I was 12– they married me off to a quite old man. He later got old, we had children. Then he died. We are not even educated ourselves”. Similarly, many girls in the developing world become victims of early marriage and drop out from school. Participants were fully aware and understood how they were deprived of education in their childhood because they were girls. The women also discussed the priority given to boys’ education as a way of improving their own security in old age. A social norm that exists in Bangladesh among the women and chil- dren is the belief that children have to take care of their parents once they grow up. One woman participant said: I have stopped her education completely. Now the son goes to school and he also has to pay a lot of fees there. But he will help me so I am still sending him. But I still cannot educate my younger daughter… I have to pay rent and for food so I cannot afford it. 120 Participants thus reflected on gender discrimination. According to their experience, un- derstanding and perception, gender discrimination is generated in poverty, and it reinforces pov- erty and passes poverty on to the next generations. Moore (2001, p. 9), in a study on intergen- erational transmission of poverty, states a similar view (see also Engle, Castle & Menon, 1996; Kabeer, 1998) according to which “a negative intergenerational impact of gender-discriminatory investments clearly emerges—a vicious cycle of low investment in women and low investment in girls”. The voices of some of the victims from Bangladesh add another dimension to poverty analysis. In many cases, the poverty generating process resides in women’s subordinate social and economic position in society (Chant, 2006), no matter where women live in the world. 6.4. Passing poverty from one generation to the next In this section, I would like to present shortly the theory that emerged from the results presented above in the earlier sections. A summary of open codes, axial codes, categories and core category, which are drawn from focus group discussions of poor children and women groups is presented in Figure 7. In the process of analysis applying grounded theory approach, I went through a variety of dimensions. The process started out with assigning open codes by reading FGD transcripts and ended up in finding a core category, which theorizes on child poverty. This research has disclosed that child poverty, according to FGDs among poor chil- dren and women, plays a vital role in the prolongation of the process of developing, expanding, extending,transmitting and persisting poverty from one generation to the next. In many cases, it is likely to enhance, generate, and regenerate further in a progression of intergenerational trans- mission of poverty. This progression is grounded in child poverty. Three different dimensions (categories A, B, C) are likely to be found interconnected to this progression of children’s resid- ing in poverty; these are: 1) Encompassing multidimensionality, 2) Spending life in distress, and 3) Having intergenerational and gendered dimension (see Figure 6 and Figure 7).The subsequent sections contain a description of these interrelated dimensions (presented in Figure 7) that work together to generate and transfer child poverty to the next generation. 121 The first dimension (Category A) “Encompassing multi-dimensionality”draws atten- tion to children living in poverty being victims of multiple forms of deprivation (axial code 1), particularly in relation to non fulfilment of basic needs. The parental incapability (axial code 2) in relation to impaired conversion factors39 is one of the vital factors that make poorchildren’s suffering more acute. This is particularly true where parents are the only source of investment in children. In such cases, parents face difficulties to pay even little attention to children’s wellbeing. Therefore, children who grow up in poverty are, most probably, less likely to end their poverty in a generation. The aspirations expressed (axial code 3) as a means for breaking the cycle of pov- erty are far from being realized. As their achievement in education suffers in connection with pa- rental incapability, this not only requires the willingness, but also the ability to materialize them. Their care givers’ incapability restricts their achieving proficiencies and skills that are required for the potential paid employment in future. The second dimension (Category B) is about “spending life in distress” in child poverty. In this category, the accounts of the consequences of child poverty were clustered. At the first sub- category (axial code 1) in this dimension, the food insecurity such as food shortage, not having adequate and edible food in hunger was voiced. If unsatisfied hunger continues to exist longer, the risk of severe negative upshots increase in children’s physical, psychological and human develop- ments. In particular, children who are raised with unsatisfied hunger face different health related problems such as malnutrition. Clearly, it would place them at higher hazards at present and even in their later life. This means that they are less likely to be efficient, skillful and active to acquire a well paid job in future in their adulthood. Children raised their voice not only against hunger, some of their assertions centred also in relation to the poor dwellings. This has been clustered around the second sub-category (axial code 2). It is not surprising that the dwellings of the poor participants were not well protected or constructed, as demonstrated in this study. For instance, they live in badly built housing (unsafe shelter), have less access to improved water, and lack infrastructure for good sanitary practice. Furthermore, poor living conditions may also lead to a series of long-lasting disadvantageous out- 39 Parental incapability in these cases is associated with impaired personal conversion factors. These are, for example, low earning and education, less access to resources etc. 122 comes. For example, these outcomes could be: getting sick, developing chronic health problems, not attending school, dropping out etc. In any case, less achievement in education would be a cause of great distress for not getting a well-paid job or having just casual jobs during adulthood. Furthermore, articulating poor access to public services is another sub-category (axial code 3) which gives a picture of children’s, who are raised in poverty, less access to the exist- ing public services (mainly, in health and education sectors). Similar to many other developing countries, inadequacy and corruptions in service delivery were reported as two major challenges. These contribute to restrict the poor children from getting full advantages of the existing public provisions of welfare in parental poverty. Along with other sub-categories (axial codes 1, 2 and 3), the issue child abuse (axial code 4) in association with children’s belongingness to the lower economic groups has been brought into notice in Category B. This problem further increases the likelihood of blocking children’s success in getting education. The final dimension (Category C) highlights on the “intergenerational and gendered dimension” of child poverty. Some of the vital issues were grouped together around this category. As the voices of the poor participants (both children and women) were heard, these issues were more likely found to contribute to child poverty. For example, parental poverty (axial code 1) was identified as one of the main contributors to child poverty. On one hand, this is typical for an economy where public provision of welfare is extremely limited, while on the other the parental poverty to provide for children is prominent. Moreover, insufficient income and less or no access to resources also contribute to parental poverty. As a result it blocks the route to education for their children. Moreover, parents’ low level of education serves to perpetuate gender discrimination by undervaluing the benefits of investing in girls. Gender discrimination in the current generation produces the future mothers (the girls of the current generation) with little or no education which contributes to the poverty of their own children in the future, hence continuing a never ending cycle. From the analysis of three aforementioned dimensions (categories), it can be logically predicted that these interrelated factors work together to perpetuate the intergenerational trans- 125 6.5. Summary and discussion This chapter is based on four focus group interviews consisting of one women’s group and three children’s groups (children aged 6-14). Samples were drawn from three different slums of Dhaka, Bangladesh. Data has been analyzed using a grounded theory approach. The aim of this study is to present how stakeholders of child poverty conceive child poverty, its causes and its impacts. When discussing child poverty, all the focus groups emphasized the lack of food and other basic needs. The definition of child poverty extracted from their conception is a lack of basic human needs such as food, as a result of parental incapability. This suggests that child poverty is a state of children’s “elementary deprivation” (of being adequately nourished, being in good health, etc.) due to parental inability to purchase the necessities, as described in Sen’s basic capabilities and deprivation approach (Sen, 1993, pp. 40-41). This definition of child poverty is applicable to a hungry, unclothed, homeless child, regardless of where they live. The children were primarily interested in talking about the deprivation and suffering in poverty whereas the women were more interested in the reasons behind parental (their own) incapability. In addition, child poverty has been found to be characterized by a gap between aspirations and the likelihood that they will be realized, where children’s willingness to go to school was hampered by the gap between a desire for education and the realistic possibilities of achieving this goal. Public service authorities, despite the state guarantees of free access to educa- tion, did little or nothing to ensure that children could attend school. The participants’ responses indicated that child poverty is seen as stemming from the failure of parental “entitlement to basic human needs, which they cannot achieve through various legal channels” (Sen, 1995, p. 52). Sen (1981) regards this as the major reason for Bangladesh’s famine in 1974, as well as for poverty in other countries in similar situations. Thus, child poverty was conceived by our participants as multidimensional, and to have multidimensional influences on children’s well-being. It can be defined as a circumstance where the children are deprived of basic human needs due to the failure of parental entitlement and capa- bility in a country where state support is very limited, and where children live in poor conditions 126 that cause huge suffering. This definition leads us to the interpretation that, no matter where a child lives in the world, poor children (living in a similar situation to the participants in this study in Bangladesh) are deprived of their rights to basic human needs, as set out in the “Convention on the Rights of the Child” (see UNICEF, 2006, p. 10). The conception of child poverty from the participants’ perspectives highlights parental poverty, insufficient income to meet basic requirements and gender discrimination as factors re- sponsible for child poverty. The participants emphasized poor education, employment status and the presence of a sole-earner in the household as some of the main reasons for parental poverty and insufficient income to meet basic requirements. The discussions in the women’s group focused primarily on their vulnerability as parents in single-earner families, especially in larger families. Although children discussed their difficulties in living in both female-headed and male-headed households having single-earners, they perceived themselves to be more vulnerable in households with a female head. Similar results have been found in many other studies (for example, Chant, 2006; Dieden & Gustafsson, 2003) for countries such as the USA, Australia, Africa, and Eastern and Western Europe. These studies suggest that single parent families headed by women are more likely to be poor. From a gender perspective, this study provides a broader understanding of poverty, which allows a better grasp of the multidimensional aspects of gender disadvantage. Children also raised questions about intra-household resource distribution, which is a long-standing feminist concern (Chant, 2006). Early marriage, the age difference between spouses (where the husband dies be- fore the wife due to greater age), and boys’ education receiving priority, were conceived in terms of gender perspectives. Regarding early marriage and age difference in Bangladesh, Siddique (1998) argues that the gender gap regarding age at marriage means that women are more at risk of losing a partner since they get married at a much younger age than men. This is evident in our data. Unfortunately, women participants, who were themselves victims of early marriage and who reflected on its drawbacks, were still prepared to repeat the same mistake with their own daughters. On the other hand, some boys in the children’s groups expressed pride in being loved 127 and preferred by their parents. It is clear that this attitude is firmly rooted in the culture, which is very common in South Asia and many of the developing countries. Thus, girls are more vulner- able as a group to being even more “underprivileged” in underprivileged, poor households. Corruption is also seen as a factor that blocks poor children’s access to whatever univer- sal public services exist in Bangladesh (for example, free primary education and healthcare). This is witnessed by the fact that most of the focus groups revealed the non-existence of such services due to the misuse of public funds. This discussion raises a serious question regarding corruption in public hospitals and schools. The impact of corruption is clearly present in most of our group discussions, which suggests similarities with many other developing countries. This limitation of public service is less pervasive in developed countries where children’s education and healthcare are supported by relatively generous transfers (Bradbury et al., 2001; Corak, 2006). The category “Spending life in distress” focuses on the impact of child poverty on chil- dren; children’s own experience of poor living conditions is analyzed. Children’s discussions centred on their helplessness in the face of hunger, starvation, inedible food and miserable living conditions. Women participants’ perceptions supported most of the children’s views. However, they also talked about their frustration over their inability to protect their children from the dis- tresses of poverty, e.g. not being able to secure treatment for health problems. The result is that children grow up in poor health. Data revealed that not only do children grow up in poor health, they also lack education, which is in part due to child abuse that occurs at school. This, too, is seen as stemming from child poverty since poor children are often ridiculed and mistreated which can lead to children dropping out of school. Despite their fundamental willingness to go to school, the children become reluctant in the face of the violence, segregation and humiliation directed at them from peers and school teachers, simply because they are poor. A similar situation is evident in many other countries. For instance, in the North American context, Iceland (2006) describes similar patterns of behav- iour; “In the educational system, discrimination against minority groups contributed eventually to access to unequal facilities”. The overall view of the data shows that poverty is passed on in a system of intergenerational poverty. It is a never-ending process in which children are born into 128 parental poverty and grow up with poor education, especially the girls, due to gender discrimi- nation. Thus, the process leading to child poverty is repeated by each generation, with very few exceptions. 129 Chapter 7 Do economic growth and inequality always matter in child poverty reduction? Evidence from Bangladesh and China40 7.1. Introduction The present chapter––based on Begum, Deng and Gustafsson (2012)––is part of extended ef- forts to go beyond a country and time specific knowledge of understanding child poverty. Prior to this chapter, the discussion centres on some of the basic and relative questions in association with child poverty’s extent, characteristics etc. in Bangladesh. Qualitatively, I also made efforts to understand how children experience poverty in the preceding chapter. Not surprisingly, widespread child poverty was found in Bangladesh. Among the vari- ous characteristics, household heads’ education, place of residence etc. were found to have pro- found effects on child poverty. The previous chapters include efforts to study the extent, characteristics and experience of child poverty, but so far I have only done this in the context of a single country and at a specific point of time. But the concept of child poverty is multifaceted and complex, an in depth analysis of this phenomenon requires further studies of it that analyse it across countries and over time. This chapter deals with reasons for changes of child poverty over time as well as differences from a cross country perspective in terms of issues related to growth, income inequality and poverty. To 40 This chapter summarises the article Economic growth and child poverty reduction in Bangladesh and China by Syeda Shahanara Begum, Quheng Deng, Björn Gustafsson, published in Journal of Asian Economics, volume: 23 (2012), pages: 73–85. An earlier version of this article was presented at the International Confer ence on Experiences and Challenges in Measuring National Income and Wealth in Transition Economies, September 19-21, 2007, Beijing, China, jointly organised by the International Association for Research on Income and Wealth (IARIW) and the National Bureau of Statistics (NBS), China. 130 see how this works, two countries, China and Bangladesh, were selected due to their experience of economic growth as well as the availability of large sample survey data from these countries that the co-researchers and I had access to. This chapter presents a comparative analysis which is done using a decomposition frame- work. Through this, poverty differences are attributed to differences in the: i) mean child income, ii) demographic composition, and iii) distribution of child income. As discussed below, the ad- dressed research questions here are: 1) In an assessment based on a World Poverty Line set to 1 USD PPP per person and per day, to what extent does child poverty of Bangladesh relate to that of China, particularly the South West of China?. 2) Based on experiences from Bangladesh and China, is poverty reduction only a question of income growth? The study has employed International Poverty Line. It should be noted here that this dissertation uses the same poverty line for defining and identifying poor children (for details, see chapters 4 and 5 in particular). It is mentioned earlier that this poverty line is often-used and widely familiar among social scientists, but is not an uncontroversial alternative (see Chapter 2, Chapter 4 and Chapter 5). For the most part, this line has been used as the monetary measure of household economic welfare, and only children in households below it are deemed to be “poor.” By this view, “child poverty” is defined in this study as children’s living in a household that has a disposable income of less than one dollar per person per day in Purchasing Power Parities. Once the definition of “poor children”, along with the poverty line for identifying chil- dren in poverty are clearly set, in the next step, large sample surveys are used for child poverty comparison across Bangladesh and China and across time. The study period was chosen on the basis of availability and suitability of data for the purpose of child poverty comparison across the two countries. The selection of these periods includes: the mid-1990s to 2000 (from the year 1995/1996 to 2000) for Bangladesh. On the other hand, the period for China was chosen from the late 1980s––(from the year 1988 to 1995) to the beginning of 2000s (from the year 1995 to 2002). Selecting Bangladesh and China for this comparative study was primarily out of conve- nience. However, there were a number of conceptual reasons that influenced the decision to com- pare these two countries specifically (see Chapter 4 for more details). First, it is certain that these 131 countries do not represent all developing countries; however, two countries of Asia were covered. In doing so, it has proven useful in providing insights into the association among growth, inequal- ity and child poverty reduction, and other factor as well such as the demographic composition of child poverty in these countries of the developing world. China and Bangladesh are among the countries that have larger proportion young indi- viduals in their population, and we can assume that many of these children of these developing countries are going to be poor. In any event, children are not poor by choice; child poverty is the outcome of the economic activities of parents or other adult household members and the cir- cumstances around it are not subject to individual control. Child poverty is structured within the societies in which it appears and is possible to be affected by policy measures. With this in view, a broad consensus has converged in the poverty literature and among observers about the desirability of combating child poverty; the discourse or increased awareness regarding child poverty has thus become a subject of great interest among policymakers and re- searchers alike. For the most part, however, in richer countries and in the Commonwealth of Inde- pendent States than in low- and middle-income countries, a mounting concern has been strongly emphasized that shows a growing focus on child poverty. As an aspect of this, United Nations Children’s Fund (UNICEF) monitors child poverty in rich countries, South-eastern Europe and the Commonwealth of Independent States (Corak, 2006a; UNICEF, 2007a). Contributions to the academic literature on child poverty include, for example, Bradbury et al. (2001), Corak, Fertig and Tamm (2008), Vleminckx and Smeeding (2001), and Gordon et al., (2003). To date, such past efforts, particularly, for child poverty comparison in association with extent and evolution across Asia, or across low- and middle-income countries, have typically not been noticed. This is one of the most important reasons for conducting this comparative study. As discussed further below, this chapter extends the imagination on the reasons for child poverty over time and across countries–to move forward through research that has economic growth and inequality relevance. Next, the analytical and policy debate since last few decades on the question of “how eco- nomic growth affects poverty” has drawn considerable attention and increased research aware- 132 ness. The position one takes in that debate would build a further extension to the literature on growth, inequality and poverty reduction. In this case, decomposition analysis to explain growth, inequality and poverty nexus across countries may offer a deeper understanding on the issue stated. What makes it difficult in considering many countries under study are funding and time limitations. Considering these, two growing nations have been focused as two case examples–– Bangladesh during the second part of the 1990s, and the People’s Republic of China from the late 1980s to the beginning of the new millennium. As aspect of China, an amazing growth record of the last 30 years stimulated by a policy of opening-up41 is well documented. By comparison, Bangladesh’s experience of recent emerg- ing, strong and rapid economic growth, though starting from a lower level, since 1990 also has caught attention of the analytical and policy discourse. In addition to comparing these two coun- tries as entities, a comparison between Bangladesh and the south-western region of China seems to be equally worthy. Here, the considerations taken into account are: existing i) large regional differences in levels of income in China; ii) the south-western region of China, with a considerably lower than average income level, as well as it is in terms of size, location and population are more similar to Bangladesh than any other parts of China; iii) comparing Bangladesh with southwest China thus provides another investigation of how economic growth affects child poverty. This chapter is relevant to current interest in understanding how growth inequality and child poverty nexus work in these two countries. These interests can be developed further by asking what can be learnt from these countries in an Asian perspective in a rapidly changing developing world. Even though these two countries are clearly not representative of all developing countries, the study is useful in providing insight into the literature on child poverty, growth and inequality nexus of a regional level of the developing world. The assumption that economic growth is a prerequisite for poverty reduction in a longer time perspective is not without controversy. Primarily, the arguments on this issue rely on the threshold that the theorists use in defining poverty. Supporters of the view as stated above would 41 The reform and open-door policy in the late 1970s of China is called opening-up policy in which interna tional trade policy, foreign investment, greater openness to the international economy. 133 prefer to use an absolute definition of poverty. By this view, “poverty” is defined as living under a predetermined fixed poverty line (which is defined as “absolute poverty”). Some observers–– who use a relative definition of poverty––argue against this view. These dissenting views do not include the assumption that economic growth leads to less “poverty”. Supporters of the latter define poverty in relative terms in which “poverty” is defined as living under a poverty line that is updated to the general level of living. Related ideas of disagreements have been explored extensively in the literature. The roots of these views are embedded in the thought that economic growth is the outcome of many differ- ent processes. As it happens, each of these views can incorporate specific empirical observations and assessment based judgments as well. In case economic growth is concentrated to those worse off, it becomes obvious that income inequality as well as poverty decreases. While if growth is concentrated to those better off, inequality increases and poverty may or may not decrease. Rapid economic growth in the developing world in recent times has caught attention of the analysts and in policy debates. One of the major reasons behind this is being directed to the policies of “opening up of previously more closed economies to allow for increased international trade and foreign investment”. If this process of economic growth leads to rapid industrialisation, domestic capital owners as well as skilled and semi-skilled urban workers would be benefited. It should be noted here that, in Bangladesh’s case, most semi-skilled urban workers reside in the industrial areas where economic growth is concentrated. However, household members in urban areas send remittances to contribute to their household expenditure in rural areas. Such a growth does not necessarily spread instantly to rural areas where typically the bulk of the country’s poor reside. In a recent contribution, Khondker and Koolwal (2010) reach a simi- lar conclusion in the case of rural Bangladesh. In the context of Pakistan, Naschold (2011) further documents that while income in some areas rises, other regions may experience geographical poverty traps. Once they do, it becomes obvious that there can be episodes where positive eco- nomic growth does not click together with poverty reduction in the population (see also Winters et al., 2004). Episodes of no poverty reduction despite economic growth can also be the result of demographic changes (widely defined). What is more, if ever larger/smaller proportions of the 134 population belong to greater/lesser poverty-prone segments, this counteracts/reinforces impulses towards poverty reduction coming from a growing economy. Empirically measuring the positive contribution of economic growth over time on poor households can be challenging, as noted by Khandker and Koolwal (2010), unless the mecha- nism is applied exactly through which the poor are likely to be benefitted from economic growth. Given that growth and poverty reduction must not necessarily go together, it is not surprising that questions of the connection between economic growth and poverty reduction are subject to considerable research efforts. A strand of the literature favours that economic growth provides a contribution to both the scientific and policy debate about different countries and periods have been investigated using various research strategies. Multiple studies analysing how poverty and economic growth vary in single countries exist (Datt & Ravallion, 1992); and another genre of research centres on analysing many countries (for example, Fosu, 2011). For evidence on this point, Thurlow and Wobst (2006) is often cited in the literature on economic growth. By applying a general equilibrium and micro-simulation model, this recent study investigates how the sectoral structure of growth affects poverty with reference to Zambia. These authors further argue that not all growth is equally good for the poor. Among macro studies, for example, a classic reference is often made to Dollar and Kraay (2002). Kraay (2006) further applies a decomposition framework to analyse data mainly from the 1990s covering 80 developing countries in which poverty is defined using the World Bank’s “one-dollar-a-day” poverty line. An analysis of spells of poverty change leads to the conclusion that cross-country differences in growth, especially in the medium- and long run, are the dominate factors explaining changes in poverty. A combination of data sets is used in Loayza and Raddatz (2010) who use a combination of databases. These include the same database used by Kraay (2006) on poverty spells, World Bank data on sectoral value added, and Purdue University’s Global Trade Analysis Project data- base (GTAP) on labor shares (for details see Loayza and Raddatz, 2010). By combining data on poverty spells with information on sector value added and on global trade, the authors investi- gate how the sector composition of growth affects the economy’s capacity to reduce poverty and 135 conclude that growth in sectors intensively using unskilled labour has the largest potential for reducing poverty. Montalvo and Ravallion (2010) analysed how GDP from primary, secondary and ter- tiary sectors in China has affected poverty using provincial data from the first part of the 1980s to 2001. Their results support the view that the primary sector (mainly agriculture) has been the main driving force in poverty reduction. It should be noted that during the spells studied here, the agriculture sector in China grew more slowly than the industrial sector. Khandker and Koolwal (2010) examine how infrastructure and credit expansion in rural Bangladesh may have contrib- uted to household income growth and/or poverty reduction, using a new-pooled dataset of three household panel surveys administered during the period 1991–2001. They reach a similar conclu- sion that growth does not necessarily spill over into poverty reduction for the poorest households. However, to my knowledge there is no previous study such as this one that focuses on how eco- nomic growth affects the extent of poverty among children. To recapitulate briefly, child poverty is defined in this study as living in a household with a disposable income lower than the often used international poverty line of one-dollar-a day PPP, promoted by the World Bank. Applying this approach, for any given point in time, child poverty has been more extensive in Bangladesh than in China. During one of the three spells of rapid eco- nomic growth studied, child poverty did not decrease profoundly in China. Here, a more unequal distribution of child income between 1988 and 1995, too, can be viewed as offsetting poverty reduction impulses of economic growth. However, for the other two spells in the period studied, economic growth was in step with child poverty reduction; this was the evolution in Bangladesh for 1995/1996 to 2000 as well as for China from 1995 to 2002. The cross-country comparisons show that the lower child poverty rates in China can mainly be attributed to the country’s higher average child income level, while differences in income inequality and demographic composition are of lesser importance. When trying to under- stand why in the mid-1990s southwest China had lower child poverty rates than Bangladesh, dif- ferences in demographic composition are found to be fundamental. In southwest China, children lived in families with fewer children than in Bangladesh. However, a few years later, the gap in 136 child poverty between southwest China and Bangladesh had widened, with differences in mean child income playing a larger role. The study thus illustrates that economic growth and differences in income levels affect child poverty differences across time and across countries. However, it also shows that economic growth does not automatically lead to less child poverty. For understanding changes over time and across countries in child poverty, it can be necessary to also consider changes/differences in the distribution of child income as well as in the demographic composition. The rest of the chapter proceeds as follows: the next section includes information on Bangladesh and China as the context for the comparison.. Section 7.3 introduces the data sets used in this chapter. A description on child poverty in China and Bangladesh is presented in Section7.4. The accounting framework is presented in Section 7.5, and the results are shown in Section7.6. Finally, the chapter comes to an end by making concluding remarks in Section 7.7. 7.2. Context Some information on cross-country differences seems to be useful when trying to understand the extent and changes of child poverty in Bangladesh and China during the 1990s. This can lead to a better understanding of the relationship between some of the economic and social development indicators. To better illustrate, some of this information is presented in Table 14, the discussion based on which continues as follows. GNI per capita is higher in China and life expectancy at birth longer. Growth has been more rapid in China, thus further widening the income gap between the countries. However, the distribution of income/consumption is less uneven in Bangladesh. In addition, the age structure of the populations differs evidently. Children construct a larger proportion of the total population in Bangladesh. China has reached a further stage in the process of demographic transition with a lower birth rate and under-age-five mortality rate. These developments are consistent with the adoption of China’s one child policy and Bangladesh’s policy of encouraging no more than two children. 137 Furthermore, although a process of rapid urbanisation is taking place in both countries, the majority of people in Bangladesh and China still live in rural surroundings. Yet, the countries differ in one important aspect regarding the dimension rural-urban. The reflection of China’s ur- ban–rural income gap is demonstrated in the literature. For example, in a recent paper, Sutherland and Yao (2011) describe China’s urban–rural income gap as massive in international standards, and even bigger than in other Asian countries. One possible explanation concerns the role played by the household registration––“Hukou”––system introduced in China in the 1950s. The “Hukou” system is considered to be one key obstacle to migration, by preventing rural migrants from accessing all the benefits associated with legal residence in cities (Sutherland & Yao, 2011; Mullan, Grosjean & Kontoleon, 2011). In contrast, a counterpart of such a system does not exist in Bangladesh. The Hukou system categorises people and only those classified as urban residents had the right (at that time) to live in a city. The development policy adopted, soon after PR (the People’s Republic of) China was formed, was meant to prioritise the better-off urban population. Rural households had to deliver quotas of agricultural products thereby funding the industrialisation that took place in urban China (see, for example, Knight & Song, 1999; Whyte, 2010). Sicular et al. (2007) attribute about half of the contemporary rural to urban income gap to differences in endowments, most importantly education. 138 As noted above, the Hukou system has functioned as a barrier to movements of the less-privileged rural inhabitants into the cities. This ends up in very little mass poverty within its registered urban population. However a new kind of poverty has appeared in the cities of China. One explanation concerns the role played by the in-migration of rural residents who do not have an urban Hukou. Another possible explanation arises from the policy of economic restructuring that appeared in China during the latter part of the 1990s and its consequences. This has led to increased unem- ployment and decreased labour market participation. When assessing poverty in urban China, poverty lines higher than one-dollar-a-day PPP are usually used (see Riskin & Gao, 2010; Apple- ton et al., 2010). As it happens, the urban to rural income gap is larger in China than in many Table 14. A comparison of Bangladesh and China. Bangladesh China Southwest Chinaa Population, millions 1990 116.00 1 135.00 180.00 2009 162.00 1 331.00 197.00 Average population growth 1990-2009, percent 1.80 0.80 0.50 Proportion of people aged 0–14 years, 2009, percent 31.00 20 b .00 20.00 Surface , thousand sq km 144.00 9 600.00 2339.00 Population density, people per sq. km 2009 1 246.00 143.00 84.00 PPPb GNI c per capita USD 2009 1 560.00 5 490.00 3370c.00 GDP average annual growth 2000–2009, percent d 5.9 d 10.9d 11.00 Literacy rate adults (%) of population 15 and above, 2009 c 56.00 94.00 90.00 Urbanisation of total population 1990 2009 20.00 28.00 27.00 44.00 13.00 38.00 Gini coefficient for income /consumption 2005, pe r cent 31.00 41.50 n.a Under-five mortality rate per 1000 1990 2009 148.00 52.00 46.00 19.00 64.00 n.a Life expectancy at birth 2006, years c 65e 73e.00 68f.00 Crude birth rate. per 1000, 2009 21.00 12.00 11.00 Notes: Sources for Bangladesh and China are with the exceptions denoted World Bank (2011a). a Southwest China is here defined to include the following province level units: Sichuan, Guizhou, Yunnan and Tibet. It also includes Chongqing which until 1996 was a part of Sichuan. Sources are China Statistical Yearbook and Provincial Statistical Yearbooks, various issues; China One Per Cent Population Survey 2005. b Includes Taiwan. Source: World Bank (2011a), with the following exceptions: c http://data.worldbank.org dWorld Bank (2011b) 139 other countries. Unlike the Hukou system, the residents’ movements throughout the country are unrestricted in Bangladesh. This being said, one must acknowledge that the situation is dissimilar in Bangladesh’s cities where big cities in particular have experienced large population influxes due to expectations of getting more prompt employment opportunities. Despite Bangladesh belonging to the group of the ten most populous countries in the world (a list headed by China), China has a population nine times as large. Bangladesh’s territory is much smaller than China’s as well. Of China’s 31 province level units, Liaoning (144 900 sq km) of northeast China has an area similar in size to Bangladesh (World Bank, 2007a). Within China, regions differ considerably regarding economic development. The eastern region is most developed, while the western region, particularly its southern part, is lagging behind. With an average income level and a population size more closely resembling that of Bangladesh, and in addition being situated closer to Bangladesh, it seems motivated to compare child poverty and its development in Bangladesh with its counterpart in southwest China. 7.3. Data For studying child poverty in Bangladesh, microdata taken from household surveys conducted by the Bangladesh Bureau of Statistics for the years 1995/96 and 2000 was used. A two-stage strati- fied random sampling was utilized for both surveys. In the first stage, 252 rural and 190 urban Primary Sampling Units (PSUs) were selected from the sample frame with probability propor- tional to size. The PSUs were chosen from 14 different strata (5 rural and 9 urban). At the second stage households were selected from each PSU. The Household Income and Expenditure Survey (HIES) 2000 made use of almost the same sample design as the Household Expenditure Survey (HES) 1995 with slight modifications (see also Bangladesh Bureau of Statistics, 2003). The modi- fications were made in Statistical Metropolitan Areas and PSUs; however, sample sizes were kept the same for both surveys. Also, the definitions for rural and urban were the same in both surveys. Due to the changes in questionnaires some variables needed to be defined as closely as possible, for example, household income, parents of the children, etc. Both surveys cover rural as well as 140 urban areas. From the microdata, three regions were defined, which differ in their distance to the coast as well as economic development (Rahman & Hossain, 2009). For this, the administrative divisions were used as building blocks: the most developed coastal or southern region consists of Chittagong, Barisal and Khulna, the mid-central region of Dhaka and Sylhet, and the less devel- oped inland or northern region of Rajshahi. The Chinese data come from the three waves of the China Household Income Project (CHIP) survey conducted in 1989, 1996 and 2003 for the reference years 1988, 1995 and 2002. This means that the study presented in this chapter has access to surveys on both rural and urban areas. The 1988 rural survey covers 28 provinces (see Eichen & Zhang, 1993), while nine were not retained in the 1995 survey. These same provinces and also Guangxi and Xinjiang appear in the 2002 survey. The 1988 urban survey was drawn from 10 provinces to represent the Coastal, Central and Western provinces and the same territories were also surveyed in 1995 and 2002 with the addition of Sichuan for the latter two years. Sichuan was split into two administrative regions at the provincial level, Sichuan and Chongqing, in 1997, both of which are surveyed in 2002 rural data (for further details on the 1995 and 2002 surveys, see Li et al., 2008). The age at which a person is no longer regarded as a child depends on social circum- stances that might vary across countries; in most cases, countries have their own techniques for a definition of child within the range of particular age. In the context of Bangladesh, for example, the age for defining a child is very complicated (for details, see Chapter 4). In this study, a child is defined as being age 14 and under. In both countries, children start school at age 6 or 7. Primary school is typically for 5 or 6 years in China and 5 years in Bangladesh, while junior high school is for 3 years in addition to junior years of primary schooling. 141 The number of children and adults (as defined above) in the five samples and their sub-samples used in the study are reported in Table 15. As can be seen in this Table, a total of around 16 000 children in the two samples for Bangladesh and in the Chinese sample for 1995. The larger size of the Chinese sample of 1988 is partly due to sampling and partly due to birth rates being higher previously. The fall in birth rate in China is also the reason for the Chinese sample of 2002 not having more than 9 000 children. The two later sub-samples for southwest China comprise about 2 000 children. Table 15 . The number of children and adults in the samples for Bangladesh 1995 and 2000 and for China 1988, 1995 and 2002. All Urban Rural Numbe r Percentage Numbe r Percentage Numbe r Percentage Ban g ladesh 1995 Children 16569 42.43 4910 38.97 11659 44.07 Adults 22482 57.57 7688 61.03 14794 55.93 Entire 39051 12598 26453 2000 Children 15071 39.13 4388 35.71 10683 40.73 Adults 23447 60.87 7899 64.29 15548 59.27 Entire 38518 12287 26231 China 1988 Children 30809 24.98 - - 13653 26.60 Adults 92517 75.02 - - 37682 73.40 Entire 123326 51335 1995 Children 16090 21.54 - - 8160 23.49 Adults 58621 78.46 - - 26579 76.51 Entire 74711 34739 2002 Children 9428 15.97 - - 6403 17.80 Adults 49623 84.03 - - 29565 82.20 Entire 59051 35968 Southwest China 1988 Children 3577 27.07 - - 2784 28.37 Adults 9635 72.93 - - 7029 71.63 Entire 13212 9813 1995 Children 2132 20.19 - - 1356 22.37 Adults 8426 79.81 - - 4706 77.63 Entire 10558 6062 2002 Children 2082 17.44 - - 1498 19.82 Adults 9859 82.56 - - 6060 80.18 Entire 11941 7558 Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002. 142 Children make up 42 per cent of all persons in Bangladesh in the sample for 1995/96 and 39 per cent of the sample for 1999/2000 (see Table 15). For China the corresponding proportions are much lower, starting at 25 per cent in 1988 and falling to 16 per cent in 2002. In both coun- tries, at any point in time, children make up a somewhat larger share in rural areas than in urban areas. Children constitute a slightly higher population share in southwest China than in China as a whole. Table 16 offers descriptive statistics of five samples and their sub-samples using selected vari- ables. The Table presents both rural and urban sub-samples for Bangladesh, and only rural sub- sample for China; the reason for this is, as reported in the next section, poverty as defined in this Table 16 . Samples of children by background characteristics (in percentage). Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002. Age Ethnic No. of children Parental education Region <=6 Minority 1 2 3 4 & more One High Other Both Low Coastal Middle Inland Bangladesh 1995 Entire 43.22 9.51 22.49 27.14 40.87 38.35 39.54 22.10 42.31 35.59 22.10 Urban 40.47 11.04 25.74 27.80 35.42 38.92 44.05 17.03 45.01 37.96 17.03 Rural 44.38 8.86 21.12 26.86 43.16 38.12 37.64 24.24 41.17 34.59 24.24 2000 Entire 42.55 11.15 26.29 27.95 34.61 39.77 38.50 21.73 43.44 34.83 21.73 Urban 40.47 13.58 30.95 27.55 27.92 40.36 43.21 16.43 46.72 36.85 16.43 Rural 44.38 10.16 24.38 28.11 37.36 39.53 36.56 23.91 42.09 34.00 23.91 China 1988 Entire 35.79 8.80 31.82 40.31 19.18 8.70 26.49 41.35 32.16 36.29 39.98 23.73 Urban 34.12 4.06 68.90 28.49 2.45 0.15 59.41 36.06 4.53 36.47 44.83 18.70 Rural 36.39 9.25 22.10 43.21 23.69 11.00 17.94 42.59 39.47 36.34 38.43 25.23 1995 Entire 30.33 7.83 42.72 38.28 15.47 3.52 36.67 45.82 17.51 32.62 40.98 26.40 Urban 31.08 4.88 87.16 12.34 0.49 - 78.53 20.64 0.83 33.35 38.84 27.81 Rural 30.14 8.88 29.72 45.66 20.11 4.51 24.50 52.84 22.66 32.39 41.37 26.30 2002 Entire 26.93 10.06 61.76 30.71 6.62 0.91 43.21 45.44 11.34 32.84 40.06 27.10 Urban 28.35 7.00 92.77 7.29 - - 85.01 14.48 0.51 30.44 41.62 27.94 Rural 26.74 11.48 46.99 41.92 9.75 1.34 24.05 59.62 16.32 34.25 39.06 26.69 Southwest China 1988 Entire 35.43 24.27 30.00 35.90 19.71 14.40 17.87 41.24 40.89 Urban 30.26 13.98 65.70 31.53 2.27 0.50 50.00 42.06 7.94 Rural 36.91 27.23 19.83 37.14 24.68 18.35 8.64 41.00 50.36 1995 Entire 31.55 27.65 56.43 28.42 12.10 3.05 37.10 38.30 24.59 Urban 71.13 12.50 91.75 8.25 - - 79.32 20.03 0.65 Rural 33.12 36.54 36.21 39.97 19.03 4.79 12.71 48.87 38.43 2002 Entire 29.99 29.25 54.56 31.12 11.53 2.79 33.08 46.49 20.43 Urban 30.14 14.04 96.23 3.77 - - 83.89 15.56 0.56 Rural 29.92 35.18 38.32 41.79 16.02 3.87 14.38 57.87 27.74 143 study is virtually absent within the urban sub-sample for China. Based on earlier reports on gender bias among children in China it comes as no surprise that the ratio between boys and girls is higher in China than in Bangladesh. In fact, it increases across the surveys. As an example of the attention in literature that the sex ratio in China has drawn, Johansson and Nygren (2001) verified the trends in sex ratios at birth in China. Findings from this study provide additional insights into the earlier findings about the explanations such as underreporting of female births, gender-specific abortion, and female infanticide to account for the missing girls. Turning next to the age of children, most children in China are seven to fourteen years of age, while such dominance of school children is less pronounced in Bangladesh. As illustrated in Table 16, around ten per cent of children in the countrywide samples for China belong to ethnic minorities and in southwest China about one out of three. In contrast, the child population of Bangladesh is rather homogeneous when it comes to ethnicity. As only a very small number of children are classified as an ethnic minority group in the surveys, the fraction is not reported here. Traditionally, and historically households used to have many children in Bangladesh. Even though the situation has changed since 1980s, most children in Bangladesh live in a household including at least three children, while this is unusual in China. This is illustrated in Table 16. In Bangladesh, there is not much difference in parental education between rural and urban areas, while the opposite is the case in China where rural parents have considerably lower educations. The Table does not indicate any dramatic differences in education level between parents in rural Bangladesh and rural China. As school systems are not the same in the two countries it is not easy to compare education levels. 7.4. Describing child poverty Social science scholars have put forward diverse standpoints and definitions of what comprises poverty. These definitions differ from narrow to broad views, based on the populations and the methodologies used. In the case of cross-country comparisons, how to define and measure pov- 144 erty is not crystal clear. This makes it very difficult to make comparisons across populations. For some time, the World Bank’s one-dollar-a-day in Purchasing Power Parity poverty line has been in extensive use when assessing poverty on a global scale, for example, on a regular basis in the dominant World Development Reports, published by the World Bank (see Chapter 2). The justification for this level is that it stands for a poverty line typically in use in poor countries (see Ravallion et al. 1991; Deaton, 2006). For the study presented here the above mentioned poverty line is employed. In turn, this line is converted into local currencies using available conversion factors to derive a poverty line for child poverty estimation in the countries under study. By using this method, the following is found while assessing child poverty and its relevant issues: i) for Bangladesh, the poverty line is set to Tk. 404.71 in 1995/96 and Tk. 500.16 in 1999/2000 per person per month;42 ii) for China, the poverty line is set to 425 Yuan per year in 1988, 934 Yuan per year in 1995 and 1010 Yuan per year in 2002 in rural China.43 The poverty line is set to 450 Yuan, 1025 Yuan, and 1134 Yuan in corresponding years for urban China. A child is defined as poor if living in a household with a disposable per capita income lower than the poverty line specified. One would acknowledge that this is not the only possible alternative for performing cross-country comparisons of child poverty. As noted earlier, the advantage of this alternative is that for some time it has been the one most widely used. Dissenting views on using this particular poverty line include, for example, Reddy and Pogge (2010) and Himanshu (2008) who have put forward a detailed critique of the approach. In a reply, Ravallion (2010) powerfully argues against this view (see Chapter 2 for more details). There is also the issue of appropriate conversion factors being subject for discussion. For example, new (though not necessarily ideal) conversion factors for China have been derived, leading to revisions of time series of poverty in China (Chen & Ravallion, 2010). This means if 42 We compute 1 USD poverty line for the year 2000 as follows: (U.S. Consumer Price Index in 2000/ U.S. Consumer Price Index in 1985) *[Official Exchange Rate (Tk / USD in 2000)] x Purchasing Power Parity Conversion factor to the official exchange rate. That is 1 USD PPP = (168.8/105.5) * 52.1 * 0.2 = 16.672 Taka per day, which is equal to 500.16 Taka per month. The pov erty line for 1995 was obtained by deflating with Consumer Price Index for Bangladesh. 43 Personal communication on the poverty line in 2005 with Shaohua Chen at the World Bank. 145 consensus on a new international poverty line including appropriate conversion factors emerges, it would be motivated to revise the results presented here to see how robust they are. The national poverty line is different from, and not comparable with the poverty line, that is used here for Ban- gladesh. The reason for this is that the national poverty line is consumption based and higher for both years (see BBS, 2003). For China, the poverty line used here is higher than the poverty line used when the National Bureau of Statistics of China (NBS) reported the extent of poverty in rural China for the years studied here. However, more recently NBS has adopted a low-income line that is closer to the one-dollar-a-day poverty line. In 2008, after the period here studied, the National Bureau of Statistics introduced a new poverty line for urban China set to 1 196 Yuan per person and year, a level higher than the low-income standard in use, and much higher than the poverty line that had been previously in use. Table 17. Child poverty rates in Bangladesh 1995 and 2000, China and Southwest China 1988, 1995 and 2002 (in percentage). Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002. Note: The child poverty rates for urban China are less than 0.5%, therefore, breakdowns for urban China are not reported in the Table. Age Nationality No. of Children Region <=6 7-14 Han Minority 1 2 3 4 One High Other Both Low Coastal Middle In land Bangladesh 1995 Entire 41.57 35.15 24.06 32.37 37.63 44.41 3.16 23.66 48.58 34.87 34.39 49.56 Urban 17.21 14.27 5.17 10.92 16.48 21.16 1.43 7.57 26.53 15.18 10.77 28.23 Rural 50.92 44.57 33.98 43.38 46.84 52.44 9.57 34.45 54.93 43.34 46.02 55.87 2000 Entire 29.64 26.85 16.12 22.51 28.92 35.37 3.28 19.19 33.93 21.65 26.44 42.56 Urban 14.06 13.57 4.19 10.16 13.65 22.53 0.00 8.67 19.73 10.39 13.40 23.02 Rural 35.45 32.70 22.67 28.96 35.06 39.31 6.56 25.59 38.24 26.38 32.77 48.08 China 1988 Entire 23.47 17.94 18.35 35.24 9.55 19.47 30.33 36.95 12.00 18.90 27.11 12.13 20.68 31.27 Rural 29.37 22.73 23.44 39.69 17.27 22.81 31.26 36.95 22.65 22.91 27.91 15.20 27.18 37.70 1995 Entire 21.37 17.14 16.40 39.61 8.16 22.18 33.71 34.92 10.72 19.80 31.77 7.14 17.17 34.31 Rural 27.49 22.14 21.38 46.39 15.18 23.83 33.82 34.51 20.56 21.57 32.43 9.14 21.56 45.15 2002 Entire 9.14 5.66 5.75 14.14 3.18 10.22 19.23 24.42 2.51 7.66 18.61 3.29 5.24 12.60 Rural 13.38 8.29 8.54 18.23 6.02 11.03 19.23 24.42 6.44 8.47 18.88 4.65 7.84 18.72 Southwest China 1988 Entire 27.31 19.70 18.39 34.07 9.13 21.18 32.77 38.83 Rural 33.69 25.84 24.60 39.11 17.57 26.31 33.62 39.14 1995 Entire 30.46 23.44 20.03 37.59 9.56 38.94 56.98 75.38 Rural 45.66 37.38 36.00 44.93 22.81 43.54 56.98 56.98 2002 Entire 14.90 10.56 9.91 16.58 4.31 20.06 21.25 29.31 Rural 20.54 14.67 14.93 19.17 8.36 20.77 21.25 29.31 146 Child poverty rates for Bangladesh and China and the sub-samples are reported in Table 17. As the table indicates, the samples (when relevant) are fragmented by age, ethnicity, number of chil- dren, parental education and region. Table A in the Appendix provides corresponding information for the composition of child poverty. First looking at the total samples in Table 18, child poverty rates in Bangladesh are found to drop from 38 per cent in 1995/96 to 28 per cent in 2000. In China the child poverty rate was considerably lower, 20 per cent in 1988, but had remained almost un- changed in 1995. However, in 2002 the child poverty rate in China was down to 7 per cent. When observing the two countries as whole units, economic growth and poverty reduction are thus no- ticed to move in tandem during two of the spells studied, but much less so during the third (China from 1988 to 1995). Not surprisingly, child poverty rates are higher in Bangladesh than in China. Virtually all poor children in China live in rural areas and the child poverty rate in urban areas is more or less zero. For urban Bangladesh on the other hand, a child poverty rate of 15 per cent in 1995/96 is reported, which was only marginally reduced in 1999/2000. In contrast, child poverty rates in rural Bangladesh fell from 47 per cent in 1995/96 to 34 per cent in 1999/2000. For China, changes in child poverty rates in the country as a whole are entirely driven by child poverty reductions in rural areas.44 At the beginning of the new millennium, child poverty rates in rural southwest China were of about the same magnitude as in urban Bangladesh, thus lower than in rural Bangladesh. Child poverty rates are generally a few percentage units higher among the youngest chil- dren than among school-aged children. In the samples for China as a whole and for the southwest region, child poverty rates are higher for ethnic minorities than for the majority. A general pattern is found of child poverty rates being highest in families with many children. There is a strong negative relation between parental education and child poverty rates in the data for Bangladesh, whereas a slightly weaker counterpart is found in the later surveys for China. In both countries child poverty rates are highest in the inland regions, lowest in the coastal. 44 An increase is reported in the child poverty rate in rural southwest China from 29 percent in 1988 to 40 percent in 1995, but note that it is based on a relatively small sample and other poverty indices (reported below) do not show an increased extent of poverty. 147 7.5. A framework for making poverty comparisons This section includes a discussion about the framework that is employed in the study for making poverty comparisons based on which the current chapter is made, A decomposition framework for poverty comparisons are attributed to Danziger and Gottschalk (1995). In this accounting frame- work, poverty differences between situations A and B are due to three components: differences in average income, differences in demographic composition and differences in the distribution of income. This framework can be used to study changes in poverty over time in a single country, but also to study differences in poverty at one point in time across two countries. The framework was first used to study the evolution of poverty in the United States (see Danziger & Gottschalk, 1995; Iceland, 2003), and (for unknown reasons) seems not to have been widely applied to low- and middle-income countries previously. The general idea of the framework is to quantify how child poverty would have differed as a result of three separate forces: income differences, demographical differences and distribu- tional differences. As further explained below, the first of these shows how the extent of child poverty would have differed due to income growth only. The second shows how child poverty would differ in the case of differed demographic composition only. This component is typically not present in studies of how economic growth in developing countries affects poverty, and can be deemed to be an advantage of the approach. The third component shows how the extent of child poverty would have differed in the case of unchanged average child income, but where the distribution of child income changed between the two situations compared. The decomposition builds on several steps, beginning with computing the growth com- ponent. Starting from an actual base distribution (A), we assign each child an income based on the assumption that the difference in average child income observed between distributions A and B is equally shared within the population. From this distribution we compute the extent of child poverty. This simulation maintains the demographic composition and income inequality of distribution A, but has the mean child income of distribution B. The first simulation allows us to estimate what every demographic group’s child poverty would have been in the second situation if only mean child income differed. 148 The next step weights these group-specific child poverty indices by demographic com- position of the child population as it was observed in distribution B. This second simulation incorporates the inequality of the base situation, but has the mean child income and demographic composition of the second situation. The difference between child poverty from the two simula- tions equals the difference in child poverty due to demographic differences. When applying this framework we must choose along which variables demographic composition should be defined. In this application we choose to use parental education along with the number of children in the household as these characteristics are shown to affect child poverty rates. The difference in child poverty that is accounted for by differences in inequality in child income is computed as a residual from the outcome of the second set of simulations and the real situation. By construction, the sum of these three components (the difference attributable to dif- ferences in mean child income, to demographic differences and to differences in child income inequality), will equal the observed differences in child poverty. It should be noted, only FGT0 is reported here, while results computed for other poverty indices belonging to the family of indices proposed by Foster et al. (1984) are reported in Begum, Deng and Gustafsson (2012). Discussed next are the results from the decomposition framework. Child poverty rates are reported in Table 18 for the samples as well as the various sub- samples. 149 In Begum, Deng and Gustafsson (2012) it is shown that other poverty indices belonging to the FGT family45 give different views on how the extent of poverty developed in urban Bangladesh from 1995/1996 to 1999/2000, and that not all indices indicate an increase in poverty in southwest China as a whole between 1988 and 1995. 45 FGT family includes FGT0, FGT1 and FGT2. FGT0, FGT1 and FGT2 assess child poverty by number af fected (“incidence”), how poor the poor are on average (“intensity”) and how the poverty deficit (up to the poverty line) is distributed among the poor children (“inequality”) respectively. Table 18. Child poverty rates (in percentage) for Bangladesh (in 1995 and 2000), China and Southwest China (in 1988, 1995 and 2002). Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002. Child poverty rates percent Rural Bangladesh 1995/1996 0.4739 1999/2000 0.3390 Urban Bangladesh 1995/1996 0.1546 1999/2000 0.1376 Entire Bangladesh 1995/1996 0.3793 1999/2000 0.2804 Rural China 1988 0.2514 1995 0.2375 2002 0.0965 Entire China 1988 0.1992 1995 0.1842 2002 0.0660 Southwest China: (rural) 1988 0.2874 1995 0.4012 2002 0.1642 Southwest China: (whole) 1988 0.2239 1995 0.2566 2002 0.1186 150 7.6 Results The question arises whether economic growth does have an impact on poverty reduction. To answer this, at first, results of the decomposition illuminating change over time in each country are reported in Table 19. Starting with Bangladesh, it is found that economic growth was a rather strong force for poverty reduction in the rural region, but rather weak in the urban region. This is not surprising as Bangladesh experienced rapid agricultural growth during the period studied, as discussed in several studies. For example ADB (2001) reports a more than five per cent annual growth in agricultural production value for the years 1997 to 2000 (as opposed to a growth of two per cent per year during the years 1991 to 1996). Many factors are deemed to have contributed to this development: expansion of the rural non-farm sector, expansion of micro credit programs, and reform of agricultural and other sectors (see also World Bank, 2005a). Indications of change in income inequality are found to be small in rural areas, and their signs differ by poverty indices, as in urban Bangladesh. Demographic change was found to be of very small importance for pov- erty development in Bangladesh. Table 19. The decomposition of percentage-point change in child poverty rate, Bangladesh 1995 to 2000 and Entire China 1988 to 1995, 1995 to 2002 and 1988 to 2002. Ba ngl a d e sh (19 96- 20 00 ) Rural Urban Tota l c hild pov erty rat e s c hil d poverty ra te s c hild poverty rat e s Actual chang e in pover ty ra te 0.1348 -0.0 1 6 9 -0. 09 8 9 (1) Ec onomic chang es 0.1365 -0.0269 - 0. 1036 (a) Gro wth i n mean adjuste d in c o me 0.1281 -0.0014 -0. 07 1 1 (b) Cha ng e i n incom e inequ al i t y 0.00 84 -0.02 55 -0. 03 25 (2) Demographi c changes 0 .00 1 7 0. 01 0 0 0 . 00 4 7 (a) Pa ren tal edu c ation 0 .00 7 3 0.01 7 1 0.01 1 1 (b) No. Of chil dre n 0.00 60 - 0. 00 95 - 0. 00 76 (c) Inte ra ction 0 .00 0 4 0.00 2 3 0.00 1 1 Chi na a s a whol e 19 88- 19 95 1995- 20 02 1988- 20 02 c hild pov erty rat e s chil d poverty ra te s c hild poverty rat e s Actual chang e in pover ty ra te -0.01 49 -0.1 1 8 2 -0. 13 3 2 (1) Ec ono m i c chang e s 0.005 9 - 0. 10 30 - 0. 12 17 (a) Gro wth i n mean adjuste d in c o m e -0.08 93 -0.1317 -0. 16 5 0 (b) Cha n g e i n inc o me inequ a l i ty 0.09 5 2 0. 02 8 7 0 . 04 3 3 (2) Demographi c changes -0 .02 09 -0.01 5 2 -0. 01 1 6 (a) Pa ren tal edu c ation -0.01 25 - 0.0039 -0. 0041 (b) No. of c hi l d r en -0.01 32 -0.01 38 -0. 00 79 (c ) Inte r a c ti o n 0.00 4 8 0. 00 2 5 0 . 00 0 4 Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002. 151 Turning to China, and starting with the period 1988 to 1995, increased parental educa- tion and fewer children per household are found to work towards decreased child poverty, but the movement towards poverty reduction from economic growth was considerably more forceful.46 However, a worsening of the income distribution was working in the other direction, particularly when assessing poverty by the head count ratio, though less so when applying the distribution sensitive poverty index. Strong growth in average household income in China from 1988 to 1995 did not do much to improve the situation of the poor children in China, a finding that can be traced back to the sectors in which growth appeared. Economic growth was fastest in the eastern part of the country and less impressive in the western part where the highest poverty rates were found. During the period 1995 to 2002, decreased numbers of children per household continued to work towards less child poverty in China. However, the magnitude of this effect is marginal compared to the poverty-reducing impulses that came from economic growth. Such impulses were somewhat larger than during the preceding period. Increased income inequality did not forcefully counteract the development towards less child poverty as it did in the first period. This means that when summing up the development over the two periods (that is when analysing the period 1988 to 2002), we find that the development of child poverty is to a large extent driven by economic growth, slightly reinforced by demographic change, but also somewhat counteracted by a worsening income distribution. 46 Results for rural China reported in Table A in the Appendix are not surprisingly similar. 152 The question that now arises is why child poverty is less extensive in China than in Bangladesh. Table 20 showing the decomposition results for the mid-1990s and the late-1990s provides some insight and the answer differs somewhat between the two comparisons. However, in both cases the main reason for child poverty being less extensive in China is a higher mean income. Demo- graphic differences contribute to this, particularly in the comparison made for the mid-1990s. Differences across countries in the income distribution work towards reducing the cross-country poverty disparity. This is particularly the case when comparing poverty of the mid-1990s. Table 20 . The decomposition of percentage-point difference in the poverty rate: cross-country comparison of Bangladesh with China (rural regions and entire countries). Rural regions of Bangladesh and China Entire Bangladesh and China Child poverty rates Child poverty rates 1995/96 Actual difference in poverty rate -0.2364 -0.1950 (1) Economic differences -0.1383 -0.1126 (a) Difference in mean adjusted income -0.2775 -0.2460 (b) Difference in income inequality 0.1392 0.1334 (2) Demographic differences -0.0981 -0.0824 (a) Parental education -0.0704 -0.0646 (b) No. of children -0.0445 -0.0412 (c) Interaction 0.0167 0.0234 2000/2002 Actual difference in poverty rate -0.2425 -0.2144 (1) Economic differences -0.2030 -0.1926 (a) Difference in mean adjusted income -0.2507 -0.2439 (b) Difference in income inequality 0.0477 0.0513 (2) Demographic differences -0.0396 -0.0218 (a) Parental education -0.0295 -0.0167 (b) No. of children -0.0178 -0.0104 (c) Interaction 0.0078 0.0053 Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002 153 When comparing poverty in Bangladesh with poverty in southwest China in the mid-1990s, see Table 21, the importance of demographic differences stands out. At that time, child poverty was clearly less extensive in southwest China than in Bangladesh. However, the decomposition indi- cates that if southwest China would have had the same demographic composition as Bangladesh (primarily having the same average number of children), there would have only been a slight dif- ference in child poverty rates.47 But when making comparisons at the beginning of the new mil- lennium the situation is different. The higher mean income in southwest China is the predominant explanation for poverty being less extensive in southwest China. The decomposition analyses have shown that all three components considered are of importance for the poverty comparisons. Rapid economic growth has large potential for lowering the extent of poverty, but it is not sufficient for reducing child poverty. This is a conclusion well in agreement with what can be found in the literature on growth and poverty in the developing world. For example Ravallion (2001, p. 1812) concludes: “The poor typically do share in the 47 However, poverty measured by other indices would have been smaller in southwest China, see Begum et al. (2012). Table 21. The decomposition of percentage-point difference in the child poverty rate: cross- country comparison between Bangladesh and Southwest China. Sources: Authors’ calculations from HES 1995, HIES 2000, CHIP 1988, 1995 and 2002 Rural Bangladesh and Southwest rural China Entire Bangladesh and Southwest China Child poverty rates Child poverty rates 1995/1996 Actual difference in poverty rate -0.0727 -0.1227(1) Economic differences 0.0641 -0.0165 (a) Difference in mean adjusted income -0.0407 -0.2019 (b) Difference in income inequality 0.1048 0.1854 (2) Demographic differences -0.1369 -0.1062 (a) Parental education -0.0802 -0.0808 (b) No. of children -0.0655 -0.0613 (c) Interaction 0.0089 0.0359 2000/2002 Actual difference in poverty rate -0.1748 -0.1618 (1) Economic differences -0.0961 -0.1221 (a) Difference in mean adjusted income -0.1208 -0.2070 (b) Difference in income inequality 0.0247 0.0849 (2) Demographic differences -0.0787 -0.0397 (a) Parental education -0.0546 -0.0299 (b) No. of children -0.0409 -0.0206 (c) Interaction 0.0167 0.0108 154 benefits of rising aggregate affluence, and they typically do suffer from economic contractions. But there is a sizable variance around the ‘typical’ outcomes for the poor.” Similarly Bigsten and Shimeles (2007) in an exercise of asking if Africa can reduce poverty by half by 2015, find that while strong focus on growth is the only viable option for some countries, changes in income distribution can have a large effect on poverty in others. The findings also indicate that for poverty comparisons, differences in demographic com- positions can be of importance, a possibility typically not considered in the same literature. When studying poverty in the United States from 1949 to 1999 for different groups using the same framework as that used here, Iceland (2003) found that income growth explains most of the trends in poverty. Rising inequality in the 1970s and 1980s was especially important in explain- ing increases in poverty among Hispanics, whereas changes in demographic (family) structure played a significant role for children and African Americans through 1990. However, changes in demographic (family) structure no longer had a significant association with trends in poverty for any group in the 1990s. 7.7. Conclusions The conclusion drawn in Begum, Deng and Gustafsson (2012), on which this chapter is based, is presented below. The discussion proceeds in this chapter to describe child poverty in Bangladesh and China as well as to investigate reasons for differences across the two countries using har- monised microdata. The reasons for changes over time during periods of rapid economic growth in both countries were studied. The study is based on large samples and a poverty line set to one USD per day. True, this is not the only alternative of defining poverty in cross-country studies. We are aware that in the future the foremost approach could be one in which the international poverty line is set to 1.25 USD per day and that the issue of how to convert this into local currencies will be solved somewhat differently than the method used in this chapter. The comparisons of child poverty were made using a decomposition framework according to which poverty differences are attributed to differences in mean child income, demographic differences and differences in the distribution of income. 155 Child poverty is very much a problem for rural children in both countries. Not surpris- ingly we have reported that child poverty is more extensive in Bangladesh than in China. Out of the three spells of rapid economic growth studied, child poverty was found to have decreased profoundly during two spells, while much less during a third (China from 1988 to 1995). A more unequal distribution of income in China between 1988 and 1995 largely offset the poverty reduc- ing impulses coming from economic growth. However, in Bangladesh from 1995 to 2000 and in China from 1995 to 2002, economic growth was much in tandem with child poverty reduction. A pattern of child poverty rates being highest in families with many children was found in both countries. Child poverty is negatively related to parental education level in Bangladesh and in China in the mid-1990s and thereafter, but much less so than in 1988. Ethnic minority children are more poverty prone than the majority in China, while the Bangladeshi population is more ethnically homogeneous as are the groups of poor children. The cross-country comparisons show that the lower child poverty rates in China can mainly be attributed to a higher average child income level than in Bangladesh. When trying to understand why in the mid-1990s southwest China had lower child poverty rates than Bangla- desh, it was found that differences in demographic composition are central. In southwest China children lived in families with fewer other children than in Bangladesh. However, a few years later, the difference in the extent of poverty between southwest China and Bangladesh had wid- ened and was now driven by differences in mean income. The study thus illustrates that economic growth and differences in income levels are significant for child poverty differences over time and across countries. However, it also shows that economic growth does not by necessity lead to a lessening of child poverty. Similarly, dif- ferences in mean income are not the only factors that affect poverty differences across countries. In addition to economic growth, changed distribution of income as well as changed demographic composition can affect how poverty develops. 156 Appendix Table A. The Composition of Poverty in Bangladesh and China (percent). Note: The composition of urban child poverty has not been reported in the Table since the numbers of urban children living in poverty are found to be negligible in China (14, 22 and 3 in1988, 1995 and 2002, respectively). Age Nationality No. of children Parental education region <=6 7-14 Han Minority 1 2 3 4 One High Other Both Low Coastal Middle Inland Bangladesh 1995 Entire 47.38 52.62 6.03 19.19 26.93 47.85 0.24 17.45 82.31 35.26 35.85 28.88 Urban 45.06 54.94 3.69 18.18 29.64 48 .48 0.70 18.12 81.18 38.21 30.70 31.09 Rural 47.69 52.31 6.35 19.33 26.55 47 .76 0.18 17.35 82.47 34.86 36.56 28.58 2000 Entire 44.98 55.02 6.41 21.11 28.82 43.66 0.19 23.36 76.45 30.71 36.30 32.99 Urban 40.57 59.43 4.14 22.85 27.32 45 .70 0.00 27.85 72.15 30.46 42.05 27.48 Rural 45.71 54.29 6.79 20.82 29.07 43 .32 0.22 22.61 77.17 30.76 35.34 33.90 China 1988 Entire 42.16 57.84 85.36 14.64 15.25 39.41 29.20 16.13 16.13 39.64 44.23 21.91 41.15 36.15 Rural 42.50 57.50 85.28 14.72 15.18 39.21 29.45 16.17 16.36 39.28 44.36 21.68 40.98 37.33 1995 Entire 35.18 64.82 82.97 17.03 18.93 46.09 28.31 6.68 21.17 48.87 29.96 12.65 38.19 49.16 Rural 34.88 65.12 82.55 17.45 18.99 45.82 28.64 6.55 21.18 47.92 30.90 12.44 37.56 50.00 2002 Entire 37.32 62.68 78.46 21.54 29.74 47.59 19.29 3.38 16.23 52.13 31.64 16.40 31.83 51.77 Rural 37.05 62.95 78.32 21.68 29.29 47.90 19.42 3.40 16.01 52.15 31.85 16.50 31.72 51.78 SW China 1988 Entire 43.19 56.81 62.74 37.26 12.33 33.96 28.84 24.97 9.93 37.25 52.81 Rural 43.25 56.75 62.69 37.31 12.13 34.00 28.88 25.00 9.95 37.17 52.88 1995 Entire 37.46 62.54 58.24 41.76 21.02 43.14 26.87 8.90 11.55 48.48 39.96 Rural 37.68 62.32 58.19 41.81 20.59 43.38 27.02 9.01 11.43 48.38 40.19 2002 Entire 37.66 62.34 59.11 40.89 19.84 52.63 20.65 6.88 8.61 54.51 36.89 Rural 37.38 62.62 58.94 41.06 19.51 52.85 20.73 6.91 8.64 54.32 37.04 157 Chapter 8 Challenges and approaches to child poverty in Bangladesh: A qualitative study 8.1. Introduction What needs to be done about child poverty is indeed a political issue. Of great concern is also the successful and effective implementation of policy and programs. Good governance is crucial to alleviate poverty and thereby sustain development and growth in developing countries such as Bangladesh (Sen & Hulme, 2006; Stern, 2002). This is especially important when the issue of child poverty is a central concern. The widespread poverty and vulnerability among children and their households in developing countries provide a powerful motivation to identify appropriate policy measures (Barrientos & Delong, 2006, 537-38; Deolalikar, 2005; Harper, 2004). Despite the recent improvements in the poverty situation during the 1990s and onwards, Bangladesh public administration, which is bureaucratic in nature, has been found to be inad- equate to deal with the massive poverty and to advance a fast track in poverty reduction (Hulme & Shepherd, 2003; Kabeer, 2009; Zafarullah, 2006). Consequently, a significant number of children live in extreme poverty and enormous misery. This raises the question of what the barriers are which inhibit poverty reduction and which lead to a significant number of children remaining in a vicious cycle of poverty. I believe that studies using qualitative methods, in order to listen to the many voices of the principal victims and to include their perceptions in the policy making process, can play an important role in answering the question posed above. In this context, an increasing demand for the use of qualitative methods is well supported (Attree, 2006; Fortier, 2006). Establishing child poverty in Bangladesh as a main focus, I will attempt to analyze gaps between the government 158 initiatives that have been taken and the initiatives that are needed, according to the perceptions of principal victims (poor children and women). Also, I intend to convey the voices of the child victims of poverty, since research on child poverty received less attention compared to research on adult poverty, especially in the context of Bangladesh. A premise of this chapter is that their recommendations need to be addressed in order to create a world where children’s development will not be restricted because of, at least, absolute poverty. In the following sections some of the poverty alleviation programs in Bangladesh will be presented. I will then attempt to review institutional weaknesses in policy implementation in Bangladesh. I conclude this chapter with an overall summary and discussion of the results. 8.2. Poverty alleviation programs Although literature on the impact of poverty alleviation programs on the poorest of the poor in Bangladesh are mixed, studies (e.g., Osmani, 2006; Sen & Hulme, 2006) are in agreement that Bangladesh witnessed accelerated growth and faster reduction of poverty in 1990s compared to 1980s, and that this trend has continued since. Despite this improvement, high income inequal- ity in both urban and rural areas remains a major concern (Murgai & Zaidi, 2005; Quisumbing, 2007), and around seventeen one fifth of Bangladesh’s population––a higher proportion of whom are children––continue to live in extreme poverty (BBS, 2011). To tackle the widespread poverty Bangladesh predominantly follows the World Bank’s anti-poverty strategy. This includes (according to Bangladesh Government’s documents): a pro- poor economic growth policy to increase income and employment opportunities, human develop- ment to enhance the capabilities of the poor by providing them with more education and better health services, gender-sensitive development to reduce gender discrimination, provision and ex- pansion of safety nets, expansion of participatory governance to empower the poor and reform of the services in the public sector (see Islam, 2004). The implementation of the policy is reflected, to some extent, in the poverty alleviation programs. These include means tested cash transfers focusing on girls’ education, and aid for the 159 poor elderly and for people of working age, food-in-kind transfers, typically aimed at the poor and often combined with micro credit programs (Islam, 2004; World Bank, 2007). These pro- grams include the Food for Work (FFW) program, the Vulnerable Group Development (VGD) program, later named the Income Generation for Vulnerable Group Development (IGVGD) Pro- gram, and the Food for Education (FFE) program (Ravallion, 2006). In addition, the Vulnerable Group Feeding (VGF), Gratuitous Relief (GR) and Test Relief (TR) programs address transitory food insecurity problems of vulnerable groups. These programs are administered through a wide variety of responsible ministries, NGOs and microfinance institutions. For example, Grameen Bank, BRAC (Bangladesh Rural Advance- ment Committee), and the UN organizations are all involved. Studies (Amin et al., 2003; Pitt, Khandker, Chowdhury, & Millimet, 2003; Weiss, 2005) demonstrate that programs targeting poverty reduction run by the micro finance institutions have positive impacts in improving the poverty situation to some extent. However, there are also many example of studies, including those mentioned above, that argue that programs are more successful at reaching the poor but less successful at reaching the vulnerable (Amin et al., 2003). A greater proportion of this excluded population is comprised of children. Studies (Grindle, 1980; Turner & Hulme, 1997) show that “around the world, commonly, there are large differences between what is planned and what hap- pens in the field. These ‘implementation gaps’ are often very great in developing countries” (as cited in Matin & Hulme, 2003, p. 651). Bangladesh is not an exception in this regard; there are differences between official policies and their implementations in practice (Landell Mills, 2002; Rahman & Razzaque, 2000). The multidimensional policy ambitions of the government of Bangladesh stand in glar- ing contrast to the slow progress of poverty reduction. One of the important reasons for this gap is that the above policies ignored or overlooked the context of institutional weakness in policy implementation that is very common in many developing countries such as Bangladesh (Islam, 2004; World Bank, 2007). One such weakness, among other things, is corruption, which is deeply rooted in the country’s political and administrative system (Dowlah, 2006). Corruption, in the words of Mutebi (2008, p. 147), is a “catch-all term that embraces a wide array of illicit and most- 160 ly particularistic behaviours including bribery, extortion, fraud, nepotism, graft, speed money, pilferage, theft, embezzlement, falsification of records, kickbacks, influence-peddling and so on”. Most studies stress that corruption has permanent detrimental effects, and according to Myrdal (as cited in Mutebi, 2008, p. 148), is a “force that slows down development”. 8.3. Institutional weaknesses in policy implementation The traditional manner of administering society, particularly the bureaucratic aspect, is ill suited to cope with the situational demands of multidimensional and complicated problems, such as poverty (Olsen, 2004; Scott, 2003). Decades ago Merton (1957/2004) had already identified and drawn attention to the dysfunctional aspects of bureaucracy, including trained incapacity, dis- placement of goals, red tape and inefficiency. The exact meaning of red tape is often unclear.48 Citizens’ blame on government of having too much red tape often points to excessive rules and regulations, waste of scarce resources, and inefficiency (Bozeman, 2000). Generally, the citizens of Bangladesh accuse the government in a similar vein and their common view is that public ad- ministration is trapped in red tape. At this point, bureaucratic inefficiency is regarded as a limitation of public policy imple- mentation in tackling the massive household poverty (Stern, 2002; Temple & Sattar, 2000) in which children live. This tends to lead to bad governance, which, in addition to limited resources, leads to bad policies, and thus interrupts in creating an environment of long term policy imple- mentation (Hulme & Shepherd, 2003). For bad governance, Temple and Sattar (2000) blame suc- cessive Bangladeshi governments that have been less accountable and responsible to citizens and more subject to illegitimate and undemocratic influences. In this context, performance of local government and the bureaucratic system is criticized as poor at reducing the poverty of citizens (Hulme & Shepherd, 2003). Such criticism has also been voiced by United Nations (UN) organi- zations and other human service organizations (World Bank, 2007) who are working with poverty in this region. 48 Generally, the concept of red tape is used to indicate complex and timewasting paper work and process before getting something done. This expression refers to the conventional official practice–– a former of ficial British custom in particular––of tying up documents with red ribbon. 161 Street-level bureaucrats (e.g., teachers, police and doctors) in Bangladesh, as described in Lipsky (1980/2004) usually have significant leeway in their practices to translate policy directives in their own way, which might jeopardize the accomplishment of central policy goals. In this con- text, Parnini (2006, p. 208) argues that institutionalized and voluntary groupings of autonomous individuals could respond more effectively to ensure good governance and democratic consolida- tion by working in different societal sectors. Zafarullah (2007,p. 169) argues that the bureaucratic elite protects itself from horizontal integration and refuses entry to non-state actors such as inter- est groups and NGOs who want to contribute to policy-making. Given the secluded and elitist role of the administrative cadres in Bangladeshi social politics and the gap between policy goals and outcomes, these positions might better be filled by street-level policy entrepreneurs rather than street-level bureaucrats in order to get things ac- complished (Petchey, Williams, & Carter, 2008, pp. 59-76). Since national policy priorities do not guarantee local action, the role of such entrepreneurs would be to ensure that, by registering unmet needs and developing multi-agency solutions. The fact that poverty reduction is a national policy priority does not lead to proper ac- tion being taken locally. A well-known phenomenon is that slippage can occur when subordinate members of organizations responsible for implementing public policies fail to comply with na- tional policy directives (Van Meter & Van Horn, 1975). Local policy entrepreneurs, who address and attempt to solve specific issues across different sectors of society, may counterbalance tradi- tional administrative cadres. 8.4. Research approach In this chapter, I attempt to explore child poverty from the principal victims’ own experience and perceptions about what initiatives or steps need to be undertaken to eliminate it, using data which emerged from interviews with five focus groups, Group 1 to Group 5. Table 7 in Chapter 4 pres- ents information briefly such as on age, sex, the numbers of members of each of these five groups. Besides, Chapter 6 contains a brief description on Group 1 to Group 4. The present Chapter is 162 based on the data from these four groups and Group 5 as well. The participants of these groups were: i) women in two groups (Group 4 and Group 5). One of these two groups was consisted of poor (and the other group had non-poor women. ii) The rest of the three groups included all poor children. The detail description on the selection criteria and other methodological approaches including selection criteria, age, sex and numbers of members of each group has been included in Chapter 4 (for data description and analysis in detail, see Chapter 4). As noted in Chapter 4, focus group interviews have the advantage of spontaneous cross checking for reliability among the group members during the discussion. Although the discus- sions among the non-poor women (due to their education) in focus groups were very well argued and focused compared with the poor groups, in some cases they tended to blame the poor (vic- tims) for their misfortune. These resulted in a few contrasting ideas (negligible) between the poor groups on the one hand and non-poor group on the other. I carefully took into account these kinds of limitations during open coding. The participants discussed a number of issues relevant to child poverty such as: “What is child poverty?” “Why do children live in poverty?” “What is the impact of poverty?” “What could be done for the children?” The last issue will be addressed as a detailed discussion in this chapter. My intention was to raise the voices of the few poor children and women to bring into the light different issues related to the alleviation of child poverty and their policy recommendations based on their own experience and perceptions. For data analysis, I use a grounded theory approach developed by Glaser and Strauss (1967), originally described in “The discovery of grounded theory” (see above, Chapter 4). This approach allowed to utilize a constant comparison method in which I compared data with data, data with category and category with category to analyze the interview texts through open coding and categorizing respectively (Charmaz, 2006; Holton, 2007). As I started coding, different labels were created as open codes for different concepts of FGD transcripts from which axial codes and categories were developed, Finally, the categories that emerged were integrated to core categories (see Figure 8). I present the results in the follow- ing section. 165 child poverty. The data from these groups provided some policy recommendations to increase the capability of the children’s parents or caregivers to raise their earnings as a step toward child pov- erty reduction. The emerged “categories” i) “getting access to education and health care services” (Category A) ii) “empowering” (Category B), iii) “requiring social justice” of the “core category” “enhancing capability and establishing social justice”, as stated above, are discussed in the fol- lowing sections (see also Figure 8). i) Getting access to education and health care services (Category A): As noted earlier, this, category emerged from “stressing adult education”, axial code A, and “describing health problems” axial code B. These axial codes were developed from the open codes: “unschooled parents”, “parents’ educational level”, “quality education for all”, “girls’ education” ,“adequate educational institutions” ,“Ill health chronic illness”, “incurable illness”, “health shocks”,” indi- viduals with health problems”. As I have discussed in details on the participants views on health problems, the axial code “describing health problem” is not focused in this chapter. The question arose whether the parents’ ability to access work and increase their income is possible given their current education, a precondition for improving employment status. One of our child respondents raised the issue clearly, stating that “…to get jobs you have to have educa- tion”. Another example from the group of poor women is: If he (husband) was more educated, had a Masters degree then he could have government jobs and we wouldn’t have to hear so many things from so many people. He wouldn’t even have to work until midnight… That would be possible if he was better educated. It is due to the lack of education that we have to be in this situation. The axial code, “stressing adult education”, was one recommendation in this context that arose in all discussions, with children, poor and non-poor groups. The discussions focused on some of the important issues that were associated with and required to take initiatives to reduce child poverty. These are: children of unschooled parents are more vulnerable in poverty; parents’ educational level is highly associated with child poverty. “One comment from a children’s group was: “If my parents study and get good jobs then they will get a lot of money and then they will become rich. That’s only if they study. If they don’t study then you will stay just as you are”. 166 Similarly, poor women and non-poor women discussed the need for adult education in combating parental poverty. More importantly, investment in children’s education was given great prominence. Chil- dren were aware of their need for education to increase their entitlement to many opportunities, which they lack when living in poverty. They understood that the simplest way to rise up from poverty was to get an education. One child showed his consciousness of education as a gateway to jobs, which would ensure entitlement to many commodities: “We have to study now. We will finish our studies and then get jobs. Then we will get money and with that money we will buy lots of things”. Similarly, another child stated that to be affluent education is a precondition: “If we study then in the future we will be rich”. Universal and compulsory primary education was a point of some discussions among the participants. Although participants understood the importance of education, they claimed that children do not have access to the so called universal and compulsory primary education. Poor children mentioned that due to a scarcity of places they were unable to gain admittance to schools and were also limited in pursuing their education by the high costs which were beyond the means of their parents. This implies that “universal” and “compulsory” primary education in Bangladesh may be a tricky and ineffective program to some extent, which itself denies some children ac- cess to education, as experienced by our participants who noted a lack of educational institutions. In this context, initiatives from the local government were suggested by the non-poor group. A participant recommended: “I think the Chairmen and Members should be assigned specific areas where they are in charge of educating the children”. Another added: “According to regions and areas”. They meant that for an effective implementation of universal and compulsory education, it should be monitored and steps should be taken by the heads (chairmen) and the members of the local government to ensure “quality education for all children”. However, the successful imple- mentation of this kind of work is again subject to corruption (e.g. bribery) at the grass root level. Neither our child participants nor our poor women group discussed who might take responsibility for ensuring their access to education. However, the group of non-poor women participants did, 167 in their discussion, look to government initiatives to establish “educational institutions”. One non-poor woman said: “The government has to take the initiative. They have to create more edu- cational institutions”. Women’s low education was an additional important point of discussions. Investment in girls’ education was encouraged in this regard. In the children’s group, some discussions made reference to the lack of, or low education among women as a contributory factor to child pov- erty. They also acknowledged that women’s education has a positive relationship with access to better and higher wage jobs and a negative relationship with child poverty: the higher the level of education among women, the better chances of high paid jobs and the lower the probability of child poverty. For example, one child said: If even my mom had studied when she was young then we wouldn’t have been so poor today. Then my mom wouldn’t have to work in other people’s houses and we wouldn’t have to eat other people’s leftovers. We would’ve been rich. The non-poor group also placed great emphasis on creating opportunities in increasing women’s educational level and blamed the traditional attitude of the society towards girls’ education. One participant of this group said: In our country, they think that girls don’t need education. I think if a mother is educated the whole family gets educated. Because a father cannot spend so much time with the family so it depends on the mother. Which is why I think women’s education is very important. Another woman agreed and added: “The poor women really lack education. They don’t know what to teach the kids. So they don’t teach them anything. So the kids are in a bad situation”. Another participant blamed the mother’s lack of education as a contributory factor in child labour saying: “Because of this lack the children become interested in earning instead and get involved with risky jobs(child labour)”. Hence, our participants’ views from all the groups show that the impact of women’s education and resource on children’s wellbeing is very important, particularly, in reducing child poverty. 168 ii) Empowerment (Category B) is derived from the axial code C where highlighting on improving knowledge and skills is portrayed. This has derived from the open codes of “giving ad- vice”, “own effort”, “guiding”, “training”, “professional help”, “job availability”, “ land access”, “increased earnings”, “higher wages”. In addition to education, other recommendations made by our participants indicated the need to empower the caregivers through additional services such as professional help or counselling services, training programs for skill development, encouraging people to make their own efforts for small business and efforts by the better off for job creation and initiatives for awareness to enable people to use their own resources. All our focus groups, including the children, argued that professional help can also play an important role in the reduction of child poverty by empowering their capability to improve their situation. One poor woman said: “It doesn’t always help by only giving money. They can help us by showing us a way in which to improve our situation”. Similarly, poor children and non-poor women were found to express positive views on professional help for the parents in poor households to help them find a job. Thus, according to all the poor women and some child participants, empowering adult households, parents in particular, would act as a safeguard of their livelihood. For instance, one poor child participant believed that assistance from someone to get a job could play a vital role in maintaining the households: “I think if someone would help (their parents) by showing the way to get a job then we could survive by getting a job”. As mentioned earlier, there are different programs run by different ministries that pro- vide professional assistance for training and rehabilitation, for different groups of people such as women, youths and persons with disabilities. Also NGOs provide education and motivation for child and mother care regarding nutrition and healthcare. Nevertheless, they are inadequate and cover only a fraction of the poor children. The poor women’s discussion suggested that em- powering parents through training programs and providing jobs can play an important role in child poverty reduction. One poor woman said in relation to this: “If they train us and employ us, then we would get rid of the problem of unemployment and have a good income”. Another poor woman expressed her belief in government initiatives and said: “They can train us for something and employ us with that. They can then give us a monthly salary during training”. 169 All the poor women also argued for counselling services for awareness among the poor, targeting empowering them for taking proper initiatives to increase income and to encourage self-dependence. For example one poor woman said: “They can advise us by telling us, Do this and increase this (income) in this way. Become aware yourselves and help yourselves”. The lack of knowledge or education, or ignorance about nutrition, along with their inability to spend on health, remains a problem in many poor countries. A non-poor participant pointed to this and in- dicated that to empower the poor parents there was a need to make them aware of this so that they would use their own resources: “They (poor parents) don’t even know what nutritional food, a wholesome meal is. (This is) due to the lack of education. They need to know how to use whatever income (resource) they have”. Although our poor women’s group showed their willingness to be employed by acquiring skills through training, one woman in this group expressed her positive attitude towards women’s labour market participation, preferring jobs that are only suitable for women. Working in a (gar- ment) factory, she does not consider suitable for women; rather her preference was for the profes- sions such as trainers (teachers). She said: Many girls in our country are unemployed. They end up working in garment factories. There are many types of jobs that women can do that they can train us for. After the train- ing, they can train other people or work themselves. They can alleviate poverty to a great extent in this way. The above-mentioned woman’s view reflects the idea that training is a source of women’s em- powerment as a continuous social process where, after being trained to do a certain trade, women would be likely to be employed. This would safeguard them and their families by becoming responsible managers of their livelihoods. In this, their children might be the most important ben- eficiaries since women are likely to invest a large proportion of their incomes in their children and families (Moore, 2001, p. 9). The latter attitude to work within special trades would restrict the scope for women to be employed in many other trades. This sort of traditional attitude towards women’s labour market participation, among women as well as men, acts as a braking force in women’s empowerment process in developing countries and developed countries. Unfortunately, 170 this is deeply rooted in a male dominated society of which our participants are a part; therefore, awareness among the population is needed to empower the poor women, particularly, for child poverty reduction. The “job availability” and “income” were described as scarce in the focus group discus- sions. It is “enhancing job availability” is an axial code derived from the open codes of job avail- ability/land access, higher wages, and increased earnings. The groups considered that, by making job available within reach of the caregivers of children, household income could be increased. They considered that sufficient household income is a precondition for the reduction of child pov- erty. One child participant said: “If my father could have proper work, he would earn money, buy groceries, we could eat. If my father can earn we can eat, otherwise not”. Another child added: “If they can get a job, then they can have a good income and give us an education. That will help us in the future”. Poor women in their group discussion also expressed similar views about job availability, land access and increased income or higher wages. For example, one poor woman said eagerly: “If I could do something for the rest of the day, like have another job or could study something, I could improve our situation by earning a little at a time”. Another woman said: “Just an increased income would solve our problems”. Another added: “I think if someone would help us then we could survive by getting a job”. The non-poor group looked to public initiatives to create more employment opportunities for the poor parents. For example, one woman said: “The government can provide more employ- ment”. They also suggested that the development of industry, communication and transport could play vital role for enlarging employment opportunities. Particularly, both the poor and non-poor women’s group proposed that the rich people should take the initiative to strengthen the role of the private and voluntary sectors in industrial development. This would complement government policy for expanding employment opportunities, particularly in rural areas. One non-poor woman said: We are a poor country but we still have many rich people. If these people would establish their industries and factories in the countryside instead of the city then there would auto- 171 matically be employment in the villages. In addition, the discussants in the same group also realized that development of infrastructure is just as important as industrial development. As one woman put it: “Then we also have to develop the communication and transportation systems”. iii) Requiring social justice Category C: Axial code E (Uttering for earning support) was derived from the open codes of loans/saving, governmental loans. Financial support and saving money were alternative ideas to combat poverty, which came from the groups of children. Financial support such as that for the development of small businesses by the parents was a sug- gestion by a poor child in the following quotation: “Well if someone would help us financially then we (the households/parents) could have a business of our own and make our situation better”. Another option that most children recommended was that saving money would improve their situ- ation, although they also mentioned the fact that whatever income their parents earned was spent. In this context, one child pointed out: “…All that we earn now gets spent…So my parents don’t have any money left in their hands.” Participants in the poor women’s group focused on the fact that financial support and credit from the government could contribute to reducing poverty by increasing income from small scale ac- tivities such as a poultry farm. One participant in this group commented: ...If I had some money then I could have a farm there and raise some poultry and could build a house there…Whenever I go there I get very upset. If I had some money I could raise some goats and increase my income. Another woman in the same group believed that the public credit program could be offered to sup- port people’s own efforts as a way out of the present situation for the households of poor children: “The government can give us loans to help us to increase our income, realizing that we need to do something and that we cannot get by in this way”. Although some participants explained that state involvement in credit programs has a critical role to play in reducing poverty by increasing income, they reported the lack of such programs. On the one hand, they admitted that NGOs do offer credit programs while on the other they criticized the high rates of interest. Our participants 172 recommended lowering the interest rate so that the poor might benefit from these programs. One poor woman mentioned this in the discussion saying: There are no such programs from the government but the NGOs do give such loans… they give us loans when we build cooperative societies. They charge a high interest rate… but they must take back less interest from us… they take more from us than they give us in loans. Non-poor women’s views also reflected their expectations for providing the poor with financial support, targeting the children in order to bring them out of poverty. An example from their dis- cussion was: What I wanted to say is child poverty… is caused by the poverty of the parents. There- fore, the first aim has to be to take away the poverty of the parents. When they (the par- ents) are established, the poverty of the children will automatically be reduced. What I think is that because they are parents, education is a difficult thing for them. Therefore financial support is more important. social justice: In all the groups, participants discussed two main obstacles in getting access to children’s schooling at primary level. These are the limited places and hidden costs (by which they meant bribery or corruption since education at public schools is supposed to be free of cost). One child provided an example of how limited places in school restricted poor children’s access: “In the school they admitted people on (date). Then they (parents) took me there but they (the school) had already admitted too may people (students), so they (the school authority) didn’t take (admit) me”. Participants in most of the groups talked about some hidden costs and improper distri- bution of available equipment and services in public education and health care centres, which they claimed was corruption on the part of the street level bureaucrats whom they face in their everyday lives. For example, according to the poor women’s group, the corruption of teachers in schools prevents poor people from gaining access to the available services provided by the state. Regarding the availability of cash transfers for the girls in secondary school, one woman 173 expressed her mistrust of the street level bureaucrats and alleged that there were corruption and some hidden costs, which they had to pay in such a case, saying: “They (teachers) say, ‘There is no money’. The teachers keep the money themselves in the name of admission fees, this and that”. Another example of hidden costs provided in almost all focus group discussions related to access to public health care centres. The members in the women’s groups and most of the chil- dren’s groups stated that free public hospitals, for example Dhaka Medical, are not really free; they often have to pay hidden costs. They are expected to buy medicine, to pay doctors, admission fees and so on. To be admitted and stay there costs money. Children also discussed health expen- diture in terms of purchasing medicine, doctors’ fees and many more costs: “It costs money to buy medicine”. Another child added: “Also to pay the doctors”. “If you go inside (to be admitted) Dhaka Medical it costs money”. This also might be a matter of mistrust of the poor towards public administration. If these hidden costs are legitimate, the lack of authentic information about legal fees for the public services concerned is one of the main reasons to form this kind of mistrust. Thus, a serious question was raised regarding the corruption in public schools and hos- pitals, which has been perceived from the discussions of the participants as mistrust towards the street level bureaucrats (for example, personnel in the administration). As expressed in most of our focus group discussions, it is evident in many developing countries that “users may be re- quired to pay in bribes more than the official price or discouraged access to basic service when seeking a public service” (Kaufmann, Montoriol-Garriga, & Recanatini, 2008, p. 1). The impact of this situation was strongly visible in most of our group discussions; poor households, includ- ing children, dropped out of school or never went to school, and remain untreated, carried their disease or died due to the lack of health care in Bangladesh. These prevent proper human develop- ment, which leads most likely to intergenerational persistence of poverty from parents to children and to their children. In this regard, a straightforward solution suggested by the poor women’s group was justice. For example, according to one participant: “Things should be distributed prop- erly by the organizations that are in charge”. Moreover, poor women participants expressed their mistrust, and were very critical of the governments that have been in power for most of the time since Bangladesh’s independence. 174 Their mistrust related to many of the poverty alleviation programs and they claimed that these programs existed primarily as propaganda and did not reflect reality. At one stage of the discus- sion, a poor woman exclaimed: “What will the government do for us?” in response to the question “What measures can the government take to improve the situation?” She added: We see on TV that they do many things. But those have never reached us. We didn’t get anything when the political party (‘x’) was in power and now that it’s the another political party, (y’) we don’t get anything either. We have never gotten any help. Another woman added: We have heard and seen on TV that they give things here and there but we have never gotten it ourselves. If we had gotten anything like that it would probably have helped a lot and we could get even better. Based on the experience of participants expressed in focus group interviews, I believe that ac- countability and transparency of public action regarding distribution and delivery of services are urgently needed so that corruption and hidden costs can be distinguished by the lenses of legiti- macy. Thus, access to public services can be ensured and the mistrust of the recipients, including poor children, can be eliminated so that social justice can be established. Together with other initiatives, these are the minimum that is necessary to ensure access to healthcare and education for the children, and this is essential to prevent future generations from being born, growing up and living their lives in poverty. 8.6. Summary and discussion This chapter addresses the various poverty alleviation programs in Bangladesh. I also described the characteristics and shortcomings of the present public administration for alleviating child poverty, and I analyzed what, according to groups of , poor children, poor women and non-poor women should be done to eliminate child poverty. The impact of such poverty alleviation pro- grams can be described in a few words in the majority of cases failure, or in the least cases, limited success. Yet, Bangladesh has experienced a slow but steady rate of poverty reduction during the 175 1990s, although the situation in the country demanded much more. This increases our concerns, since there is the risk that we face a future with an extremely large number of children living in poverty, and the present system allows a process of intergenerational transmission of poverty to continue. As I detailed in the previous sections (about the core category “To increase the capabil- ity for earning more”), the empirical data put forward the recommendations regarding improving capabilities for increasing incomes of parents. These suggestions were made to get rid of child poverty. By utilizing their own human capital, individuals and households living under conditions of extreme poverty are more likely to be enabled to meet their basic needs and to solve their own problems. In this context an educated person is more likely to be capable of tackling all aspects of their lives including her/his children’s healthcare and education. As Harper (2004, p. 1) stated: An educated person can make better use of health and other facilities, enhance their own children’s education, make demands on local services and use their knowledge and skills to enhance all other aspects of their lives. In particular, our participants’ understanding was very clear that illiteracy reduces poten- tial of earnings. They suggested ensuring adult education, universal education for children, and women’s education and empowerment in order to secure access to more work and earnings, and vice versa. The participants’ views reflect need to increase their capability and potential to realize rights and responsibilities. One interesting point is that the majority of recommendations put forward by our dis- cussion groups were initiatives directed towards increasing job related parental income. This indicates that in Bangladesh’s cultural context, the participants were either unaware of the gov- ernment responsibility to care for and support its children as citizens, or they were aware of the government’s inefficiency and the inadequacies of the bureaucratic administrative system. How- ever, interestingly, they discussed the dysfunction of the bureaucracy, the corruption, as one of the main obstacles for poor children to access the existing services in education and healthcare. Reviewing the literature and the claims made by the participants regarding their denial of or lim- ited access to public service delivery, it could be suspected that corruption (see Dowlah, 2006), 176 formalism and red tape prevent poor children and their households from acquiring “the set of dif- ferent commodity bundles through various legal channels” (Sen, 1995, p. 52). Child and parental entitlement requires more attention and poor people need to become more aware of their human rights (Parnini, 2006, p. 209). I found a great deal of consensus on most of the recommendations made by our focus group participants concerning the existing policy and programs adopted by the Bangladesh gov- ernment as recommended by the World Bank. Therefore, the question arises as to why poverty reduction has proceeded so slowly. Bangladesh’s public sector is a traditionally administered bureaucratic system, which is centrally organized and rule-bound. Dysfunctional features of a bureaucracy, such as trained incapacity, displacement of goals, red tape and inefficiency create problems in meeting situational demands. The bureaucratic public administration also fails to ad- dress the pervasive poverty and poverty-related problems, which remain, despite the fact that a large number of policy and programs for poverty alleviation exist. Turning to the empirical findings, our participants also suggested increasing the income of children’s care givers to combat child poverty. They proposed primarily financial support, public credits and reducing the costs associated with education and healthcare in particular. They also proposed government initiatives for increasing the availability of jobs, and increased income through land access, industrialization etc. In addition, our participants not only recommended financial support, but also suggested empowerment measures (such as professional help, train- ing and supporting their own efforts) in order to increase earnings and build capacity among the parents. A combination of programs that enhance both capability-building and ensured income of the caregivers of the children might be one possible way of breaking the vicious cycle of child poverty in the developing countries. In the context of Bangladesh, as I discussed earlier, many of these programs already exist according to our literature review. However, our participants raised the question of how well these programs actually function in practice. The charges against poverty alleviation programs such as public healthcare and education that they made are: hidden costs, mistrust in relation to service delivery by the street level bureaucrats and corruption. Con- 177 stitutional commitments to social justice are at stake according to our participants. This underlines the importance of what Rothstein and Uslaner (2005, p. 41-72) wrote about social trust; people who believe that in general other people in society can be trusted tend to have more faith in democratic institutions in their society. However, many countries with low levels of social trust, they argue, may be stuck in a sort of social trap. As Rothstein and Uslaner (2005, p. 70) explained: “Social trust will not increase because massive social inequality prevails, but the policies that could remedy this situation cannot be established precisely because there is a lack of trust”. From this perspective, public policy for poverty reduction needs to aim at breaking through the above-mentioned social trap by reducing the massive social inequality, and increasing social trust through accountability and transparency. What is more, the gap between the existence of theoretical policy in propaganda and its practical implementation needs to be bridged, in order to benefit the households in which the victims of poverty, the poor children, live. There is also a need for additional and alternative policies which target these children. In particular, child pov- erty reduction issues need to be highly prioritized in policy agenda as a specific aspect of poverty reduction. More importantly, to eradicate child poverty, Bangladesh needs to ensure political coor- dination, policy consistency, a coherent public service and a guarantee of political accountability. To accomplish such ambitions, changes in policy implementation are needed, primarily at the local level. The public administration needs to promote a horizontal integration with voluntary organizations and key actors (street-level policy entrepreneurs) so as to place child poverty issues on the local policy agenda. Such integration demands the development of the ability to network and cooperate across organizational and sector boundaries. Above all, service providers must be accountable to the poor children and their parents or care givers. In this, the clarity or transpar- ency of the function of public administration is vital, to reduce the mistrusts of the clients, and through this control corruption among public service providers. 178 Chapter 9 Final thoughts Children represent a country’s future, which is why their well-being is a societal concern; they cannot change their situations on their own and collective action is needed to mitigate vulnerable situations in poverty (Bradbury et al., 2001, p. 1). This is applicable to all children no matter where they live. The following steps are crucial to reduce child poverty in Bangladesh in view of the studies included in this dissertation and my value judgment. Where child poverty is found to be extensive and children constitute a greater share of population, child poverty reduction should be, in practice, first and foremost: 1) central to all poverty reduction policy agenda; and 2) given the highest priority by policy makers who aim to reduce poverty. On one hand, it requires a clear understanding of how many children are poor, how poor they are, etc., while on the other hand more appropriate and effective policy measures for child poverty reduction is needed. To do this the country needs an official poverty line to quantify child poverty and its changes. In this context, this dissertation recommends for an urgent initiative in producing and implementing an appropriate official poverty line for monitoring and analysing child poverty in Bangladesh. More generally, it must be realized that policy instruments should aim to interrupt inter- generational poverty transmission. Lessening child poverty requires more specific and central attention as a strategy of all forms of poverty reduction. This is, of course, a somewhat difficult task for a developing country like Bangladesh. secondly, a crucial determining variable of child poverty in Bangladesh has been found to be the household head’s educational level. More resources should be allocated to the educa- tion sector, as well as to how it is designed. One policy suggestion would, therefore, be to take effective and realistic measures to increase the educational or skill level of the household heads 179 (for both sexes) to increase their employability (either salaried or self-employment), thus benefit- ing children by increasing household income. Adult education programs with study loan for the illiterate can play an important role in this regard. In addition, long-term policy measures should be taken for ensuring compulsory primary and secondary education for all children. Cash trans- fers schemes for schooling should be introduced that include all children, regardless of sex and region. This will provide an increase in household income that will presently reduce poverty rate and increase the share of educated adults in the future. Both of these short and long-term impacts will obviously influence intergenerational transmission of poverty. Also, the government should restructure the present education system to make it more effective and attractive to children to reduce school dropout. All policy measures should be directed towards equalizing the differences in access to public services at the divisional, rural and urban level. In particular, policy makers should focus on the children of Rajshahi division, where the child poverty rate is highest. Actions are required to improve the quality and accessibility of education and healthcare (Harper, 2004). thirdly, although this dissertation has not examined health indicators extensively, Chap- ter 5 examined the notion of well-being, which showed that the living conditions of the poor children are not favourable for healthy physical and psychological development and are likely to create health problems. These findings are very consistent, clear and very detailed, as expressed in the participants’ experiences and perceptions in Chapters 6 and 7. For any country, it is evident that child development can be challenged if health problems in childhood are not treated and it can be a primary cause for the development of poverty, and likely to reinforce intergeneration poverty persistence. Therefore, I suggest bringing all poor children under a comprehensive health care system, which will ensure preventive and curative basic health care services such as nutri- tional supplements, food security and preventive and curative health care. What is more, based on the experience of participants expressed in focus group inter- views, this dissertation argues that the accountability and transparency of public action regarding distribution and delivery of services, particularly in health care and education sectors are urgently needed so that corruption and hidden costs can be distinguished through the lense of legitimacy. 180 Thus, access to public services can be ensured and mistrusts of the recipients including poor children can be eliminated. Together with these, other initiatives, to ensure at least access to basic health care and education for the children, are needed to prevent our next generations from being born and growing up and continuing their lives in poverty. fourthly, since child poverty rates differ spatially, all relevant policy measures should be directed towards equalizing the differences in access to public services in all sectors at the divisional, rural and urban level. The development in infrastructure including roads, bridges and electricity is to be speeded up, although Bangladesh has done well in this respect (Sen & Hulme, 2006). In particular, policy makers should focus on the children of Rajshahi division where child poverty is found at the highest rate. Bangladesh is committed to be a “welfare state” according to its constitution; however, it presently lacks resources, not surprisingly, to provide comprehensive welfare provisions and coverage for its huge population, including children. If the country had sufficient resources, perhaps things would be different. It is documented that limited resources is an obstacle in initiating child poverty reduction strategies in Bangladesh, which is very similar to many developing countries. Policies directed to sustainable and rapid economic growth, as ex- perienced in China during later years, would be needed, together with policies to reduce income inequality to diminish extensive child poverty in Bangladesh. Through this the poor children and their carers can benefit from the opportunities that economic growth creates. finally, establishment of good governance and administrative reforms are preconditions for effective policy implementation. Studies highlight that together with limited resources, bad governance leads to bad policies (Hulme & Shephard, 2003), which is very crucial for creating an environment of long term policy implementation. The political instability, corruption, conflict and violence also have been identified to be associated with bad governance in many countries including Bangladesh. Therefore, good governance along with adequate finance is preconditions for “creating an environment for savings, investment, risk-taking and employment creation for all forms of future poverty reduction”(Hulme & Shepherd, 2003; Moore, 2001). More precisely, it would likely then be very useful to have such policy instruments to provide the country’s next generation with a world, at least, free of “absolute poverty”. This cannot be expected with the 181 presence of corruption in the administration and mistrust of the citizens about their activities. 9.1. Few words on the findings from the empirical results Bangladesh has achieved considerable poverty reduction, high economic growth, decrease in mal- nutrition and infant mortality rates, increase in adult literacy, gender parity in school enrolment, etc. since 1990s (see Osmani, 2006; Sen & Hulme, 2006; GRM International, 2010). However, a large proportion of the country’s population, a greater share of which is children, continues to live in poverty. This chapter presents some concluding remarks on the findings of the studies pre- sented in this dissertation: something about their limitations; something about how specific child poverty is in Bangladesh, and what implications they have (which has been described earlier in different studies) on how to study, understand and deal with child poverty. Child poverty is structured within the societies in which it appears and is possible to af- fect by policy measures. A large consensus among observers of the desirability of combating child poverty is evident and child poverty has become a subject of great interest among policymak- ers and researchers alike. However, concerns about child poverty have been more outspoken in richer countries and in the Commonwealth of Independent States than in low- and middle-income countries. In this regard, this dissertation attempts to add to the knowledge of child poverty in the context of developing countries with special attention to Bangladesh. It aimed at, firstly, inves- tigating the extent and the characteristics of child poverty in Bangladesh. Secondly, it sought to get deeper knowledge about the definitions, reasons and impacts of child poverty. Thirdly, child poverty for international comparison was studied, and compared the reasons for the differences of changes of child poverty between Bangladesh and China were analysed and discussed. Finally, this dissertation has included poor children’s and mothers’ perception on what needs to be done to combat or reduce it. Addressing child poverty as a main focus in research has not been very common until the recent past, in comparison with studies on adult poverty in the context of developing countries. Firstly, I attempted to quantify child poverty in Bangladesh and analyze how it varies with dif- 182 ferent variables. To my knowledge, no previous study has attempted to do this. The estimation is procured from national microdata of Bangladesh, HIES—1995 and in particular the HIES—2000. However, it is important for future research to study the development of child poverty in Bangla- desh over time by including, for example, the data of HIES—2005 and the next available round. I acknowledge that this dissertation does not include the information on poor children living out- side households, as they are not included in the survey. These limitations aside, it was found that in the year 2000 the child poverty rate in Bangladesh was 28 percent, while applying a poverty line set to 1 USD in PPP year per day, and 72.8 percent if doubling the poverty line. Secondly, more comprehensive knowledge about child poverty has been gained by com- plementing quantitative and qualitative methods, as it is viewed from more than one standpoint (see Brannen, 2004). I took the advantages of using survey data in this dissertation for counting or identifying poor children, and for analyzing the characteristics of child poverty of Bangladesh and China as well, and comparing these two countries. This contributed to our knowledge about the impact of economic growth and income inequality on poverty, in this case, on child poverty. The main findings are discussed here. first, child poverty in Bangladesh displays a pattern very similar to the ones of other countries in some aspects. For example, similar to many other countries (Dieden & Gustafsson, 2003; Gordon et al., 2003; UNIICEF, 2006) poverty is found to be more significant among the children than the adults. The main findings support that child poverty in Bangladesh should be regarded as a major social problem since the country has a large number of young generation under the age of 14 and they constitute a large proportion of the total population; the poverty rate for children is higher than for adults in terms of both demographic and household characteristics. Child poverty is more extensive in Bangladesh than in China, and is a more heightened problem for rural children in both the countries (see Chapter 5 and 8). second, household heads’ characteristics, especially their educational level and locations of residences, play a vital role for the probability of a child being poor in Bangladesh. The prob- ability of being poor is high (59 percent) if: i) the household lives in the rural part of Rajshaih division, and ii) the head of the household is illiterate. On the contrary, a household living in urban 183 Chittagong with largely educated household head is much less likely to be poor (2 percent). The importance of education for not being in poverty was also very clear in the thoughts of the par- ticipants in the qualitative study. The qualitative studies emphasize the poor groups’ participants’ trust in the importance of education, as they expressed their belief that having educated house- hold members reduces child poverty. third, one of the most striking themes emerging from the discussions is that defining child poverty in terms of the World Bank’s definition of International Poverty Line is very restrict- ed, compared with what was derived from the participants of focus group interviews (Chapter 6). Although this poverty line is widely used for measuring and analyzing poverty, child poverty, in particular, requires, according to the participants of focus group discussions, addressing more is- sues than only income. At this point, child poverty may be defined as: the state of inadequacy of households’ earning to pay for food and house rent, including inability to live in a safe shelter, and living under miserable conditions, such as the footpath, instead. In addition, the victim’s voice indicates that child poverty is also associated with the feeling of misery, dishonour or humiliation and also a matter of disempowerment because of being poor. Thus, by making some victims’ voices heard, this study adds other dimensions to poverty research. fourth, as evaluated by the World Bank’s one-dollar per day poverty line, the extensive child poverty in Bangladesh mainly due to the higher inequality in income distribution and lower level economic growth compared to China. Of relevance to the two countries studied, economic growth and differences in income levels are found to affect the level of child poverty differences across time and across countries (Chapter 7). However, the same study also shows that though the differences in mean income play an important role, differences in demographic compositions (such as parental education and number of children) also affect child poverty differences across developing countries. fifth, in addition to statistical inferences, this dissertation also examined child poverty from some of its principal victims’ perspectives about how they conceptualize and experience poverty. The knowledge on child poverty in this study is partly generated and grounded on some 184 of the principal victims’ own experience of and reflections on child poverty (Chapter 6 and 8). I believe that it is useful to inquire and put emphasis on their expressed needs and thoughts, as they are the main actors who deal with child poverty in their everyday lives. To listen and to raise some of the voices of the victims of child poverty, focus group interviews are appropriate method to maximize the quality and quantity of data (see Denzin & Lincoln, 2000). sixth, the participants’ conceptualization of child poverty, and its causes and impacts are not very different to what child poverty studies in other parts of the world show (see Chant, 2006; Kalaycioglu, 2006). Although the experience of child poverty, as demonstrated in focus group discussions in Bangladesh, is very common in the case of many other developing countries, poor children in rich countries experience it differently in several important aspects. For example, the children participants in my focus groups expressed their experiences of extreme hardship such as starvation just like many poor children in developing countries. However, this is extremely unusual for child poverty in rich countries. seventh, gender differences were not well studied in the quantitative study as there are limitations to examine this issue in survey data that I used. However, qualitative studies present some reflection on this. The quantitative study in Chapter 5 is about quantifying child poverty and reporting its profile, and it does not focus on causes of poverty. The qualitative study can help to sort out possible causes in this context. However, it cannot quantify the importance of it. There- fore, one can arrive at hypotheses from the qualitative study, and then design (another) quantita- tive study. Qualitative study on the poor men can be used to investigate how hard facts (e.g. low income) are felt and experienced. Poor children are found, in qualitative studies; to face gender discrimination, to be the victims of child abuse and humiliation, and to lack empowerment in their everyday life. Other significant aspects are found to contribute to child poverty such as corruption, mistrust and child abuse. Similar to in many other developing countries, many factors are found to further aggravate an already dire situation by limiting poor children’s access to existing public services and facili- ties. What is common in all studies that are presented in this dissertation is that low level of paren- tal or household head’s education is very important in contributing to child poverty in Bangladesh 185 as in many developing countries. eighth, notably, poverty in Bangladesh is noticed to be transferred in a system of inter- generational poverty (Chapter 6)—from parents to their children, further on to their children and so on. Three different dimensions (encompassing multidimensionality, spending life in distress, and having intergenerational and gendered dimension) are found to be interconnected with the progression for children’s residing in poverty (Chapter 6). Child poverty reduction has been iden- tified as one of the main challenges of putting a ceiling on IGT of poverty. This is a risk for all poor children in the world, no matter where they live. But, it is probably greater in developing countries than in welfare states that ensure children’s access to basic needs at a minimum standard including comprehensive schools as defined by the states (for example Sweden and other Scan- dinavian countries). In this process, parental poverty, low education, fewer resources, and gender discrimination altogether put children’s lives in distress and in poverty, which is very common for the poor children in the developing countries as well as in many developed countries. ninth, to raise some of the victims’ voices and to involve them in the policy making process, this dissertation (Chapter 7) takes the responsibility to present the opinions of a limited number of poor children about what they really need and what they expect from the policy mak- ers. Not surprisingly, participants claimed that public services for the poor children are out of reach for many of them. The data from these groups provided some policy recommendations to increase the capability (for example, universal education) of the children for human development that would aid to interrupt intergenerational transmission of the poverty process. Additionally, they suggested for adult education, and women’s education to enable them to gain access to more work and earnings. One interesting point is that the majority of our discussion groups’ recom- mendations were initiatives directed towards increasing parental income with the exception of universal education for children. In their opinion, providing financial supports , taking empower- ment measures, stopping corruption etc. to build parents’ capacities may contribute to increase their income. A labour intensive industrialization, particularly, at rural areas, has been another alternative suggestion for ensuring job availability and increased income. finally, the secondary data on public policy and program in Bangladesh’s bureaucratic 186 system of administration, and empirical data show that there is a gap between reality and existing program in propaganda. One of the malfunctions of bureaucracy, as in many developing coun- tries, is corruption, along with other problems such as lack of resources; it creates an environment of failure in implementing a huge number of poverty alleviation programs in Bangladesh. The participants spoke of, for example, some hidden costs in public services, which should be free for the poor children according to the state policy, and they claimed those to be due to corruption. However, this also might be a matter of mistrust of the poor towards public administration. If these hidden costs are legitimate, the lack of authentic information about legal fees of concerned public services is one of the main reasons to form this kind of mistrust. 9.2. Policy recommendations Children represent a country’s future, which is why their well-being is a societal concern; they cannot change their situations on their own and collective action is needed to mitigate vulnerable situations in poverty (Bradbury et al., 2001, p. 1). This is applicable to all children no matter where they live. The following steps are crucial to reduce child poverty in Bangladesh in view of the studies included in this dissertation and my value judgment. Where child poverty is found to be extensive and children constitute a greater share of population, child poverty reduction should be, in practice, first and foremost: 1) central to all pov- erty reduction policy agenda; and 2) given the highest priority by policy makers who aim to re- duce poverty. On one hand, it requires a clear understanding of how many children are poor, how poor they are, etc., while on the other hand more appropriate and effective policy measures for child poverty reduction is needed. To do this the country needs an official poverty line to quantify child poverty and its changes. In this context, this dissertation recommends for an urgent initia- tive in producing and implementing appropriate official poverty line for monitoring and analysing child poverty in Bangladesh. More generally, it must be realized that policy instruments should aim to interrupt inter- generational poverty transmission. Lessening child poverty requires more specific and central 187 attention as a strategy of all forms of poverty reduction. This is, of course, a somewhat difficult task for a developing country like Bangladesh. secondly, a crucial determining variable of child poverty in Bangladesh has been found to be the household head’s educational level. More resources should be allocated to the educa- tion sector, as well as to how it is designed. One policy suggestion would, therefore, be to take effective and realistic measures to increase the educational or skill level of the household heads (for both sexes) to increase their employability (either salaried or self-employment), thus ben- efiting children by increasing household income. Adult education programs with study loan for the illiterate can play an important role in this regard. In addition, long-term policy measures should be taken for ensuring compulsory secondary education for children. Long-term policy measures should be taken to ensure primary and secondary education for all children. Cash trans- fers schemes for schooling should be introduced that include all children, regardless of sex and region. This will provide an increase in household income that will presently reduce poverty rate and increase the share of educated adults in the future. Both of these short and long-term impacts will obviously influence intergenerational transmission of poverty. Also, the government should restructure the present education system to make it more effective and attractive to children to reduce school dropout. All policy measures should be directed towards equalizing the differences in access to public services at the divisional, rural and urban level. In particular, policy makers should focus on the children of Rajshahi division, where the child poverty rate is highest. Actions are required to improve the quality and accessibility of education and healthcare (Harper, 2004). thirdly, this dissertation has not examined health indicators extensively. However, Chapter 5 examined the notion of well-being, which shows that the living conditions of the poor children are not favourable for healthy physical and psychological development and are likely to create health problems. These findings are very consistent, clear and very detailed, as expressed in the participants’ experiences and perceptions in Chapters 6 and 7. For any country, it is evident that child development can be challenged if health problems in childhood are not treated and it can be a primary cause for the development of poverty, and likely to reinforce intergeneration 188 poverty persistence. Therefore, I suggest bringing all poor children under a comprehensive health care system, which will ensure preventive and curative basic health care services such as nutri- tional supplements, food security and preventive and curative health care. What is more, based on the experience of participants expressed in focus group inter- views, this dissertation argues that the accountability and transparency of public action regarding distribution and delivery of services, particularly in health care and education sectors are urgently needed so that corruption and hidden costs can be distinguished through the lense of legitimacy. Thus, access to public services can be ensured and mistrusts of the recipients including poor children can be eliminated. Together with these, other initiatives, at least to ensure access to basic health care and education for the children, are needed to prevent our next generations to be born and grow up and continue their lives in poverty. fourthly, since child poverty rates differ spatially, all relevant policy measures should be directed towards equalizing the differences in access to public services in all sectors at the divisional, rural and urban level. The development in infrastructure including roads, bridges and electricity is to be speeded up, although Bangladesh has done well in this respect (Sen & Hulme, 2006). In particular, policy makers should focus on the children of Rajshahi division where child poverty is found at the highest rate. Bangladesh is committed to be a “welfare state” according to its constitution; however, it presently lacks resources, not surprisingly, to provide comprehensive welfare provisions and coverage for its huge population, including children. If the country had sufficient resources, perhaps things would be different. It is documented that limited resources is an obstacle in initiating child poverty reduction strategies in Bangladesh, which is very similar to many developing countries. Policies directed to sustainable and rapid economic growth, as ex- perienced in China during later years, would be needed, together with policies to reduce income inequality to diminish extensive child poverty in Bangladesh. Through this the poor children and their carers can benefit from the opportunities that economic growth creates. finally, establishment of good governance and administrative reforms are preconditions for effective policy implementation. Studies highlight that together with limited resources, bad governance leads to bad policies (Hulme & Shephard, 2003), which is very crucial for creating 189 an environment of long-term policy implementation. The political instability, corruption, conflict and violence also have been identified to be associated with bad governance in many countries including Bangladesh. Therefore, good governance along with adequate finance is preconditions for “creating an environment for savings, investment, risk-taking and employment creation for all forms of future poverty reduction”(Hulme & Shepherd, 2003; Moore, 2001). More precisely, it would likely then be very useful to have such policy instruments to provide the country’s next generation with a world, at least, free of “absolute poverty”. This cannot be expected with the presence of corruption in the administration and mistrust of the citizens about their activities. 9.3. Summing up The potential of millions of children around the world who live in poverty is left unexplored and unrealized. The children interviewees’ hopes or dreams of becoming pilots, doctors, engineers or teachers will most likely also be unrealized, since they may have already dropped out of school. Could we or should we ensure that these children realize their dreams? Could we or should we guarantee their education as long as they want to continue? Do we or should we agree that the rights declared in the Convention on the Rights of the Child should be realized? And if so, how do we do this? By placing child poverty reduction at the centre of the policy agenda, this dissertation suggests: an alternative, capable and corruption-free public administration for a faster economic growth of Bangladesh; an increase in income level and a reduction in income inequality by re- distributing wealth from the rich to the poor, to ensure its children a better standard of living, together with “a comprehensive social provision of health care, education and social protection” (Harper, 2004). These steps are very important to tackle “deprivation” in extensive and multidi- mensional child poverty, to ensure the “entitlement” to access basic human needs of children and their households, to empower them and to increase their “capabilities” to function or “participate in the society”. Regulatory frameworks need to be reformed and their realistic implementation is required to combat all forms of gender discrimination, humiliation and abuse against the children, 190 with particular attention to the poor. Mass media and other development partners can play a vital role in this regard. With this in view followed by the encouragement of media interest and cover- age of the issue are therefore needed to include in bringing about change. Particularly, poverty alleviation programmes should aim to break the vicious cycle of intergenerational poverty by targeting poor children and their parents. These measures need to include universal education and healthcare for children, increased and guaranteed parental or household income, the elimination of gender discrimination on both the household and the nation- al level, paying special attention to the education of girls by creating awareness of its importance. Good governance, together with adequate finances, are preconditions for creating an environment where all forms of future poverty reduction can flourish. Bangladesh is in urgent need of corrup- tion-free public administration in order to implement its poverty alleviation policies successfully and bring its children out of the vicious cycle of poverty. These measures are essential to ensure that the country’s next generations are born and grow up at least with suitable shelter, with adequate and nutritious food, with appropriate cloth- ing, and with minimum health care and education necessary for children’s growth and develop- ment. These policy recommendations are neither very easy tasks to implement, nor impossible to achieve. Political commitment for a solidaristic society and its implementation in reality are es- sential both from the political parties in the government and the opposition, where redistribution and proper utilization of existing resources will serve as main sources of resource allocation. This will probably provide one of the most important ways to prevent our children and their children to be trapped in the vicious cycle of extensive child poverty from generations to generations. 191 References ADB. (2001). Rural development priorities for poverty reduction in Bangladesh. Resident Mis- sion, Dhaka: Asian Development Bank (ADB). Retrieved June 1, 2010, from http://www.adb.org/documents/epps/ban/rural_development/rural_development.pdf Afsar, R. (1999). Rural-urban dichotomy and convergence: Emerging realities in Bangladesh. Environment and Urbanization, 11 (1), 235-46. Ahmed, S. M. (2005). Exploring health-seeking behavior of disadvantaged populations in rural Bangladesh. Doctoral dissertation, Stockholm: Karolinska University Press. Akram-Lodhi, A. H., Borras, S. M. Jr., Kay, C., & McKinley, T. (2007). Neoliberal globalization, land and poverty: Implications for public action. In A. H. Akram-Lodhi, S. M. Borras Jr., & C. Kay, (Eds.), Land, poverty and livelihoods in an era of globalization: Perspectives from developing and transition countries (Chapter 12, pp. 383-96). New York: Routledge. Amin, S., & Basu, A. M. (2004). Popular perceptions of emerging influences on mortality and longevity in Bangladesh and West Bengal. Population Studies, 58, (3), 357-64. Amin, S., Rai, A. S., & Topa, G. (2003). Does microcredit reach the poor and vulnerable? Evidence from northern Bangladesh. Journal of Development Economics, 70 (1), 59-82. Anand, S., Segal, P., & Stiglitz, J. E. (Eds.). (2010). Debates on the measurement of global pov- erty, Oxford: Oxford University Press. Anam,T. (2007, August 14). Trouble water in Bangladesh. BBC News South Asia. [Online]. Re- trieved June 1, 2010, fromhttp://news.bbc.co.uk/2/hi/south_asia/6943873.stm Appleton, S., Song, L. & Xia, Q. (2010). Growing out of poverty: Trends and patterns of urban poverty in China 1988-2002. World Development, 38 (5), 665-678. Atkinson, A. B. (1987). On the measurement of poverty. Econometrica, 55 (4), 749-64. Atkinson, T., Cantillon, B., Marlier, E., & Nolan, B. (2002). Social indicators: The EU and social inclusion. Oxford: Oxford University Press. Attree, P. (2006). The social costs of child poverty: A systematic review of the qualitative evi- dence. Children & Society, 20 (1), 54-66. Backman, K., & Kyngäs, H. A. (1999). Challenges of the grounded theory approach to a novice researcher. Nursing and Health Sciences, 1 (3), 147-53. Banerjee, L. (2007). Floods disasters and agricultural wages in Bangladesh. Development and Change, 38 (4), 641-64. Barrientos, A., & DeJong, J. (2006). Reducing child poverty with cash transfers: A sure thing? Development Policy Review, 24 (5), 537-52. Baulch, B., & Hoddinott, J. (2000). Economic mobility and poverty dynamics in developing countries. Journal of Development Studies, 36 (6), 1–24. 192 BBS. (2003). Report of the household income & expenditure survey, 2000. Dhaka: Bangladesh Bureau of Statistics (BBS). BBS. (2006). Preliminary report on the household income and expenditure survey, 2005. Dhaka: Bangladesh Bureau of Statistics (BBS). BBS. (2007). Statistics Bangladesh, 2006. Dhaka: Bangladesh Bureau of Statistics (BBS). BBS. (2007a). Bangladesh data sheet. Dhaka: Bangladesh Bureau of Statistics (BBS). BBS. (2008). Key indicators on Report of Sample Vital Registration System (SVRS) 2008. Dhaka: Bangladesh Bureau of Statistics (BBS). BBS. (2009). Statistical pocket book of Bangladesh, 2008. Dhaka: Bangladesh Bureau of Statis- tics (BBS). BBS. (2011). Preliminary Repost on Household Income & Expenditure Survey-2010. Dhaka: Bangladesh Bureau of Statistics (BBS). Begum, S. S. (2005, April 25-27). How children experience poverty: A study on child poverty in Bangladesh. Paper presented at the international conference of UNICEF and the Gradu- ate Programme in International Affairs (GPIA). New York: New School University. Begum, S. S., Deng, Q., & Gustafsson, B. (2012). Economic growth and child poverty reduction in Bangladesh and China. Journal of Asian Economics, 23, 73-85 Berg, B. L. (2007). Qualitative research methods for the social sciences (6th Ed.). Boston: Allyn & Bacon. Bhuyan, A. R., Khan, H. A. R., & Ahmed, S.U. (2001). Rural urban migration and poverty: The case for reverse migration in Bangladesh (Monitoring Adjustment and Poverty [MAP] Focus Study Series 10). Dhaka: Centre on Integrated Rural Development for Asia and the Pacific (CIRDAP). Bigsten, A., & Shimeles, A. (2007) Can Africa reduce poverty by half by 2015? Development Policy Review, 25 (2), 147–66. Bilson, A., & Cox, P. (2007).Caring about poverty: Alternatives to institutional care for children in poverty. Journal of Children and Poverty, 13 (1), 37-55. Bird, K. (2007, November). The intergenerational transmission of poverty: An overview (Chro nic Poverty Research Centre [CPRC]Working Paper 99). Manchester: CPRC. Blinkenberg, L. (1998). India-Pakistan: The history of unsolved conflicts (Vol. 1). Odense, Denmark: Odense University Press. Booth, C. (1892). Life and labour of the people in London (1889–1903). London: Macmillan. Bourguignon, F. (2004, February 04). The poverty-growth-inequality triangle. Paper presented at the Indian Council for Research on International Economic Relations. New Delhi, India. Bourguignon, F., & Chakravarty, S. R. (2003). The measurement of multidimensional poverty. Journal of Economic Inequality, 1 (1), 25-49. 193 Boyden, J., & Cooper, E. (2009). Questioning the power of resilience: Are children up to the task of disrupting the transmission of poverty? In T.Addison, D.Hulme, & R.Kanbur (Eds.) Poverty dynamics: Interdisciplinary perspectives (Part 3, Chapter 13, pp. 289- 308).Oxford: Oxford University.Press. Bozeman, B. (2000) Bureaucracy and red tape. Upper Saddle River, NJ: Prentice Hall. Bradbury, B. (1999). Child poverty across the industrialised world. Family Matters, 54(Spring/ Summer), 65-69. Bradbury, B., Jenkins, S. P., & Micklewright, J. (Eds.). (2001). The dynamics of child poverty in industrialized countries. Cambridge: Cambridge University Press. Brannen, J. (2004). Working qualitatively and quantitatively. In C.Seale, G. Gobo, J. Gubrium, & D. Silverman (Eds.), Qualitative research practice (Part 4, Chapter 20, pp. 312-26). London: Sage Publications. Calás, M. B., & Smircich, L. (2006). From the “woman’s point of view” ten years later: Toward a feminist organization studies. In S. Clegg, C. Hardy, & W. Nord, (Eds.), The Sage hand- book of organization studies (2nd ed., Chapter 8, pp. 284-346). London: Sage Publica- tions. Cameron, S. (2011). Whether and where to enrol? Choosing a primary school in the slums of urban Dhaka, Bangladesh. International Journal of Educational Development, 31 (4), 357–366. Campaign for Popular Education (CAMPE). (2007). Education watch 2006: Financing primary and secondary education in Bangladesh: Overview. Dhaka: CAMPE. Case, A. (2006). The primacy of education. In A. V. Banerjee, R. Bénabou, & D. Mookherjee (Eds.), Understanding poverty (Part II, Chapter 18, pp. 269-84). Oxford: Oxford Univer- sity Press. Chan, K. W., & Buckingham, W. (2008). Is China abolishing the Hukou System? The China Quarterly,195 (September), 582–606. Chant, S. (2006). Contributions of a gender perspective to the analysis of poverty. In J. S. Ja- quette & G. Summerfield (Eds.), Women and gender equity in development theory and practice:Institutions, resources, and mobilization (Chapter 1, pp. 87-106). Durham: Duke University Press. Charmaz, K. (2005). Grounded theory in the 21st century: A qualitative method for advancing social justice research. In N. K. Denzin & Y. S. Lincoln (Eds.), The Sage handbook of qualitative research (3rd ed., Chapter 20, pp. 507-35). Thousand Oaks, California: Sage Publications. Charmaz, K. (2006). Constructing grounded theory. A practical guide through qualitative analy- sis. London: Sage Publications. Chen, S., & Ravallion, M. (2001). How did the world’s poorest fare in the 1990s? Review of In- come and Wealth, 47 (3), 283-300. Chen, S., & Ravallion, M. (2004). How the world’s poorest fared since the early 1980s. The World Bank Research Observer, 19 (2), 141-69. 194 Chen, S., & Ravallion, M. (2010). China is poorer than we though, but no less successful in the fight against poverty. In S. Anand; P. Segal, & J. E. Stiglitz (Eds.), Debates on the mea- surement of global poverty (Part II, Chapter 13, pp. 327-40). Oxford: Oxford University Press. Chen, S., & Ravallion, M. (2010a). The developing world is poorer than we thought, but no less successful in the fight against poverty. Quarterly Journal of Economics, 125 (4), 1577- 1625. Chen, G., & Williams, D. W. (2007). How political support influences red tape through devel- opmental culture. The Policy Studies Journal,35 (3), 419-36.Conticini, A., & Hulme, D. (2007). Escaping violence, seeking freedom: Why children in Bangladesh migrate to the street. Development and Change, 38 (2), 201–27. Corak, M. (2006). Do poor children become poor adults? Lessons from a cross country com- parison of generational earnings mobility. In J. Creedy & G. Kalb (Eds.) Dynamics of inequality and poverty (Vol. 13), Research on economic inequality (Part II, pp. 143-88). Oxford: Elsevier Ltd. Corak, M. (2006a). Principles and practicalities for measuring child poverty.International Social Security Review, 59 (2), 3-35. Corak, M., Fertig, M., & Tamm, M. (2008). A portrait of child poverty in Germany. Review of Income and Wealth, 54 (4), 547–571. Corcoran, M. E., & Chaudry, A. (1997).The dynamics of childhood poverty. Journal of Children and Poverty, 7 (2), 40-54. Corbin, J., & Strauss, A. L. (2008). Basics of qualitative research (3rd ed.). ThousandOaks, Cali- fornia: Sage Publications. Coudouel, A., Hentschel , J. S., & Wodon, Q. T. (2002). Poverty measurement and analysis. In J. Klugman, (ed.),Asourcebookforpovertyreductionstrategies(Vol. I), Core techniques and cross-cutting issues (Part 1, Chapter 1, pp. 29-74). Washington, DC: The International Bank for Reconstruction and Development/The World Bank. Cournoyer, D. E., & Klein, W. C. (2000). Research methods for social work. Allyn Bacon. CPRC. (2004-05). Chronic poverty report 2004-05. Manchester: Chronic poverty research centre (CPRC). Datt, G., & Ravallion, M. (1992). Growth and redistribution components of changes in poverty: A decomposition to Brazil and India in the 1980s. Journal of Development Economics, 38 (2), 275–295. Danziger, S., & Gottschalk, P. (1995). America unequal. Cambridge, MA: Harvard University Press. Davis, P. (2009). Poverty in time: exploring poverty dynamics from life history interviews in Bangladesh.In: T.Addison, D.Hulme, & R.Kanbur (Eds.) Poverty dynamics: Interdisci- plinary perspectives (Part II, Chapter 7, pp. 154-82). Oxford: Oxford University Press. Deaton, A. (1997). The analysis of household surveys: A micro econometric approach to develop- ment policy.Baltimore, MD: Published for the World Bank (by) Johns Hopkins University Press. 195 Deaton, A. (2006). Measuring poverty. In A. V. Banerjee, R. Bénabou, & D. Mookherjee (Eds.), Understanding poverty (Chapter 1, pp. 3-16). Oxford: Oxford University Press. Denscombe, M. (1998). The good research guide for small-scale social research projects. Buck- ingham: Open University Press. Denzin, N. K., & Lincoln, Y. S. (2000). Handbook of Qualitative Research (2nd ed.). Thousand Oaks, California: Sage Publications. Deolalikar, A. B. (2005). Poverty and child malnutrition in Bangladesh. Journal of Developing Societies, 21 (1-2), 55-90. Desta, Y. (2007). Designing anti-corruption strategies for developing countries: A country study of Eritrea. Journal of Developing Societies,22 (4), 421-49. Devarajan, S., & Nabi, I. (2006). Economic growth in South Asia: Promising, unequalizing, … sustainable? Washington DC:The World Bank. Dey, I. (2007). Grounding categories. In A. Bryant & K. Charmaz (Eds.), The Sage handbook of grounded theory (Part III, Chapter 8, pp. 167-90). London: Sage Publications. Dieden, S., & Gustafsson, B. (2003). Child poverty in South Africa: An assessment based on microdata for 1995. International Journal of Social Welfare, 12 (4), 326-38. Dollar, D., & Kraay, A. (2002). Growth is good for the poor. Journal of Economic Growth, 7 (3), 195-225. Dowlah, CAF. (2002). Dynamics of food-assisted development strategies in Bangladesh. Interna- tional Journal of Social Welfare, 11 (1), 3-21. Dowlah, CAF. (2006). The politics and economics of food and famine in Bangladesh in the early 1970s: With special reference to Amartya Sen’s interpretation of 1974 Famine. Interna- tional Journal of Social Welfare,15 (4), 344-56. EC. (2007, December). Bangladesh -European Community: Country strategy paper for the pe- riod 2007-2013 (European Commission [EC] Country Strategy Paper for Bangladesh). Brussels: EC. Retrieved June 06, 2010, from http://ec.europa.eu/external_relations/ban- gladesh/csp/csp_07_13_en.pdf Eichen, M., & Zhang, M. (1993). Annex. The 1988 Household Sample Survey: Data description and availability. In K. Griffin &Z. Renwei (Eds.), The distribution of income in China (pp. 341-46). London:Macmillan. Engle, P. L., Castle, S., & Menon, P. (1996). Child development: Resilience (FCND Discussion Paper 12). Washington, DC: TheInternational Food Policy Research Institute (IFPRI). Eswaran, M. (2006). Fertility in developing countries. In A. V. Banerjee, R. Bénabou, & D. Mookherjee (Eds.), Understanding poverty (Part 1, Chapter 10, pp. 143-60). Oxford: Oxford University Press. Ferreira, F. (2010). Distributions in motion: Economic growth, inequality, and poverty dynamics (ECINEQ Working Paper 183). ECINEQ, Society for the Study of Economic Inequality. Fisher, G. M. (1995). Is there such a thing as an absolute line over time? Evidence from the 196 United States, Britain, Canada, and Australia, on the income elasticity of the poverty line. US Bureau of the Census, unpublished. http://www.census.gov/hhes/www/povmeas/pa- pers/elastap4.html Fortier, S. M. (2006). On being a poor child in America: views of poverty from 7-12-year-olds. Journal of Children and Poverty, 12 (2), 113-27. Foster, J. E. (1998). What is poverty and who are the poor? Redefinition for the united states in the 1990’s: Absolute versus relative poverty. AmericanEconomic Review, 88 (2), 335-41. Foster, J., Greer, J., & Thorbecke, E. (1984). A class of decomposable poverty measures. Econo- metrica, 52 (3), 761-66. Fosu, A. K. (2011). Growth, inequality, and poverty reduction in developing countries: Recent global evidence (UNU-WIDER Working Paper No. 2011/01). United Nations University. Glaser, B.G. (2007). Doing formal theory. In A. Bryant & K. Charmaz, (Eds.), The Sage handbook of grounded theory ( Part II, Chapter 4, pp. 97-113). London: Sage pub- lications. Glaser, B. G,. & Holton, J. (2004). Remodeling grounded theory. Forum: Qualitative Social Re- search, 5(2), Art. 4 Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for qualita- tive research.Chicago: Aldine De Gruyter. Goedhart, T., Halberstadt, V., Kapteyn, A., & Van Praag, B. (1977). The poverty line: Concept and measurement, Journal of Human Resources, 12 (4), 503-20. Gordon, D., Nandy, S., Pantazis, C. Pemberton, S., & Townsend, P. (2003). Child poverty in the developing world. Bristol: The Policy Press. Green, D. O., Creswell, J. W., Shope, R. J., & Clark, V. L. P. (2007). Grounded theory and racial/ethnic diversity. In A. Bryant, & K. Charmaz (Eds.), The Sage handbook of grounded theory (Part V, Chapter 22, pp. 472-92). London: Sage publications. Grindle, M. S. (ed.). (1980). Politics and policy implementation in the Third World. Princeton, NJ: Princeton University Press. GRM International. (2010). Bangladesh reality check 2009: Listening to poor people’s realities about primary healthcare and primary education – Year 3. Bangladesh: Swedish Inter- national Development Cooperation Agency (SIDA). Retrieved August 10, 2010, from http://www.sida.se/Global/Countries%20and%20regions/Asia%20incl.%20Middle%20 East/Bangladesh/SIDA61258en_Reality%20Check%20Bangladesh_%20Web%20.pdf Gustafsson, B. A. (1995). Assessing poverty: Some reflections on the literature. Journal of Popu- lation Economics, 8 (4), 361-81. Gustafsson, B. A., Li, S., & Sato, H. (2006). Can a subjective poverty line be applied to China? Assessing poverty among urban residents in 1999. In H. Sato and S. Li. Unemployment, Inequality and Poverty in Urban China (Part I, Chapter 8, pp. 152-72). Oxon: Routledge. 197 Haddad, L., Hoddinott, J., & Alderman, H. (1997). Intrahousehold resource allocation in develop- ing countries: models, methods, and policies.Baltimore, MD: Published for the Interna- tional Food Policy Research Institute (by) Johns Hopkins University Press. Hagenaars, A. J. M.(1986). The perception of poverty. Amsterdam: North-Holland. Haider, M. (2008). Recognising complexity, embracing diversity: working children in Bangladesh. South Asia Research, 28 (1), 49–72. Hanmer, L. C., Pyatt, G., & White, H. (1999). What do the World Bank’s poverty assessments teach us about poverty in Sub-Saharan Africa? Development and Change, 30 (4), 795- 823. Haque, A. K. M. M. (2008). Emergency powers and caretaker government in Bangladesh. Jour- nal of the Australasian Law Teachers Association, 1 (1-2), 81-94. Hardning, A., & Szukalska, A. (2000). Trends in child poverty in Australia, 1982 to 1995-1996. The Economic Record, 76 (234), 236-254. Harper, C. (2004). Breaking poverty cycles: The importance of action in childhood (Childhood Poverty Policy and Research Centre [CHIP] Briefing 8). London: CHIP. Haughton, J., & Khandker, S. R. (2009). Handbook on poverty and inequality. Washington, DC: The International Bank for Reconstruction and Development /The World Bank. Hayati, D., Karami, E., & Slee, B. (2006). Combining qualitative and quantitative methods in the measurement of rural poverty: The case of Iran. Social Indicators Research, 75 (3), 361-94. HDR. (2009). Human development report 2009. Overcoming barriers: Human mobility and de- velopment. New York: Published for the United Nations Development Programme by Palgrave Macmillan. HDR. 2010. Human Development Report 2010. 20th anniversary edition. The real wealth of nations: Pathways to human development. New York: Published for the United Nations Development Programme by Palgrave Macmillan. HDR. 2011. Human Development Report 2011. Sustainability and equity: A better future for all New York: Published for the United Nations Development Programme by Palgrave Macmillan. Himanshu. (2008). World Bank’s new poverty estimates: What are these new poverty estimates and what do they imply? Economic and Political Weekly, 43 (43), 38-43. Retrived July 15, 2010, from http://www.csh-delhi.com/team/downloads/publiperso/himanshu_world- poverty_EPW.pdf Hjalmarson, H. V., Strandmark, M., & Klässbo, M. (2007). Healthy risk awareness motivates fracture prevention behavior: A grounded theory study of women with osteoporosis. In- ternational Journal of Qualitative Studies on Health and Well-being, 2 (4), 236-45. Holton, J. A. (2007). The coding process and its challenges. In A. Bryant & K. Charmaz (Eds.), The Sage handbook of grounded theory (Part III, Chapter 13, pp. 265-89). London: Sage publications. Howe, G., & Mckay, A. (2006). Combining quantitative and qualitative methods in assessing 198 chronic poverty: The case of Rwanda. World Development, 35 (2), 197–211. Hulme, D., & Shepherd, A. (2003). Conceptualizing chronic poverty. World Development, 31 (3),403-23. Iceland, J. (2003). Why poverty remains high: The role of income growth, economic inequality, and changes in family structure, 1949-1999. Demography,40 (2), 499–519. Iceland, J. (2006). Poverty in America: A handbook (2nd ed.). Berkeley: University of California Press. Johansson, S., & Nygren, O. (1991). The missing girls of China: A new demographic account. Population and Development Review, 17 (1), 35-51. Ilon, L. (2000). Colonial secondary education in a global age: Economic distortions in Bangla- desh. Asia Pacific Education Review, 1 (1), 91-99. Islam, A. (2004). Overcoming poverty in Bangladesh: Search for a new paradigm. Bangladesh e-Journal of Sociology 1 (2), 1-20. Rettived September 30, 2008, from http://www.bangladeshsociology.org/Sardar%20%20poverty%20paper%20vol.%20 1.2,%20PDF.pdf Kabeer, N. (1998). Jumping to conclusions? Struggles over meaning and method in the study of household economics. In C. Jackson & R. Pearson (Eds.), Feminist visions of development: Gender analysis and policy (Part II, Chapter 4, pp. 91-107). London: Rout- ledge. Kabeer, N. 2003. Deprivation, discrimination and delivery: Competing explanations for child labour and educational failure in South Asia. In N. Kabeer, G. B. Nambissan, & R. Sub- rahmanian (Eds.), Child labour and the right to education in South Asia: Needs versus rights (pp. 351–93). New Delhi: Sage. Kabeer, N.,&Mahmud, S. (2009). Imagining the future: Children, education and intergenerational transmission of poverty in urban Bangladesh. IDS Bulletin,40 (1), 10-21. Kakwani, N. (2006, September). What is poverty (International Poverty Centre [IPC]One pager 22)? Brazil: IPC, UNDP. Retrieved June 04, 2010, from http://www.undp-povertycentre. org/pub/IPCOnePager22.pdf Kalaycıoğlu, S. (2006). Dynamics of poverty in Turkey: Gender, rural urban, social networks and survival strategies. In M. M. Petmesidou&C. Papatheodorou (Eds.), Poverty and social deprivation in the Mediterranean: Trend, policies and welfare prospects in new millen- nium (Part I, Chapter 8, pp. 189-218). CROP International Studies in Poverty Research. London: Zed books. Kaufmann, D., Montoriol-Garriga, J., & Recanatini, F. (2008, January). How does bribery af- fect public service delivery? Micro-evidence from service users and public officials in Peru (World BankResearch Working PaperWPS4492). Washington DC: World Bank. Re- trived August 1, 2010, from http://www-wds.worldbank.org/servlet/WDSContentServer/ WDSP/IB/2008/01/28/000158349_20080128090947/Rendered/PDF/wps4492.pdf Khan, A. (2008). Growth, inequality and poverty. A comparative study of China’s experience in the periods before and after the Asian crisis. In , B. A. Gustafsson, . S. Li, &T. Sicular(Eds.), Inequality and public policy in China (Chapter 6, pp. 145-81). Cambridge: Cambridge 199 University Press. Khandker, S. R., & Koolwal, Gayatri B. (2010): How infrastructure and financial institutions affect rural income and poverty: Evidence from Bangladesh. Journal of Development Studies, 46 (6), 1109-1137. Knight, J. & Song, L. (1999). The rural–urban divide: Eeconomic disparities and interactions in China. Oxford: Oxford University Press. Kraay, A. (2006). When is growth pro-poor? Evidence from a panel of countries. Journal of De- velopment Economics, 80 (1), 198-227. Krueger, R. A. (1997).Analyzing and reporting focus group results (Focus Group Kit, Vol. 6). Thousand Oaks, California: Sage Publications. Krueger, R. A., & Casy, M. A. (2000). Focus groups: A practical guide for applied research (3rd edition).Thousand Oaks, CA Sage Publications. Laffont, Jean-Jacques. (2006). Corruption and development. In A. V. Banerjee, R. Bénabou, & D. Mookherjee (Eds.), Understanding poverty (Part I, Chapter 11, pp. 161- 67). Oxford: Oxford University Press. Landell-Mills, P. (2002). Taming Leviathan - Reforming governance in Bangladesh: An institu- tional review. Dhaka: World Bank. Leisering, L., & Leibfried, S. (1999). Time and poverty in western Welfare States: United Ger- many in perspective. Cambridge: Cambridge University Press. Lipsky, M. (2004). Street-level bureaucracy: The critical role of street-level bureaucrats. In J. M. Shafritz, A. C. Hyde, & S. J. Parkes (Eds.), Classics of public administration (5th ed., Part II, Chapter 40, pp. 414-22). Belmont, CA: Wadsworth/Thomson Learning. (Original work published 1980). Li, S., Chuliang, L., Zhong, W., & Ximing, Y. (2008). Appendix. The 1995 and 2002 Household Surveys: Sampling method and data description. In B. A. Gustafsson, S. Li, & T. Sicular (Eds.), Inequality and public policy in China (pp. 337-54). Cambridge: Cambridge Uni- versity Press. Loayza, N., & Raddatz, C. (2010). The composition of growth matters for poverty alleviation. Journal of Development Economics, 93 (1), 137-151. Lyytikäinen, M., Jones, N., Huttly, S., & Abramsky, T. (2006). Childhood poverty, basic ser- vices and cumulative disadvantage: An international comparative analysis (Young Lives Working Paper 33). London: Young Lives. Mack, J., & Lansley, S. (1985). Poor Britain. London: George Allen & Unwin. Mahmud, S. (2003, June). Female secondary school stipend program in Bangladesh: A critical assessment (Report prepared for UNESCO Education for All [EFA] Global Monitoring Report 2003/4). Dhaka: Institute of Development Studies. Marvasti, A. B. (2004). Qualitative research in Sociology: An introduction. London: Sage Publi- cations. Matin, I., & Hulme, D. (2003). Programs for the poorest: Learning from the IGVGD program in Bangladesh. World Development, 31 (3), 647–65. 200 Mayhew, H. (1851). London labour and London poor. London: Griffin, Bohn, & Company. McKay, A., & Lawson, D. (2003). Assessing the extent and nature of chronic poverty in low in- come countries: issues and evidence. World Development, 31 (3): 425-39. Merton, R. K. (2004). Bureaucratic structure and personality. In J. M. Shafritz, A. C. Hyde, & S. J. Parkes (Eds.), Classics of public administration (5th ed., Part II, Chapter 13, pp. 109- 17.Belmont, CA: Wadsworth/Thomson Learning. (Original work published 1957). Minujin, A., Delamonica, E., Davidziuk, A., & Gonzalez, E. D. (2006). The definition of child poverty: A discussion of concepts and measurements. Environment and Urbanization, 18 (2), 481-500. Mirza, M. M. Q. (2004). Hydrological changes in Bangladesh. In M. M. Q. Mirza (Eds.), The Ganges water diversion: Environmental effects and implications (Chapter 2, pp. 13-31). Dordrecht, The Netherlands: Kluwer Academic Publishers. Montalvo, J. G, & Ravallion, M. (2010). The pattern of growth and poverty reduction in China. Journal of Comparative Economics, 38 (1), 2–16. Moore, K. (2001). Frameworks for understanding the intergenerational transmission of poverty and well-being in developing countries (Chronic Poverty Research Centre [CPRC] Work- ing paper 8). Manchester: CPRC. Retrieved June 10, 2010, from http://cprc.abrc.co.uk/ pdfs/08Moore.pdf Morgan, D. L. (1997). Focus groups as qualitative research (Rev. 2nd ed.). Thousand Oaks, California: Sage Publications. Morse, J. M. (2007). Sampling in grounded theory. In A. Bryant & K. Charmaz (Eds.), The Sage handbook of grounded theory (Part III, Chapter 11, pp. 229-44). London: Sage publications. Mullan, K., Grosjean, P., & Kontoleon, A. (2011). Land tenure arrangements and rural–urban migration in China. World Development, 39 (1), 123–133. Munshi, F. (2008). Trade liberralization and wage inequality: Empirical evidence from Bangla- desh. In F. Munshi, Essays on globalization and occupational wages (Essay 1, pp. 11-45). Doctoral dissertation. Department of Economics, Gothenburg: University of Gothenburg. Murgai, R., & Zaidi, S. (2005). Poverty trends in Bangladesh during the 1990s. Journal of Devel- oping Societies, 21 (1-2), 7-32. Mutebi, A. M. (2008). Explaining the failure of Thailand’s anti-corruption regime. Development and Change,39 (1), 147-71. Narayan, D., Patel, R., Schafft, K., Rademacher, A., & Koch-Schulte, S. (2000). Voices of the poor: Can anyone hear us. New York: Oxford University Press. Naschold, N. (2009). Microeconomic determinants of income inequality in rural Pakistan. Jour- nal of Development Studies, 45 (5), 746-768. Nilson, J. (2005). Understanding health-related quality of life in old age: A cross sectional study of elderly people in rural Bangladesh. Doctoral dissertation, Karolinska Institute, Swe- den. 201 Nolan, B., & Whelan, C. T. (2007). On the multidimensionality of poverty and social exclusion. In S. P. Jenkins, & J. Micklewright (Eds.), Inequality and poverty re-examined (Part II, Chapter 7, pp. 146-65). Oxford: Oxford University Press. Nolan, P. (1993).The causation and prevention of famines: A critique of A. K. Sen. Journalof Peasant studies, 21 (1), 1-28. Nunes, C. (2008). Poverty measurement: The development of different approaches and its techniques (ECINEQ, Society for the Study of the Economic Inequality Working Paper Series 93). Palma de Mallorca: Spain. OECD.(2008). Growing Unequal? Income Distribution and Poverty in OECD Countries. Paris: OECD. Retrieved August 10, 2010, from http://www.oecd.org/document/4/0,3343,en_2649_33933_41460917_1_1_1_1,00.html Olsen, J. P. (2004). Citizens, public administration and the search for theoretical foundations. Political Science & Politics, 37(1), 69-79. Osmani, S. (1995). The entitlement approach to famine: An assessment. In K. Basu, P. Pattanaik, & K. Suzumura (Eds.), Choice, welfare, and development: A festschrift in honour of Am- artya K. Sen (Part II, Chapter 12, pp. 253-94). Oxford: Oxford University Press. Osmani, S. R. (2006). The impact of globalization on poverty in Bangladesh (International La- bour Office [ILO] Working paper 65). Geneva: ILO. Parnini, S. N. (2006). Civil society and good governance in Bangladesh. Asian Journal of Politi- cal Science, 14 (2), 189-211. Patton, M. Q. (2002). Qualitative research and evaluation methods. London: Sage Publications. Petchey, R., Williams, J., & Carter, Y. H. (2008). From street-level bureaucrats to street-level policy entrepreneurs? Central policy and local action in lottery-funded community cancer care. Social Policy and Administration, 42 (1), 59-76. Pitt, M. M., Khandker, S. R., Chowdhury, O. H., & Millimet, D. L. (2003). Credit programs for the poor and the health status of children in rural Bangladesh. International Economic Review,44 (1), 87-118. Pradhan, M., & Ravallion, M. (2000). Measuring poverty using qualitative perceptions of consumption adequacy. Review of Economics and Statistics, 82 (3), 462-71. Quisumbing, A. R. (2007).Poverty transitions, shocks, and consumption in rural Bangladesh: Pre- liminary results from a longitudinal household survey (Chronic Poverty Research Centre [CPRC]Working Paper 105). Manchester: CPRC. Quisumbing, A. R. (2009). Investments, bequests, and public policy: Intergenerational transfers and the escape from poverty. In T. Addison, D. Hulme, & R. Kanbur (Eds.), Poverty dy- namics: Interdisciplinary perspectives (Part III, Chapter 14, pp. 267-88). Oxford: Oxford University Press. Rahman, A., & Razzaque, A. (2000). On reaching the hard core poor: Some evidence on social exclusion in NGO programs. The Bangladesh Development Studies, 26 (1), 1-35. Rahman, M. H., & Hossain, M. S. (2009). Convergence in per capita income across regions in 202 Bangladesh. The Bangladesh Development Studies, 32(1), 45–60. Rahman, R. I. (2004). Employment route to poverty reduction in Bangladesh: Role of self employ- ment and wage employment (ILO issues in employment and poverty Discussion Paper 17). Geneva: ILO. Ravallion, M. (1994).Poverty comparisons, fundamentals of Pure and AppliedEconomics (Vol. 56). Chur, Switzerland: Harwood Academic Publishers. Ravallion, M. (1998). Poverty lines in theory and practice (Living Standards Measurement Study[LSMS] Working Paper 133). Washington DC: World Bank. Ravallion, M. (2001). Growth, inequality and poverty: Looking beyond averages. World Develop- ment,29 (11) 1803-15. Ravallion, M. (2006). Transfers and safety nets in poor countries: Revisiting the trade-off and policy options. In A. V. Banerjee, R. Bénabou, & D. Mookherjee (Eds.), Understanding poverty (Part II, Chapter 14, pp. 203-30). Oxford: Oxford University Press. Ravallion, M. (2010). The debate on globalization, poverty, and inequality: Why measurement matters. In S. Anand; P. Segal, & J. E. Stiglitz (Eds.), Debates on the measurement of global poverty (Part 1, Chapter 2, pp. 25-41). Oxford: Oxford University Press. Ravallion, M. (2010a). How not to count the poor: A reply to Reddy and Pogge. In S. Anand; P. Segal, & J. E. Stiglitz (Eds.), Debates on the measurement of global poverty (Part I, Chapter 3a, pp. 86-101). Oxford: Oxford University Press. Ravallion, M., & Bidani, B. (1994). How robust is a poverty profile? The World Bank Economic Review, 8 (1), 75-102. Ravallion, M., Chen, S., & Sangraula, P. (2009). Dollar a day revised. The World Bank Economic Review, 23 (2),163-84. Ravallion, M., Datt, G., & Van de Walle, D. (1991). Quantifying absolute poverty in the develop- ing world. Review of Income and Wealth, 37(4), 345-61. Ravallion, M., & Wodon, Q. T. (2000). Does child labor displace schooling? Evidence on behav- ioral responses to an enrollment subsidy. The Economic Journal, 110 (462), C158-75. Reddy, S.G. (2009, December). The emperor’s new suit: Global poverty estimates reappraised. (Department of Economic and Social Affairs [DESA] Working Paper No. 79). New York: DESA, UnitedNations. Retrieved June 12, 2010, from http://www.un.org/esa/desa/pa- pers/2009/wp79_2009.pdf Reddy, S. G., & Pogge, T. W. (2002). How not to count the poor. Retrieved April 08, 2010, from www.socialanalysis.org. Reddy, S. G., & Pogge, T. (2005). How not to count the poor. Retrieved June 04, 2010, from http:// www.columbia.edu/~sr793/count.pdf Reddy, S. G., & Pogge, T. (2010). How not to count the poor. In S. Anand; P. Segal, & J. E. Sti- glitz (Eds.), Debates on the measurement of global poverty (Part I, Chapter 3, pp. 42-85). Oxford: Oxford University Press. Redmond, G. (2008). Child poverty and child rights: Edging towards a definition. Journal of Children and Poverty, 14 (1), 63-82. 203 Riskin, C. and Gao, Q. (2010). The changing nature of urban poverty in China. In S. Anand, P. Segal, and Stiglitz, J. (Eds.), Debates on the measurement of global poverty. Oxford: Oxford University Press. Rodas-Martini, P. (2001). Has income distribution really worsened in the South (Background Paper for the Human Development Report 2001)? Retrieved July 04, 2010, from http:// hdr.undp.org/en/reports/global/hdr2000/papers/rodasmartini.pdf Roelen, K., Gassmann, F., & Neubourg, C. (2009).The importance of choice and definition for the measurement of child poverty: The case of Vietnam. Child Indicators Research, 2 (3), 245-63. Rothstein, B., & Uslaner, E. M. (2005). All for all: Equality, corruption, and social trust. World Politics, 58 (1), 41-72. Rowntree, S. (1901). Poverty: A study of town life. London: Macmillan. Ruwanpura, K. N., & Roncolato, L. (2006). Child rights: An enabling or disabling right? The nexus between child laborand poverty in Bangladesh. Journal of Developing Societies, 22 (4), 359-78. Salmon, C. (2005). Child labor in Bangladesh: Are children the last economic resource of the household? Journal of Developing Societies, 21 (1-2), 33-54. Saunders, P., Bradshaw, J., & Hirst, M. (2002). Using household expenditure to develop an income poverty line. Social Policy & Administration,36 (3), 217-34. Scott, W. R. (2003). Organizations: rational, natural, and open systems. Upper Saddle River: Prentice Hall. Sen, A. (1976). Poverty: An ordinal approach to measurement. Econometrica, 44 (2), 219-31. Sen, A. (1981). Poverty and famines: An essay on entitlement and deprivation. Oxford: Clarendon Press. Sen, A. (1983). Poverty relatively speaking. Oxford Economic Papers, 35 (2), 153-69. Sen, A. (1985).A sociological approach to the measurement of poverty: A reply to Professor Peter Townsend. Oxford Economic Papers, 37 (4),669-76. Sen, A. (1987). Commodities and capabilities.New Delhi: Oxford University Press. Sen, A. (1992). Inequality Reexamined. Cambridge, MA: Harvard University Press. Sen, A. (1993).Capability and well-being. In M. Nussbaum & A. Sen (Eds.), The quality of life (Part I, pp. 30-53).Oxford: Clarendon Press. Sen, A. (1995).Food, economics, and entitlements. In J. Drèze, A. Sen, & A. Hussain (Eds.), The political economy of hunger: Selected essays - W I D E R Studies in Development Eco- nomics (Chapter 2, pp. 50-68). Oxford: Clarendon Press. Sen, A. (1999). Development as freedom.Oxford: Oxford University Press. Sen, B., & Ali, Z. (2009). Spatial inequality in social progress in Bangladesh. The Bangladesh Development Studies,32 (2), 53-78. Sen, B., & Hulme, D. (Eds.). (2006). The state of the poorest in Bangladesh: Chronic poverty in 204 Bangladesh - Tales of ascent, descent, marginality and persistence - The state of the poor- est 2004/2005. Dhaka and Manchester: BIDS and CPRC. Sheldin, M. G., & Shulman, L. (2004). Qualitative needs assessment of HIV services among Do- minican, Mexican and central American immigrant populations living in the New York city area. Aids Care,16 (4), 434-45. Siddique, M. A. B. (1998). Gender issues in poverty alleviation: A case study of Bangladesh. International Journal of Social Economics, 25 (6/7/8), 1095-1111. Sicular, T., Yue, X., Gustafsson, B., & Li, S. (2007). The urban-rural income gap and inequality in China. Review of Income and Wealth, 53 (1), 93-126. Silva, K. T., & Athukorala, K. (1996). South Asia: An overview. In E. Øyen, S. M. Miller, & S. A. Samad (Eds.), Poverty a global review: Handbook on international poverty research (Chapter 5, pp. 65-85). Oslo: Scandinavian University Press. Smith, A. (1776). An inquiry into the nature and causes of the wealth of nations. [Online edition © 1995-2005, Adam Smith Institute]. Retrieved, June 04, 2010, from http://www.adam- smith.org/smith/won/won-b5-c2-article-4-ss2.html Smith, B., & Moore, K. (2006). Intergenerational transmission of poverty in Sub-Saharan Africa (Chronic poverty research centre [CPRC] Working Paper 59). Manchester: Chronic pov- erty research centre. Sohlberg, P. (2006). Amartya Sen’s entitlement approach: Empirical statement or conceptual framework. International Journal of Social Welfare, 15 (4), 357-62. Stern, N. (2002, January 8). The investment climate, governance, and inclusion in Bangladesh (Speech delivered at Bangladesh Economic Association). Dhaka: Bangladesh Economic Association. Stewart, D. W., Shamdasani, P. N., & Rook, D. W. (2007). Focus groups theory and practice (2nd ed., Vol. 20).Newbury Park, CA: Sage Publications. Stiglitz, J. E. (2004). A willing world can end child poverty. In UNICEF, The state of the world’s children 2005: Childhood under threat ( Chapter 5, pp. 96-98). New York: United Na- tions Children’s Fund (UNICEF). Strauss, A. L. (1987). Qualitative analysis for social scientists. New York. Cambridge University Press. Sutherland, D., & Yao, S. (2011). Income inequality in China over 30 years of reforms. Cam- bridge Journal of Regions, Economy and Society, 4 (1), 91–105. Tanzeema, S., & Faisal, I. M. (2001). Sharing the Ganges: A critical analysis of the water sharing treaties. Water Policy,3 (1), 13-28. Temple, F. T., & Sattar, Z. (2000). Approaches to poverty reduction in Bangladesh. Dhaka: World Bank. Thurlow, J., & Wobst, P. (2006). Not all growth is equally good for the poor: The case of Zambia. Journal of African Economies,15 (4), 603-25. Tomlinson, M., Walker, R., & Williams, G. (2008). Measuring poverty in Britain as a multi- dimensional concept, 1991 to 2003. Journal of Social Policy, 37 (4), 597-620. 205 Townsend, P. (1979). Poverty in the United Kingdom. Harmondsworth: Penguin. Townsend, P. (1985).A sociological approach to the measurement of poverty: A rejoinder to pro- fessor Amartya Sen. Oxford Economic Papers,37 (4), 659-68. Turner, M., & Hulme, D. (1997). Governance, administration and development: Making the state work. London: Palgrave. UN. (2009). Rethinking poverty: Report on the world social situation 2010. New York: Depart- ment of Economic and Social Affairs, United Nations. UNDP. (2011). Project factsheet: Rural employment opportunities for public assets. Retrieved November 04, 2011, from http://www.undp.org.bd/projects/prodocs/REOPA/REOPA- CST%20factsheet%20Mar%202011%20%28extended%20project%29.pdf UNFPA. (2010). About Bangladesh. Retrieved August 4, 2010, from http://www.unfpa-bangla- desh.org/php/about_bangladesh.php UNICEF. (2004). The state of the world’s children 2005: Childhood under threat. New York: UNICEF. UNICEF. (2006). Innocenti social monitor 2006: Understanding child poverty in south eastern Europe and the commonwealth of independent states (UNICEF Innocenti Research Cen- tre Research report). Florence: UNICEF Innocenti Research Centre. UNICEF. (2007, January 10). UN General Assembly adopts powerful definition of child poverty (News notes). Retrieved June 12, 2007, from http://www.unicef.org/media/media_38003. html UNICEF. (2007a). Child poverty in perspective: An overview of child well-being in rich countries. Series. Innocenti Report Cards, 7. Retrived August 8, 2010, from http://www.unicef-icdc. org/publications/ Van Meter, D.S., & Van Horn, C. E. (1975). The policy implementation process: A conceptual framework. Administration and Society,6 (4), 445-87. Van Praag, B., Goedhart, T., & Kapteyn, A. (1980). The poverty line: A pilot survey in Europe. Review of Economics and Statistics, 62 (3), 461-65. Vleminckx, K., & Smeeding, T. (Eds.). (2001). Child well-being, child poverty and child policy in modern nations. Bristol: The Policy Press. WDI. (2007). World Development Indicators 2007. Washington, DC: The World Bank. Weiss, J. (2005). Experiences with poverty targeting in Asia: An overview. In J. Weiss (Ed.), Poverty targeting in Asia (Chapter 1, pp. 1-33). Cheltenham: Edward Elgar Publishing Limited. White, H., Leavy, J., & Masters, A. (2003). Comparative perspective on child poverty: A review of poverty measures. Journal of Human Development, 4 (3), 379-96. WHO. (2010). Part II, Table 7. World health statistics 2010. Geneva: World Health Organiza- tion (WHO) Press. Retrived July 17, 2010, from http://www.who.int/whosis/whostat/ EN_WHS10_Full.pdf Whyte, M. K. (2010). One country, two societies: Rural-urban inequality in contemporary China. Cambridge Massachusetts: Harvard University Press. 206 Wiener, C. (2007). Making teams work in conducting grounded theory. In A. Bryant & K. Charmaz (Eds.), The Sage handbook of grounded theory (Part IV, Chapter 14, pp. 293-310). Lon- don: Sage Publications. Winters, L.A., McCullock, N. & McKay, A. (2004). Trade liberalization and poverty: The evi- dence so far. Journal of Economic Literature, XLII, 72–115. Wilkinson, S. (2004). Focus group research. In D. Silverman (Ed.), Qualitative research: Theory, method and practice (2nd ed., Part IV, Chapter 10, pp. 177-99). London: Sage Publica- tions. Wint, E., & Frank, C. (2006). “Poor” to “not poor”: Improved understandings and the advantage of the qualitative approach. Journal of Sociology and Social Welfare, 33 (1), 163-77. Winters, L. A., McCulloch, N., & McKay, A. (2004). Trade liberalization and poverty: The evi- dence so far. Journal of Economic Literature, 42 (1), 72–115. World Bank. (2001). World development report 2000/2001: Attacking poverty. New York: Oxford University Press. World Bank. (2003). World Development Report 2003. Sustainable development in a dynamic world: Transforming institutions, growth and quality. New York: Oxford University Press. World Bank. (2005). Introduction to poverty analysis. World Bank. Retrieved July 28, 2010, from http://siteresources.worldbank.org/PGLP/Resources/PovertyManual.pdf World Bank. (2005a). Attaining the Millennium Development Goals in Bangladesh: How likely and what will it take to reduce poverty, child mortality and malnutrition, gender dispari- ties, and to increase school enrolment and completion? Washington DC. Human Devel- opment Unit, South Asia Region:The World Bank. World Bank. (2007). Bangladesh: Strategy for sustained growth (The World Bank Bangladesh development series. Document. Paper 18). Dhaka: The World Bank office. Retrieved Oc- tober 25, 2007, from www. worldbank.org.bd/bds World Bank. (2007a). China at a glance.Washington DC. The World Bank. Retrieved October 25, 2007, from www.worldbank.org/china World Bank. (2009). Bangladesh at a Glance. World Bank. Retrieved august 9, 2010, from http:// devdata.worldbank.org/AAG/bgd_aag.pdf World Bank. (2010). Table. Data profiles: Data for cross-country comparisons from the World De- velopment Indicators.World Bank. Retrieved august 9, 2010, from http://ddp-ext.world- bank.org/ext/ddpreports/ViewSharedReport?&CF=&REPORT_ID=9147&REQUEST_ TYPE=VIEWADVANCED World Bank.(2011a). World Development Indicators 2011. Washington D.C. World Bank. (2011b). World Development Report 2011. Washington D.C. Zafarullah, H. (2007). Bureaucratic elitism in Bangladesh: The predominance of generalist ad- ministrators. Asian Journal of Political Science, 15 (2), 161-73.