Bangladesh - Demographic and Health Survey - 2005

Publication date: 2005

Bangladesh Demographic and Health Survey 2004 National Institute of Population Research and Training (NIPORT) Dhaka, Bangladesh Mitra and Associates Dhaka, Bangladesh ORC Macro Calverton, Maryland USA May 2005 CONTRIBUTORS TO THE REPORT Ahmed Al-Sabir S.N. Mitra Shahidul Islam Subrata K. Bhadra Anne Cross Sushil Kumar Special acknowledgement: Dr. Kanta Jamil, Program Coordinator for Research, PHN Team, USAID, Dhaka for technical assistance at all steps of survey implementation, analysis, and report generation. This report summarizes the findings of 2004 Bangladesh Demographic and Health Surveys (2004 BDHS) conducted under the authority of the National Institute for Population Research and Training (NIPORT) of the Ministry of Health and Family Welfare and implemented by Mitra and Associates of Dhaka. ORC Macro provided financial and technical assistance for the survey through the financial aid provided by USAID/Bangladesh. The Bangladesh Demographic and Health Survey (BDHS) is part of the worldwide Demographic and Health Surveys program, which is designed to collect data on fertility, family planning, and maternal and child health. The opinions expressed in this report are those of the authors and do not necessarily reflect the views of USAID. Additional information about the 2004 BDHS may be obtained from: NIPORT Azimpur Dhaka, Bangladesh Telephone: 862-5251 Fax: 861-3362 Mitra and Associates 2/17 Iqbal Road, Block A Mohammadpur, Dhaka, Bangladesh Telephone: 911-5053 Fax: 912-6806 Additional information about the MEASURE DHS project may be obtained from: ORC Macro 11785 Beltsville Drive Suite 300 Calverton, MD 20705 USA Telephone: 301-572-0200 Fax: 301-572-0999 Internet: http://www.measuredhs.com Suggested citation: National Institute of Population Research and Training (NIPORT), Mitra and Associates, and ORC Macro. 2005. Bangladesh Demographic and Health Survey 2004. Dhaka, Bangladesh and Calverton, Maryland [USA]: National Institute of Population Research and Training, Mitra and Associates, and ORC Macro. Contents | iii CONTENTS Page Tables and Figures . ix Preface. xv Foreword . xvii Summary of Findings . xix Map of Bangladesh.xxvi CHAPTER 1 INTRODUCTION 1.1 GEOGRAPHY AND ECONOMY .1 1.2 POPULATION.2 1.3 POPULATION, FAMILY PLANNING AND MATERNAL AND CHILD HEALTH POLICIES AND PROGRAMS .2 1.4 ORGANIZATION OF THE 2004 BANGLADESH DEMOGRAPHIC AND HEALTH SURVEY .5 1.4.1 Survey Objectives and Implementing Organizations .5 1.4.2 Sample Design.5 1.4.3 Questionnaires .6 1.4.4 Training and Fieldwork .7 1.4.5 Data Processing .8 1.4.6 Coverage of the Sample.8 CHAPTER 2 HOUSEHOLD POPULATION AND HOUSING CHARACTERISTICS 2.1 Household Population by Age, Sex, and Residence .11 2.2 Household Composition .14 2.3 Educational Attainment of Household Members .14 2.3.1 School Attendance .17 2.4 Employment .18 2.5 Housing Characteristics.20 2.6 Household Possessions .22 2.7 Wealth Index.23 2.8 Arsenic in Household Drinking Water.24 CHAPTER 3 CHARACTERISTICS OF SURVEY RESPONDENTS 3.1 Background Characteristics of Respondents .29 3.2 Educational Attainment.32 3.3 Exposure To Mass Media .35 iv | Contents 3.4 Employment .37 3.4.1 Employment Status .37 3.4.2 Control Over Women’s Earnings.40 3.5 Women’s Empowerment .42 3.5.1 Women’s Participation in Decisionmaking: Women’s Perspective .42 3.5.2 Wife’s Participation in Decisionmaking: Husbands’ Perspective.45 3.5.3 Freedom of Movement.45 3.6 Men’s Attitudes Towards Wife-Beating .46 CHAPTER 4 FERTILITY 4.1 Introduction .49 4.2 Current Fertility Levels .50 4.3 Fertility Differentials.51 4.4 Fertility Trends.53 4.5 Children Ever Born and Living .55 4.6 Birth Intervals .57 4.7 Age at First Birth .59 4.8 Adolescent Fertility .60 CHAPTER 5 FERTILITY REGULATION 5.1 Knowledge of Family Planning Methods.63 5.2 Ever Use of Contraception .64 5.3 Knowledge and Ever Use of Menstrual Regulation .65 5.4 Current Use of Contraception.66 5.4.1 Trends in Current Use of Family Planning .67 5.4.2 Differentials in Current Use of Family Planning .69 5.5 Number of Children at First Use of Contraception .72 5.6 Problems with Current Method.72 5.7 Use of Social Marketing Brands.74 5.8 Age at Sterilization and Sterilization Regret .75 5.9 Source of Family Planning Services .78 5.10 Contraceptive Discontinuation.79 5.11 Future Intentions to Use Family Planning .82 5.11.1 Future Use of Contraception .82 5.11.2 Reasons for Not Intending to Use Contraception .82 5.11.3 Preferred Method for Future Use.83 5.12 Family Planning Outreach Services .84 5.13 Discussion about Family Planning between Spouses .87 5.14 Exposure To Family Planning Messages .87 Contents | v CHAPTER 6 OTHER PROXIMATE DETERMINANTS OF FERTILITY 6.1 Introduction .91 6.2 Marital Status.91 6.3 Age at First Marriage .93 6.4 Postpartum Amenorrhea, Abstinence, and Insusceptibility .96 6.5 Termination of Exposure to Pregnancy.99 CHAPTER 7 FERTILITY PREFERENCES 7.1 Desire for More Children. 101 7.2 Desire to Limit Childbearing . 103 7.3 Need for Family Planning Services . 105 7.4 Ideal Family Size. 107 7.5 Wanted and Unwanted Fertility. 111 CHAPTER 8 INFANT AND CHILD MORTALITY 8.1 Introduction . 115 8.2 Assessment of Data Quality. 115 8.3 Levels and Trends in Infant and Child Mortality . 117 8.4 Socioeconomic Differentials in Infant and Child Mortality. 118 8.5 Demographic Differentials in Infant and Child Mortality . 120 8.6 Perinatal Mortality . 121 8.7 High-Risk Fertility Behavior. 122 CHAPTER 9 CAUSES OF DEATH IN CHILDREN UNDER FIVE YEARS OF AGE 9.1 Introduction . 125 9.2 Description of the Data Collection Instrument . 126 9.3 Assigning Cause of Death. 126 9.4 Causes Of Death Among Children Under Five . 129 9.5 Differentials in Causes of Deaths among Children Under Five. 130 CHAPTER 10 MATERNAL AND CHILD HEALTH 10.1 Antenatal Care. 135 10.1.1 Antenatal Care Coverage . 135 10.1.2 Number and Timing of Antenatal Visits. 137 10.1.3 Health Services Received during Pregnancy. 138 10.1.4 Tetanus Toxoid Vaccinations . 140 10.2 Delivery Care . 141 10.2.1 Place of Delivery . 141 10.2.2 Assistance during Delivery . 143 10.3 Caesarean Section . 144 10.4 Postnatal Care . 145 vi | Contents 10.5 Complications during Pregnancy, during delivery or after delivery. 147 10.5.1 Knowledge of Life-Threatening Maternal Conditions. 147 10.5.2 Experience of Specific Maternal Complications around Delivery. 147 10.5.3 Treatment for Maternal Complications. 148 10.6 Childhood Vaccination . 150 10.6.1 Vaccination Coverage. 150 10.6.2 Differentials in Vaccination Coverage . 151 10.6.3 Trends in Vaccination Coverage. 151 10.7 Childhood Illness and Treatment . 154 10.7.1 Acute Respiratory Infection. 154 10.7.2 Childhood Diarrhea. 160 10.7.3 Treatment of Diarrhea . 161 CHAPTER 11 INFANT FEEDING AND NUTRITIONAL STATUS OF CHILDREN AND WOMEN 11.1 Breastfeeding and Supplementation. 165 11.1.1 Initiation of Breastfeeding . 165 11.1.2 Age Pattern of Breastfeeding. 167 11.1.3 Duration of Breastfeeding . 169 11.1.4 Complementary Feeding . 171 11.2 Micronutrient Intake. 173 11.2.1 Micronutrient Intake among Children . 173 11.2.2 Micronutrient Intake and Deficiencies among Women . 173 11.3 Nutritional Status of Children under Five . 176 11.3.1 Stunting . 177 11.3.2 Wasting . 181 11.3.3 Underweight . 182 11.3.4 Nutritional Status of Women . 182 CHAPTER 12 KNOWLEDGE, ATTITUDES AND BEHAVIOR RELATED TO HIV/AIDS AND OTHER SEXUALLY TRANSMITTED INFECTIONS 12.1 Knowledge of HIV/AIDS . 185 12.2 Knowledge of Ways to Avoid HIV/AIDS . 188 12.3 Knowledge of HIV/AIDS-Related Issues and Communication with Spouses . 193 12.4 Awareness, Prevalence, and Treatment of Sexually Transmitted Infections (STIs) . 196 Contents | vii CHAPTER 13 COMMUNITY CHARACTERISTICS. 203 CHAPTER 14 POLICY IMPLICATIONS OF THE 2004 BDHS 14.1 National Policy Environment. 209 14.2 Fertility . 210 14.3 Family Planning . 213 14.4 Childhood Health and Mortality . 215 14.5 Maternal Care. 218 14.5.1 Use of Antenatal Care. 218 14.5.2 Place of Delivery and Delivery Attendant. 218 14.5.3 Postnatal Care . 220 14.6 Knowledge of HIV/AIDS and Ways to Avoid AIDS . 220 14.7 Arsenic in Household Drinking Water. 221 14.8 Disparities . 222 REFERENCES . 223 APPENDIX A SAMPLE IMPLEMENTATION . 229 APPENDIX B ESTIMATES OF SAMPLING ERRORS . 233 APPENDIX C DATA QUALITY TABLES. 247 APPENDIX D BANGLADESH DEMOGRAPHIC AND HEALTH SURVEY PERSONNEL . 253 APPENDIX E QUESTIONNAIRES . 259 APPENDIX F SUMMARY INDICATORS.341 Tables and Figures | ix TABLES AND FIGURES Page CHAPTER 1 INTRODUCTION Table 1.1 Results of the household and individual interviews . 8 CHAPTER 2 HOUSEHOLD POPULATION AND HOUSING CHARACTERISTICS Table 2.1 Household population by age, sex, and residence . 12 Table 2.2 Population by age from selected sources. 13 Table 2.3 Household composition. 14 Table 2.4.1 Level of education of household population: women . 15 Table 2.4.2 Level of education of household population: men . 16 Table 2.5 School attendance . 17 Table 2.6 Employment status. 18 Table 2.7 Form of earnings. 19 Table 2.8 Household characteristics . 21 Table 2.9 Household durable goods and land ownership . 22 Table 2.10 Level of arsenic in household drinking water. 26 Table 2.11 Arsenic levels and knowledge of arsenic and markings on tubewells . 27 Figure 2.1 Population Pyramid, Bangladesh 2004. 12 Figure 2.2 Distribution of De Facto Household Population by Single Year of Age and Sex . 13 Figure 2.3 Trends in Percentage of Men and Women Age Six and Above With No Education by Sex and Residence. 17 Figure 2.4 Percentage of Household Respondents Who Know About Arsenic in Water, According to Background Characteristics . 25 CHAPTER 3 CHARACTERISTICS OF SURVEY RESPONDENTS Table 3.1.1 Background characteristics of respondents: women . 30 Table 3.1.2 Background characteristics of respondents: men. 31 Table 3.2.1 Level of education by background characteristics: women. 33 Table 3.2.2 Level of education by background characteristics: men. 34 Table 3.3.1 Exposure to mass media: women. 35 Table 3.3.2 Exposure to mass media: men . 36 Table 3.4.1 Employment status: women. 38 Table 3.4.2 Employment status: men . 39 Table 3.5 Decision on use of earnings . 40 Table 3.6 Decision on use of wife’s earnings . 41 Table 3.7 Women’s participation in household decisionmaking . 42 Table 3.8 Women’s participation in decisionmaking by background characteristics . 44 Table 3.9 Wife’s participation in decisionmaking. 45 Table 3.10 Freedom of movement . 46 Table 3.11 Men’s attitude towards spousal violence . 47 x | Tables and figures Figure 3.1 Differences in husband’s and wife’s ages (husband’s age minus wife’s age) . 32 Figure 3.2 Education of couples . 34 Figure 3.3 Percentage of ever-married women and all men exposed to various media at least once a week. 37 CHAPTER 4 FERTILITY Table 4.1 Current fertility rates . 50 Table 4.2 Fertility by background characteristics. 51 Table 4.3 Trends in current fertility rates . 53 Table 4.4 Percent pregnant . 54 Table 4.5 Trends in fertility by marital duration . 55 Table 4.6 Children ever born and living. 56 Table 4.7 Trends in children ever born. 57 Table 4.8 Birth intervals. 58 Table 4.9 Age at first birth . 59 Table 4.10 Median age at first birth . 60 Table 4.11 Teenage pregnancy and motherhood . 61 Figure 4.1 Age-Specific Fertility Rates by Residence. 52 Figure 4.2 Total Fertility Rates by Background Characteristics . 52 Figure 4.3 Trends in Total Fertility Rates . 54 CHAPTER 5 FERTILITY REGULATION Table 5.1 Knowledge of contraceptive methods . 63 Table 5.2 Ever use of contraception . 64 Table 5.3 Trends in ever use of family planning methods. 65 Table 5.4 Menstrual regulation. 66 Table 5.5 Current use of contraception . 66 Table 5.6 Trends in current use of contraceptive methods. 67 Table 5.7 Current use of contraception by background characteristics . 70 Table 5.8 Number of children at first use of contraception . 72 Table 5.9 Problems with current method of contraception . 73 Table 5.10 Use of pill brands. 74 Table 5.11 Use of condom brands . 75 Table 5.12 Timing of sterilization . 76 Table 5.13 Sterilization regret. 77 Table 5.14 Source of supply of modern contraceptive methods. 78 Table 5.15 Contraceptive discontinuation rates . 80 Table 5.16 Reasons for discontinuation . 81 Table 5.17 Future use of contraception . 82 Table 5.18 Reason for not intending to use contraception . 83 Table 5.19 Preferred method of contraception for future use . 84 Table 5.20 Contact with family planning fieldworkers and health fieldworkers . 85 Table 5.21 Satellite clinics . 86 Table 5.22 Discussion of family planning with husband. 87 Table 5.23 Exposure to family planning messages. 88 Figure 5.1 Trends in Contraceptive Use (%) Among Currently Married Women 10-49, Selected Surveys, 1975-2004 . 68 Figure 5.2 Trends in Contraceptive Method Mix Currently Married Women 10-49 Using a Method, Selected Surveys ,1991-2004. 69 Tables and Figures | xi Figure 5.3 Contraceptive Use and Women’s Status Indicators . 71 Figure 5.4 Distribution of Current Users of Modern Contraceptive Methods By Source of Supply. 79 CHAPTER 6 OTHER PROXIMATE DETERMINANTS OF FERTILITY Table 6.1 Current marital status. 92 Table 6.2 Trends in proportion never married . 93 Table 6.3.1 Age at first marriage: women . 93 Table 6.3.2 Age at first marriage: men. 94 Table 6.4.1 Median age at first marriage: women. 95 Table 6.4.2 Median age at first marriage: men. 96 Table 6.5 Postpartum amenorrhea, abstinence and insusceptibility . 97 Table 6.6 Median duration of postpartum insusceptibility by background characteristics . 98 Table 6.7 Menopause . 99 Figure 6.1 Trend in First Marriage of Women 20-24 by Age 18 . 94 CHAPTER 7 FERTILITY PREFERENCES Table 7.1 Fertility preferences by number of living children. 101 Table 7.2 Fertility preferences by age . 103 Table 7.3 Desire to limit childbearing. 104 Table 7.4 Need for family planning services . 106 Table 7.5.1 Ideal and actual number of children: women . 108 Table 7.5.2 Ideal and actual number of children: men . 109 Table 7.6 Mean ideal number of children by background characteristics . 110 Table 7.7 Fertility planning status . 111 Table 7.8 Wanted fertility rates . 113 Figure 7.1 Fertility Preferences Among Currently Married Women Age 10-49 .102 Figure 7.2 Percent of Currently Married Women Who Want No More Children by Number of Living Children.102 Figure 7.3 Percentage of Married Women Who Want No More Children by Number of Living Children and Background .105 Figure 7.4 Trend in Unmet Need for Family Planning by Division.107 Figure 7.5 Trend in Unplanned Births by Percent .112 CHAPTER 8 INFANT AND CHILD MORTALITY Table 8.1 Trend in early childhood mortality rates. 117 Table 8.2 Early childhood mortality rates by socioeconomic characteristics . 118 Table 8.3 Early childhood mortality rates by demographic characteristics . 120 Table 8.4 Perinatal mortality . 123 Table 8.5 High-risk fertility behavior. 124 Figure 8.1 Trends in Infant and Child Mortality . 117 Figure 8.2 Under-Five Mortality Rate by Socioeconomic Characteristics . 119 Figure 8.3 Under-Five Mortality Rate by Demographic Characteristics. 121 xii | Tables and figures CHAPTER 9 CAUSES OF DEATH IN CHILDREN UNDER FIVE YEARS OF AGE Table 9.1 Causes of death among children under five by age group . 130 Table 9.2 Causes of death among children under five by sex of child and residence . 131 Table 9.3 Causes of death among children under five by mother’s education. 132 Table 9.4 Causes of death among children under five by division . 133 Figure 9.1 Flow Chart Showing the Different Tiers Used in Assigning Cause of Death Based on Algorithms (ref) . 127 CHAPTER 10 MATERNAL AND CHILD HEALTH Table 10.1 Antenatal care . 136 Table 10.2 Number of antenatal care visits and timing of first visit . 138 Table 10.3 Health care during pregnancy. 139 Table 10.4 Tetanus toxoid injections . 140 Table 10.5 Place of delivery . 142 Table 10.6 Assistance during delivery . 143 Table 10.7 Delivery characteristics . 144 Table 10.8 Postnatal care for mother and children . 145 Table 10.9 Postnatal care by background characteristics. 146 Table 10.10 Knowledge of life threatening maternal conditions. 147 Table 10.11 Experience of complications around delivery . 148 Table 10.12 Treatment seeking for maternal complications . 149 Table 10.13 Vaccinations by source of information . 150 Table 10.14 Vaccinations by background characteristics. 152 Table 10.15 Prevalence and treatment of acute respiratory infection (ARI) . 155 Table 10.16 Prevalence of fever . 157 Table 10.17 Prevalence of diarrhea . 158 Table 10.18 Treatment of diarrhea . 159 Table 10.19 Feeding practices during diarrhea . 161 Table 10.20 Treatment of diarrhea . 162 Table 10.21 Feeding practices during diarrhea . 163 Figure 10.1 Reasons for Not Seeing Anyone for Antenatal Care . 137 Figure 10.2 Vaccinations by Background Characteristics . 153 Figure 10.3 Trends in Vaccination Coverage among Children Age 12-23 Months . 153 CHAPTER 11 INFANT FEEDING AND NUTRITIONAL STATUS OF CHILDREN AND WOMEN Table 11.1 Initial breastfeeding . 166 Table 11.2 Breastfeeding status by child’s age . 167 Table 11.3 Median duration of breastfeeding . 170 Table 11.4 Foods consumed by children in the day preceding the interview . 172 Table 11.5 Micronutrient intake among children . 174 Table 11.6 Micronutrient intake and deficiency among mothers. 175 Table 11.7 Nutritional status of children by demographic characteristics . 179 Table 11.8 Nutritional status of children by background characteristics. 180 Table 11.9 Nutritional status of women by background characteristics. 184 Tables and Figures | xiii Figure 11.1 Infant Feeding Practices by age . 168 Figure 11.2 Trends in Exclusive Breastfeeding for Children Under Six Months . 169 Figure 11.3 Trends in Complementary Feeding for Children 6-9 Months. 169 Figure 11.4 Median Duration (Months) of Breastfeeding. 171 Figure 11.5 Percentage of Children Under Five Who Are Stunted, According to Demographic Characteristics . 178 Figure 11.6 Percentage of Children Under Five Who Are Stunted According to Socioeconomic Characteristics. 178 Figure 11.7 Trends in Nutritional Status of Children Under Five . 181 Figure 11.8 Trends in the Nutritional Status of Women with Children under Five Years of Age. 183 CHAPTER 12 KNOWLEDGE, ATTITUDES, AND BEHAVIOR RELATED TO HIV/AIDS AND OTHER SEXUALLY TRANSMITTED INFECTIONS Table 12.1.1 Knowledge of HIV/AIDS and sources of AIDS information: women. 186 Table 12.1.2 Knowledge of HIV/AIDS and sources of AIDS information: men . 187 Table 12.2.1 Knowledge of ways to avoid HIV/AIDS: women. 189 Table 12.2.2 Knowledge of ways to avoid HIV/AIDS: men . 190 Table 12.3.1 Knowledge of HIV/AIDS and its prevention: women. 191 Table 12.3.2 Knowledge of HIV/AIDS and its prevention: men. 192 Table 12.4 Perceptions of HIV/AIDS-related issues . 194 Table 12.5 Discussion of HIV/AIDS with spouse . 195 Table 12.6.1 Knowledge of signs and symptoms of STIs : women. 196 Table 12.6.2 Knowledge of signs and symptoms of STIs : men. 197 Table 12.7 Gynecological health problems . 198 Table 12.8 Women seeking treatment for gynecological health problems . 199 Table 12.9 Self-reportiing of sexually-transmitted infections (STIs) and STI symptoms . 200 Table 12.10 Men seeking treatment for STIs. 201 Figure 12.1 Trends in Knowledge of HIV/AIDS Among Ever-Married Women and Currently Married Men. 188 Figure 12.2 Percentage of Ever-married Women and All Men Who have Heard of HIV/AIDS, by Background Characteristics . 188 Figure 12.3 Trends in Knowledge of Two or More Correct Ways to Avoid HIV/AIDS Among Ever-Married Women and Currently Married Men. 193 CHAPTER 13 COMMUNITY CHARACTERISTICS Table 13.1 Distance to nearest general services . 204 Table 13.2 Distance to nearest education facilities . 205 Table 13.3 Availability of income-generating organizations. 205 Table 13.4 Availability of family planning and health services . 206 Table 13.5 Distance to nearest health and family planning services . 207 CHAPTER 14 POLICY IMPLICATIONS OF THE 2004 BDHS Table 14.1 Percent reduction in mortality per year, among children under five .215 Figure 14.1 Percentage of Women Age 15-19 Ever Married .211 Figure 14.2 Percentage of Females Age 15-19 with No Education and Percentage with at Least Some Secondary Education .212 xiv | Tables and figures Figure 14.3 Median Age at Marriage and Median Age at First Birth among women age 20-24 .213 APPENDIX A SAMPLE IMPLEMENTATION Table A.1 Sample implementation: women . 229 Table A.2 Sample implementation: men. 230 Figure A.1 Urban Sampling Points . 231 Figure A.2 Rural Sampling Points . 232 APPENDIX B ESTIMATES OF SAMPLING ERRORS Table B.1 List of selected variables for sampling errors . 236 Table B.2 Sampling errors: Total sample . 237 Table B.3 Sampling errors: Urban sample. 238 Table B.4 Sampling errors: Rural sample. 239 Table B.5 Sampling errors: Barisal sample. 240 Table B.6 Sampling errors: Chittagong sample . 241 Table B.7 Sampling errors: Dhaka sample. 242 Table B.8 Sampling errors: Khulna sample. 243 Table B.9 Sampling errors: Rajshahi sample. 244 Table B.10 Sampling errors: Sylhet sample . 245 APPENDIX C DATA QUALITY TABLES Table C.1 Household age distribution. 247 Table C.2.1 Age distribution of eligible and interviewed: women. 248 Table C.2.2 Age distribution of eligible and interviewed: men . 248 Table C.3 Completeness of reporting . 249 Table C.4 Births by calendar years . 249 Table C.5 Reporting of age at death in days . 250 Table C.6 Reporting of age at death in months . 251 Preface | xv Director General National Institute of Population Research and Training (NIPORT) PREFACE The Bangladesh Demographic and Health Survey 2004 is the fourth survey of this type conducted in Bangladesh. The main objective of this survey is to provide policy-makers and program managers in health and family planning with detailed information on fertility and family planning, childhood mortal- ity, maternal and child health, nutritional status of children and mothers, and awareness of HIV/AIDS. The survey consisted of two parts: a household-level survey of women and men and a community survey around the sample points from which the households were selected. Preparations for the survey started in mid-2003 and the fieldwork was carried out between January and May 2004. Financial support for the BDHS survey was provided by the United States Agency for International Development (USAID)/Dhaka. It was implemented through a collaborative effort of NIPORT, Mitra and Associates, and ORC Macro. The findings of this report will be instrumental in assessing the achievements of family planning, nutrition, and health programs. The report provides estimates of key indicators by socioeconomic and demographic differentials. The preliminary results of the 2004 BDHS, with its major findings, were offi- cially announced through a national seminar in September 2004. The final report supplements the pre- liminary report, which was released earlier. I believe that the information obtained from this survey will help the policymakers and program managers in the formulation of new programs and monitoring the on- going programs. The Technical Review Committee (TRC) consisted of experts from government, non- government, and international organizations, as well as researchers and professionals working in the health and population areas. The TRC contributed their valuable opinions in major phases of the survey. In addition, the Technical Task Force (TTF) was formed with representatives from NIPORT, Mitra and Associates, USAID/Dhaka, ICDDR,B, the NGO Service Delivery Program, and ORC Macro to design and implement the survey. I would like to extend my thanks and appreciation to the members of the TRC and TTF for their contributions at different phases of the survey. I express my sincere thanks to the professionals of the Research Unit of NIPORT, ORC Macro, and Mitra and Associates for their sincere efforts in timely completion of the survey. USAID/Dhaka deserves special thanks for providing financial support for the survey. (Lokman Hakim) Preface | xvii Secretary Ministry of Health and Family Welfare Government of the People’s Republic of Bangladesh FOREWORD The Bangladesh Demographic and Health Survey (BDHS) is a nationally representative survey designed to obtain and provide information on the basic indicators of social progress including fertility, childhood mortality, reproductive and child health, nutritional status of mothers and children and aware- ness of HIV/AIDS. Previously, BDHS surveys were carried out in 1993-1994, 1996-1997, and 1999- 2000. The findings of the 2004 BDHS presented in this report provide up-to-date, and reliable informa- tion on a number of key health and demographic topics of interest to planners, policymakers, program managers, and researchers that will guide the planning, implementation, monitoring and evaluation of the Health, Nutrition and Population Sector Program (HNPSP) in Bangladesh. The data indicate there has been a decline in the total fertility rate and a steady increase in contraceptive use. After an almost decade- long stagnation, fertility declined to 3.0 children per woman in 2004. The 2004 BDHS findings also show a trend toward increasing utilization of health services for mothers and children. While the survey results are encouraging, there is still a long way to go to achieve the national health and demographic goals. The findings of this report together with other national surveys will enhance the understanding of important issues related to the HNPSP in Bangladesh. Information obtained from the 2004 BDHS can be used to review the progress of programs and to improve future policies and strategies. Further analysis of the BDHS data is necessary. It is hoped that academicians, researchers and program personnel will carry out such analysis and provide in-depth knowledge to guide the future direc- tion and effective implementation of the HNPSP. The successful completion of the 2004 BDHS was made possible by the contributions of a num- ber of organizations and individuals. I deeply appreciate the United States Agency for International De- velopment (USAID), Dhaka for providing financial support. I would like to thank NIPORT, Mitra and Associates, and ORC Macro for the effort they put into implementing the 2004 BDHS. (A. F. M. Sarwar Kamal) Summary of Findings | xix SUMMARY OF FINDINGS The 2004 Bangladesh Demographic and Health Survey (2004 BDHS) is a nationally repre- sentative survey of 11,440 women age 10-49 and 4,297 men age 15-54 from 10,500 households covering 361 sample points (clusters) throughout Bangladesh, 122 in urban areas and 239 in the rural areas. This survey is the fourth in a series of national-level population and health surveys con- ducted as part of the global Demographic and Health Surveys (DHS) program. It is designed to provide data to monitor the population and health situation in Bangladesh as a followup to the 1993- 1994, 1996-1997 and 1999-2000 BDHS surveys. The survey utilized a multistage cluster sample based on the 2001 Bangladesh Census and was designed to produce separate estimates for key indicators for each of the six divisions of the country—Barisal, Chittagong, Dhaka, Khulna, Rajshahi and Sylhet. Data collection took place over a five-month period from 1 January to 25 May 2004. Previous surveys included only ever- married women and currently married men; this is first DHS survey in Bangladesh to also include never-married and formerly married men, i.e., the sample for the survey was ever-married women age 10-49 and all men age 15-54. The survey obtained detailed information on fertility levels, marriage, fertility preferences, awareness and use of family planning methods, breastfeeding practices, nutritional status of women and young children, childhood mortality and causes of death of children under five, mater- nal and child health, awareness and behavior re- garding HIV/AIDS, and other sexually transmitted infections (STIs). In the previous surveys, anthro- pometric measurements (height and weight) were restricted to mothers who had a child under five years, and their young children. In the 2004 BDHS, all children under five in the household and all interviewed women had their height and weight measured. In addition, the 2004 BDHS collected information on the level of arsenic in drinking water. The 2004 BDHS was conducted under the authority of the National Institute for Population Research and Training (NIPORT) of the Ministry of Health and Family Welfare. It was implemented by Mitra and Associates, a Bangladeshi research firm located in Dhaka. Technical assistance was provided by ORC Macro through the MEASURE DHS pro- gram. Financial support for the survey was provided by the U.S. Agency for International Development (USAID)/Bangladesh. FERTILITY Fertility Levels and Trends. In 1971-1975, women in Bangladesh were having on average 6.3 children. The total fertility rate (TFR) declined to 5.1 fifteen years later, and to 4.3 in 1989-1991. The TFR plateaued at around 3.3 for most of the 1990s, when the three earlier BDHS surveys took place. Data from the 2004 BDHS indicate that after almost a decade- long stagnation, the Bangladesh fertility rate has de- clined slightly to 3.0 children per woman. Comparison of the Bangladesh TFR with fertility rates in other Asian countries that have implemented a DHS survey indicates that, with a TFR of 3.0, Bangladesh is in the mid-range among the countries—below Nepal (4.1 in 2001), Cambodia (3.8 in 2000), and the Philippines (3.5 in 2003), but above India (2.8 in 1998-1999), In- donesia (2.6 in 2002-2003), and Vietnam (1.9 in 2002). Fertility Differentials. Differentials in fertility by background characteristics are substantial. Women in rural areas have more children than their urban counterparts (3.2 and 2.5 children per woman, respec- tively). The TFR is highest in Sylhet division (4.2) and lowest in Rajshahi (2.6). As expected, women’s education is strongly associated with lower levels of fertility; the TFR decreases from 3.6 among women with no education to 2.2 among those who have at least completed their secondary education. Similar differentials are observed by wealth quintile, with the TFR decreasing from 4.0 among women in the lowest wealth quintile to 2.5 among those in the highest wealth quintile. Unplanned Fertility. Despite a steady rise in the level of contraceptive use over the past thirty years, the 2004 BDHS data indicate that unplanned pregnan- cies are common in Bangladesh. Overall, 3 out of 10 births in Bangladesh are either unwanted (14 percent) xx | Summary of Findings or mistimed and wanted later (16 percent). How- ever, the proportion of unplanned births declined from 33 percent in 1999-2000 to 30 percent in 2004. The proportion of unwanted births did not change. Fertility Preferences. There is considerable desire among currently married Bangladeshi women to stop having children. A total of 54 per- cent of women age 10-49 reported not wanting another child, and 6 percent are already sterilized. Twenty-one percent of women want to have a child but would prefer to wait two or more years. Thus, over 80 percent of women want either to space their next birth or to limit childbearing alto- gether. Only 13 percent of women would like to have a child soon (within two years). A compari- son of the 1999-200 and 2004 data shows that the proportion of women who want to limit childbear- ing has not changed. As in the 1999-2000 BDHS, a majority of ever-married women and currently married men embrace the two-child family as an ideal (2.4 and 2.3 children, respectively). FAMILY PLANNING Knowledge of Contraception. Knowledge of family planning is universal in Bangladesh. Among ever-married women, the most widely known methods of family planning are the pill (100 percent), injectables (99 percent), female sterilization (96 percent), and condom (92 per- cent); these are followed by the IUD (85 percent), Norplant (76 percent), male sterilization (73 per- cent), periodic abstinence (70 percent), and with- drawal (58 percent). Since overall knowledge of contraceptive methods was already high in 1999-2000, little change has taken place. However, knowledge of Norplant has increased from 56 to 77 percent among currently married women. Use of Contraception. The contraceptive prevalence rate (any method) among currently married women is 58 percent. The most com- monly used modern method is the pill (26 per- cent), followed by injectables (10 percent). Fe- male sterilization and male condoms are used by 5 percent and 4 percent of married women, respectively, while Norplant, the IUD, and male sterilization are each used by only 1 percent. Periodic abstinence, used by 7 percent of married women, is the most com- monly used traditional method. Trends in Contraceptive Use. Over the past three decades, use of any method of contraception by married women has increased sevenfold, from 8 to 58 percent, while use of modern methods has increased almost tenfold, from 5 to 47 percent. The same trend was observed between the 1999-2000 BDHS and the 2004 BDHS, when use of any method increased from 54 to 58 percent and use of modern methods increased from 43 to 47 percent. Trends in the contraceptive method mix show that short-term methods, especially the pill, are gaining in popularity against long-term methods, such as the IUD, Norplant, and sterilization. The pill now accounts for 45 percent of all contracep- tive use, compared with 35 percent in 1991. On the other hand, long-term methods now account for only 12 percent of all contraceptive use, compared with 30 percent in 1991. Differentials in Contraceptive Use. Women in urban areas are slightly more likely to use contracep- tive methods (63 percent) than their rural counterparts (57 percent); however, the condom is the only method that shows differentials in use by urban-rural resi- dence: 8 percent in urban areas compared with only 3 percent in rural areas. Differentials are more marked by division: use of any method varies from 32 percent in Sylhet and 47 percent in Chittagong to 64 percent in Khulna and 68 percent in Rajshahi. Contraceptive prevalence is 54 percent in Barisal and 59 percent in Dhaka. There is little variation in contraceptive use by level of education. However, women in economically better-off households tend to use family planning more than those in households in the lowest wealth quintile (63 and 54 percent, respectively). The propor- tion of women using contraception increases with in- creasing number of children. Twenty-three percent of women with no children are currently using a contra- ceptive method, compared with 62 to 70 percent of women with two or more children. Source of Modern Methods. In Bangladesh, both the public and private sectors are important sources of supply for users of modern methods (57 and 36 percent, respectively). The most common pub- lic sector source remains government fieldworkers (23 Summary of Findings | xxi percent), although their share has declined sub- stantially since 1993-1994 (42 percent). Upazila health complexes are the second most important public source (10 percent). Pharmacies (29 per- cent) provide most of the methods in the private sector (an increase from 21 percent in 1999-2000). Femicon, the most commonly used social market- ing brand of pills, is distributed through a network of retail outlets including pharmacies. Of every ten pills used in Bangladesh, three carry the Femi- con brand. Contraceptive Discontinuation. One in two contraceptive users in Bangladesh stops using their method within 12 months of starting. The most common reason for discontinuation is side effects or health problems. Discontinuation rates are highest for condoms (72 percent) and with- drawal (60 percent), and lowest for periodic absti- nence (41 percent). Unmet Need for Family Planning. Eleven percent of married women have an unmet need for family planning. Unmet need is about equally di- vided between spacing and limiting births. Unmet need declined from 15 percent in 1999-2000 to 11 percent in 2004. It has remained high in Sylhet division (21 percent), while dropping substantially in Rajshahi (7 percent) and Khulna (8 percent). Overall, 84 percent of the demand for family planning is currently being met. MATERNAL HEALTH Antenatal Care. Antenatal care coverage in- creased sharply between the 1999-2000 BDHS and the 2004 BDHS. One-third of women re- ceived an antenatal checkup from a medically trained provider in 1999-2000 compared with one- half (49 percent) in 2004. Thirty-one percent of women received antenatal care from a doctor and 17 percent received care from a nurse, midwife, or paramedic. A relatively high proportion of women received no antenatal care (44 percent), especially in Sylhet (52 percent) and Barisal (53 percent). Two in three women received at least two doses of tetanus toxoid for their most recent birth in the five years preceding the survey, 21 percent received only one tetanus toxoid injection, and 15 percent received none, which was an improve- ment since the 1999-2000 BDHS (19 percent). Delivery Care. Nationally, nine in ten births in the last five years were delivered at home; only 9 per- cent were delivered in a health facility. Delivery in a health facility is substantially higher among women who have at least completed their secondary education (44 percent), and among those in the highest wealth quintile (30 percent). The data also show that only 13 percent of babies were delivered by medically trained providers, compared with 63 percent who were deliv- ered by untrained birth assistants. Postnatal Care. Only 15 percent of women who had a non-institutional live birth in the five years pre- ceding the survey received postnatal care within two days of delivery; more than 80 percent received no postnatal care at all. Maternal Complications around Delivery. One in four births in the five years preceding the survey had at least one of the following maternal complica- tions around delivery—prolonged labor, excessive bleeding, baby’s hands or feet came first, fever with foul-smelling discharge, convulsions/eclampsia. The most common complication was prolonged labor of over 12 hours, associated with one in six live births. For 11 percent of the births, the mothers experienced excessive bleeding, and 3 percent had convulsions. Two other problems, high fever with foul discharge and baby’s hands or feet coming first, were reported for 5 and 1 percent of births, respectively. Treatment was sought from a medically trained provider for only 29 percent of the cases that had maternal complications around delivery. Nearly four in ten women with complications did not seek any care. The 2004 BDHS data confirm the findings of the 2001 BMMS, that there are two main problems regarding the treatment of maternal complications: first, a large proportion of women with potentially life-threatening maternal complications seek no health care; and second, among those who do seek health care, about half seek assistance from providers that are not medically trained. CHILD HEALTH Childhood Mortality. Data from the 2004 BDHS show that under-five mortality (88 deaths per 1,000 live births) has continued to decline thanks pri- marily to the substantial decline (20 percent) in child mortality (age 1-4 years) over the past five years. However, this still means that for the most recent five- xxii | Summary of Findings year period, one in every eleven Bangladeshi chil- dren dies before reaching age five, while one in fifteen children dies before reaching the first birthday (65 deaths per 1,000 live births). A ma- jority of infant deaths occur during the first month of life (neonatal mortality). The 2004 BDHS also collected information on causes of death. Overall, for all children under five, the two most important causes of death were: possible serious infections (31 percent) including possible ARI and diarrhea and ARI (21 percent), which particularly affect children age 1-11 months. Birth asphyxia (12 per- cent), which occurs in the first 28 days, diarrhea (7 percent), and prematurity/low birth weight (7 percent) were responsible for most of the other deaths. Childhood Vaccination Coverage. Seventy-three percent of Bangladeshi children age 12-23 months are fully immunised—most of them by 12 months of age as recommended—while 3 percent have received no vaccinations. More than nine in ten children have received BCG and the first dose of DPT and polio vaccines. While coverage for the first dose of DPT and polio is high, there is a de- cline with subsequent doses, with only about 81 percent of children receiving the recommended three doses of these vaccines. Seventy-six percent of children have received measles vaccine. Full vaccination coverage is highest in Khulna division (83 percent) and lowest in Sylhet division (62 per- cent). Mother’s education is strongly associated with children’s vaccination coverage: only 60 per- cent of children of mothers with no education are fully vaccinated compared with 92 percent of chil- dren of highly educated mothers. Child Illness and Treatment. Among chil- dren under five years of age, 21 percent were re- ported to have had symptoms of acute respiratory illness in the two weeks preceding the survey. Of these, only one-fifth were taken to a health facility or provider for treatment, and one-third received no treatment at all. Eight percent of children under five years had diarrhea in the two weeks preced- ing the survey. Of these, 16 percent were taken to a health provider. Use of oral rehydration therapy (ORT) for children with diarrhea has remained unchanged since 1999-2000, but there has been a shift toward greater use of the commercially available packets of oral rehydration salts (ORS), from 61 to 67 percent. Overall, 83 percent of the children with diarrhea received ORS, recommended home fluids (RHF), or increased fluids. Forty percent of children under five years had a fever in the two weeks preceding the survey. Of these, nearly two-thirds were taken to a provider for treat- ment, but only 19 percent were taken to a medically trained provider/facility. NUTRITION Breastfeeding Practices. Almost all (98 percent) Bangladeshi children are breastfed for some period of time. Twenty-four percent of infants were put to the breast within one hour of birth, and 83 percent started breastfeeding within the first day. This is a substantial increase when compared to the 1999-2000 BDHS data. The median duration of any breastfeeding in Bangladesh is 32 months, but it varies among divi- sions from 36 months in Khulna and Rajshahi to around 26 months in Chittagong and Sylhet. Exclusive breastfeeding of children under six months (based on 24-hour period before the survey) has not improved in the past 10 years; it remained un- changed at around 45 percent between 1993-94 and 1999-2000, and has declined to 42 percent most re- cently. Supplementary feeding of children who are also breastfed has greatly improved over the past decade. In 1993-1994, only 29 percent of children age 6-9 months received complementary foods while being breastfed, compared with 62 percent in 2004. The most commonly used complementary foods are rice, wheat, and porridge (over 60 percent); 20 to 25 per- cent of the children in this age group received other complementary foods (fruits, meat/fish/eggs, and green leafy vegetables), and a smaller proportion re- ceived dal. Feeding children with a bottle with a nipple starts very young, and three in ten infants age 2-3 months receive some food this way. Also, commercially pro- duced baby formula is more popular than it was at the time of the 1999-2000 BDHS. Intake of Vitamin A. Ensuring that children 6- 59 months receive enough vitamin A may be the sin- gle most effective child survival intervention because deficiencies in this micronutrient can cause blindness and increase the severity of infections such as measles Summary of Findings | xxiii and diarrhea. Between the 1999-2000 BDHS and the 2004 BDHS, vitamin A supplementation among children 12-59 months increased from 80 to 84 percent, but it actually dropped by half for children age 9-11 months (from 73 to 38 percent). Consumption of fruits and vegetables rich in vi- tamin A is another way to ensure that children are protected from blindness or infection. Overall, 7 in 10 children under three consumed such foods. Only 15 percent of mothers with a birth in the past five years reported receiving a vitamin A dose postpartum. Three percent of interviewed women reported night blindness during preg- nancy. Nutritional Status of Children. According to the 2004 BDHS which measured all children under five in the household, 43 percent of children are stunted and 17 percent severely stunted. Thir- teen percent of children under five are wasted and 1 percent severely wasted. Weight-for-age results show that 48 percent of children under five are underweight, with 13 percent severely under- weight. Comparison of children whose mothers were interviewed shows that in spite of the fact that child nutritional levels showed a substantial improvement from 1996-1997 to 1999-2000, since then no noticeable improvement has occurred ex- cept that the severe stunting has slightly decreased and overall wasting has increased from 10 to 13 percent. Nutritional Status of Women. The mean height of Bangladeshi women is 151 centimetres, which is above the critical height of 145 centime- tres. A high proportion of women (16 percent) are below 145 centimetres. Thirty-four percent of women were found to be chronically malnour- ished, their body mass index (BMI) being less than 18.5. One in ten women was found to be overweight or obese (BMI 25 or higher). A woman’s place of residence, level of education, and household wealth quintile are strongly associ- ated with her nutritional status. For example, 37 percent of rural women are considered thin (<18.5), compared with 25 percent of their urban counterparts. Among divisions, Sylhet has the highest proportion of women who are thin (48 percent) and Khulna the least (29 percent). Al- though Bangladeshi women with children under five years are not getting taller, there is a substan- tial improvement in mother’s nutritional status as measured by BMI. Since 1996-97, the proportion of mothers below the cutoff point of BMI of 18.5 contin- ued to drop, from 52 percent in 1996-97 to 38 percent in 2004—a decline of 27 percent in less than ten years. Arsenic in Drinking Water. Arsenic in drinking water is a hazard to human health. Its main source is arsenic-rich rocks through which the water has fil- tered. It may also occur because of mining or indus- trial activity. In Bangladesh, arsenic-contaminated water is found particularly in tubewells. Overall, in the 2004 BDHS, one in twelve households were found to have elevated levels of arsenic (equal to or greater than 50 parts per billion) in their drinking water. The problem is especially severe in Chittagong, where 22 percent of the households tested had arsenic- contaminated water; arsenic contamination is almost nonexistent in Barisal and Rajshahi (1 and 2 percent). HIV/AIDS AND STIS Awareness of HIV/AIDS. Knowledge of HIV/ AIDS among ever-married women increased from 19 percent in 1996-1997 to 31 percent in 1999-2000, and then it almost doubled to 60 percent in 2004. For currently married men, the corresponding proportions are 34, 51, and 78 percent. A respondent’s place of residence, level of edu- cation, and household wealth quintile are strongly as- sociated with HIV/AIDS awareness. Whereas 82 per- cent of women and 93 percent of men in urban areas have heard of AIDS, only 54 percent of women and 78 percent of men in rural areas have heard of the dis- ease. Education is positively associated with knowl- edge of HIV/AIDS. It ranges from 37 percent among women with no education, to 71 percent among those who have completed primary school (only), to virtu- ally all women (98 percent) who have completed sec- ondary education. A similar pattern can be found when analyzing the data by wealth quintile. Thirty-seven percent of ever-married women, 57 percent of never-married men, and 45 percent of cur- rently married men know that condom use is a way to avoid contracting HIV/AIDS, a clear improvement over the results of the 1999-2000 BDHS. About one in three married women and one in eight among all men or currently married men know that limiting the num- ber of sexual partners can prevent HIV/AIDS. Overall, six in ten women and 42 percent of men do not know any way to avoid the disease. xxiv | Summary of Findings Among respondents who know of HIV/ AIDS, seven in ten women and 84 percent of men correctly reported that a healthy looking person can have the AIDS virus. In 2004, 29 percent of ever-married women were able to cite two or more correct ways to avoid contracting HIV/AIDS. Since 1999-2000, the unprompted knowledge of at least two correct ways to avoid HIV/AIDS has increased substan- tially among ever-married women (from 7 to 29 percent) and moderately among currently married men (from 19 to 26 percent). Awareness of Sexually Transmitted Infections (STIs). Knowledge of STIs is generally lower than that of HIV/AIDS. Ninety-four percent of women and 78 percent of married men still do not have any knowledge of STIs. Knowledge of STIs is highest among women and men who have completed secon- dary education, 19 and 38 percent, respectively. Map of Bangladesh | xxv Introduction │ 1 INTRODUCTION 1 1.1 GEOGRAPHY AND ECONOMY Bangladesh is located in the northeastern part of South Asia and covers an area of 147,570 square kilometers. It is almost entirely surrounded by India, except for a short southeastern frontier with Myanmar and a southern coastline on the Bay of Bengal. It lies between latitudes 20° 34′ and 26° 38′ north and longitudes 88° 01′ and 92° 41′ east, and has a tropical climate. The Moguls ruled the country from the 13th century till the 18th century, when the British took over and administered the subcontinent until 1947. During British rule, Bangladesh was a part of India. In 1947, the independent states of Pakistan and India were created with the present Bangladesh territory as a part of Pakistan. Bangladesh emerged on the world map as a sovereign state on March 26, 1971 after fighting a 9-month war of liberation. Most of Bangladesh consists of low, flat and alluvial soil. The most significant feature of the landscape is the extensive network of large and small rivers that are of primary importance to the socioeconomic life of the nation. Chief among these, lying like a fan on the face of the land, are the Ganges-Padma, Brahmaputra-Jamuna, and Megna rivers. The climate of Bangladesh is dominated by seasonal monsoons. The country experiences a hot summer season with high humidity from March to June; a somewhat cooler but still hot and humid monsoon season from July through early October; and a cool, dry winter from November to the end of February. The fertile delta is frequented by natural calamities such as floods, cyclones, tidal bores, and drought. For administrative purposes, the country is divided into 6 divisions, 64 districts, and 496 upazilas (subdistricts) (BBS, 2001:19). Muslims constitute almost 90 percent of the population of Bangladesh, Hindus constitute about 9 percent, and others constitute about 1 percent. The national language of Bangladesh is Bangla, which is spoken and understood by all. Agriculture is the overwhelmingly dominant sector of the economy, occupying 80 percent of the total population and contributing 25 percent of gross domestic product (GDP). The average per capita income is as low as $444 and more than one-third of the population live below the absolute poverty line (UNDP, 2003). Rice, wheat, jute, sugarcane, tobacco, oilseeds, and potatoes are the principal crops. The country produces about 51 million kilograms of tea per year, a sizeable quantity of which is exported to foreign markets after meeting the internal demand. Bangladesh produces about 1,057,000 metric tons of superior quality jute annually and 16 percent of export earnings come from raw-jute manufactures (BBS, 2001). Industry, although small, is increasing in importance as a result of foreign investments. Unemployment/underemployment is a serious problem, and pressure on the land in rural areas has led to movement of people from rural to urban areas. Bangladesh ranks 138th in the Human Development Index (HDI) as presented in the Human Development Report of 2004 with a HDI value of 0.519. With this HDI value, the country belongs to the category of medium human development countries. However, Bangladesh ranks among the three last countries in this category, followed by Sudan and Cameroon. Within the region, the position of Bangladesh is better only than Pakistan, which ranks 144 in the HDI. 2 │ Introduction Bangladesh is still struggling to emerge from the realm of poverty. Bangladesh ranks 72nd among 94 developing countries in terms of the Human Poverty Index (HPI). The HPI is a multidimensional measure of poverty for developing countries; it takes into account social exclusion, lack of economic opportunities, and deprivations in survival, livelihood and knowledge. 1.2 POPULATION Bangladesh is the most densely populated country in the world, excluding city-states such as Hong Kong and Singapore. The country has a population of about 140 million, with a corresponding population density of more than 900 per square kilometer. During the first half of the last century the population increased by only 45 percent. This slow increase was due to a combination of high birth rates and high death rates. In the second half of the century, population growth was rapid, tripling during the period. The relatively young age structure of the population indicates continued rapid population growth in the future. According to the 2001 census, 39 percent of the population is under 15 years of age, 57 percent are between 15 and 64 years, and 4 percent are age 65 or over (BBS, 2003:51). This young age structure constitutes built-in “population momentum,” which will continue to generate population increases well into the future, even in the face of rapid fertility decline. The population projections indicate that the population will increase rapidly even after attaining replacement-level fertility because of the echo effect of the high fertility experienced in the past. The Bangladesh Population Policy indicates that the population should stabilize at 210 million by 2060, if replacement-level fertility is reached by 2010. This estimate of future population size is reasonably consistent with the World Bank projections from 1994 (Bos et al., 1994), and the United Nations projections 1996 revision (United Nations, 1996), both of which estimated a mid-21st century population of 218 million. However, there is wide disparity between the estimates of the Bangladesh Government and others on the time when the population would stabilize. The World Bank boldly forecast a final stationary population of 263 million by mid-22nd century (2150), whereas others have not projected beyond the mid-21st century. Recently however, the United Nations has revised their estimate for 2050 by 25 million (or 11 percent) to 243 million, apparently on the basis of the decade long fertility plateau (United Nations, 2004). This recent and very substantial upward revision of the mid-century population by the United Nations seems unduly pessimistic because a five-year delay in attaining replacement-level fertility adds only 3 percent to the population at any point in time. Nevertheless, Bangladesh still faces many decades of continued population growth, and efforts to slow that growth need to continue, through the family planning program, and increasingly, through social and health interventions that will facilitate further fertility decline, so that progress towards economic development is not hindered. 1.3 POPULATION, FAMILY PLANNING AND MATERNAL AND CHILD HEALTH POLICIES AND PROGRAMS Family planning was introduced in Bangladesh (then East Pakistan) in the early 1950s through the voluntary efforts of social and medical workers. The government, recognizing the urgency of moderating population growth, adopted family planning as a government-sector program in 1965. The policy to reduce fertility rates has been repeatedly reaffirmed by the Government of Bangladesh since liberation in 1971. The First Five-Year Plan (1973-1978) of Bangladesh emphasized “the necessity of immediate adoption of drastic steps to slow down the population growth” and reiterated that “no civilized measure would be too drastic to keep the population of Bangladesh on the smaller side of 15 crore (i.e., 150 million) for sheer ecological viability of the nation” (GOB, 1994:7). From mid- 1972, the family planning program received virtually unanimous, high-level political support. All Introduction │ 3 subsequent governments that have come into power in Bangladesh have identified population control as the top priority for government action. This political commitment is crucial in understanding the fertility decline in Bangladesh. In 1976, the government declared the rapid growth of the population as the country’s number one problem and adopted a broad-based, multisectoral family planning program along with an official population policy (GOB, 1994:9). Population planning was seen as an integral part of the total development process, and was incorporated into successive five-year plans. Policy guidelines and strategies for the population program are formulated by the National Population Council (NPC), which is chaired by the prime minister. Bangladesh’s population policy and programs have evolved through a series of developmental phases and have undergone changes in strategies, structure, content, and goals. In the mid-1970s, the government instituted the deployment of full-time, local Family Welfare Assistants (FWAs)— community-based family planning motivators and distributors who numbered almost 24,000 at the height of the program a few years ago. A social marketing program to promote the sale of birth control pills and condoms was also initiated in the mid-1970s. Another characteristic of the population program is the involvement of more than 200 nongovernmental organizations (NGOs). Since 1980, the program has stressed functionally integrated health and family planning programs. The goal is to provide an essential package of high quality, client-centered reproductive and child health care, family planning, communicable disease control, and limited curative services at a one- stop service point. The Fifth Five-Year Plan, the Health and Population Sector Program (HPSP) 1998- 2003 was formulated keeping in view the principles of the Health and Population Sector Strategy (HPSS) with a single sector for both health and population. The main objective of the HPSP was to ensure universal access to essential health care services of acceptable quality and to further slow population growth. Moreover, to overcome the multidimensional problems and to meet the challenge according to the spirit of the International Conference on Population and Development (ICPD), the Government of Bangladesh launched the Health, Nutrition and Population Sector Program (HNPSP) in 2003. This aimed to reform the health and population sector. The program entails provision of a package of essential and quality health care services responsive to the needs of the people, especially those of children, women, elderly and the poor. Recently, the government adopted the Bangladesh Population Policy with the objectives to improve the status of family planning, and maternal and child health, including reproductive health services and to improve the living standard of the people of Bangladesh through striking a desired balance between population and development in the context of the Millennium Development Goals (MDGs) and a Poverty Reduction Strategy Paper (PRSP). The objectives of the population policy include the following: • Reduce the total fertility rate (TFR) and increase the use of family planning methods among eligible couples through raising awareness of family planning; • Attain a net reproduction rate equal to one by the year 2010 so as to stabilize population around 2060; • Ensure adequate availability and access of reproductive health services, especially family planning services to all including information, counseling and services for adolescents; • Improve maternal health with emphasis on reduction of maternal mortality; • Reduce reproductive tract and sexually transmitted infections (RTI/STIs) and prevent the spread of HIV/AIDS; • Reduce infant and under-five mortality rates; • Reduce maternal and child malnutrition; 4 │ Introduction • Promote and actively support programs for elimination of gender disparity in education, health and nutrition; • Ensure early childhood development programs; • Ensure and support gender equity and empower women; • Develop the human resource capacity of planners, managers and service providers, including improved data collection, research and dissemination; • Actively support measures to provide food and social security and shelter for the disadvantaged including the elderly, destitute, physically and mentally retarded persons; • Actively support measures to regulate and reduce rural to urban migration; • Support measures for environmental sustainability with emphasis on access to safe drinking water; • Support poverty alleviating strategies and a conducive environment for improved quality of life; • Ensure coordination among relevant Ministries in strengthening population and development linkages and making their respective mandates and implementation strategies more population- focused. The government’s policy of providing health care is based on the principles of universal coverage and accessibility; optimum utilization and development of human resources for health; appropriate use of technology; gender equity; improvement of the quality of life; priority service for the most vulnerable groups including women, children, and the poor; and promotion of health as an integral part of overall socioeconomic development. Private-sector involvement in both health and population services is being encouraged. There has been considerable variation across the health and population sector in effectiveness of services. Some projects have very effectively targeted certain diseases or conditions, while others have been less effective. Under the HPSP, since 1998, the adoption of a sector-wide program approach has brought its own set of issues. However, throughout these strategy changes, the family planning program has continued to function reasonably effectively, and contraceptive use has steadily become more widespread. Numerous factors have contributed to the increase in contraceptive use over the past 20 years. The elements identified as having contributed to the success of the program are 1) strong political commitment to family planning programs by successive governments, 2) successful promotion of a small family norm through information and educational activities and other multisectoral programs, 3) establishment of a widespread infrastructure for delivering family planning and health services down to the village level, 4) increased involvement of nongovernmental organizations to supplement and complement the government’s efforts, 5) flexibility to make policy and programmatic adjustments in response to emerging needs, and 6) strong support of the program by the international aid community (GOB, 1994:36). The success achieved so far in the national family planning program is encouraging and has increased the confidence that it is possible to achieve further progress. But there remain several issues of concern, such as the tremendous growth potential built into the age structure as a consequence of past high fertility. Because of the increasing population entering childbearing age, the program will have to expand efforts substantially just to maintain the current level of contraceptive use. If demand for family planning also increases, that will put even more strain on the program. Other concerns are lack of a steady supply of contraceptives from external sources, which affects program performance; the need for further improvement in access to and quality of facilities and services; and the need for men to participate more actively in family planning acceptance. Introduction │ 5 1.4 ORGANIZATION OF THE 2004 BANGLADESH DEMOGRAPHIC AND HEALTH SURVEY 1.4.1 Survey Objectives and Implementing Organizations The Bangladesh Demographic and Health Survey (BDHS) is intended to serve as a source of population and health data for policymakers and the research community. In general, the objectives of the BDHS are to: • Assess the overall demographic situation in Bangladesh • Assist in the evaluation of the population and health programs in Bangladesh • Advance survey methodology. More specifically, the objective of the BDHS survey is to provide up-to-date information on fertility and childhood mortality levels; nuptiality; fertility preferences; awareness, approval, and use of family planning methods; breastfeeding practices; nutrition levels; and maternal and child health. This information is intended to assist policymakers and administrators in evaluating and designing programs and strategies for improving health and family planning services in the country. The 2004 BDHS survey was conducted under the authority of the National Institute for Population Research and Training (NIPORT) of the Ministry of Health and Family Welfare. The survey was implemented by Mitra and Associates, a Bangladeshi research firm located in Dhaka. ORC Macro of Calverton, Maryland, provided technical assistance to the project as part of its international Demographic and Health Surveys program, and financial assistance was provided by the U.S. Agency for International Development (USAID)/Bangladesh. 1.4.2 Sample Design The sample for the 2004 BDHS covered the entire population residing in private dwellings units in the country. Administratively, Bangladesh is divided into six divisions. In turn, each division is divided into zilas, and in turn each zila into upazilas. Each urban area in the upazila is divided into wards, and into mahallas within the ward; each rural area in the upazila is divided into union parishads (UP) and into mouzas within the UPs. The urban areas were stratified into three groups, i) Standard metropolitan areas, ii) Municipality areas, and iii) Other urban areas. These divisions allow the country as a whole to be easily separated into rural and urban areas. For the 2001 census, subdivisions called enumeration areas (EAs) were created based on a convenient number of dwellings units. Because sketch maps of EAs were accessible, EAs were considered suitable to use as primary sampling units (PSUs) for the 2004 BDHS. In each division, the list of EAs constituted the sample frame for the 2004 BDHS survey. A target number of completed interviews with eligible women for the 2004 BDHS was set at 10,000, based on information from the 1999-2000 BDHS. The 2004 BDHS sample is a stratified, a multistage cluster sample consisting of 361 PSUs, 122 in the urban area and 239 in the rural area. After the target sample was allocated to each group area according to urban and rural areas, the number of PSUs was calculated in terms of an average of 28 completed interviews of eligible women per PSU (or an average of 30 selected households per PSU). Mitra and Associates conducted a household listing operation in all the sample points from 3 October 2003 to 15 December 2003. A systematic sample of 10,811 households was then selected from these lists. All ever-married women age 10-49 in the selected households were eligible respondents for 6 │ Introduction the women’s questionnaire. For the men‘s survey, 50 percent of the selected households were chosen through systematic sampling. Interviewers interviewed one randomly selected man, regardless of marital status, in the age group 15-54, from each of the selected households. It was expected that the sample would yield interviews with approximately 10,000 ever-married women age 10-49 and 4,400 men age 15- 54. 1.4.3 Questionnaires The BDHS used a Household Questionnaire, a Women’s Questionnaire, a Men’s Questionnaire, and a Community Questionnaire. The contents of these questionnaires was based on MEASURE DHS+ model questionnaire. These model questionnaires were adapted for use in Bangladesh during a series of meetings with the Technical Task Force, which consisted of representatives from NIPORT, Mitra and Associates, USAID/Dhaka, ICDDR,B’s Center for Health and Population Research, Bangladesh, Pathfinder/Dhaka, and ORC Macro (see Appendix D for a list of members). Draft questionnaires were then circulated to other interested groups and were reviewed by the BDHS Technical Review Committee (see Appendix D). The questionnaires were developed in English and then translated into and printed in Bangla. In addition, two versions of a Verbal Autopsy Questionnaire were used. One version was for neonatal deaths (0-28 days old at death) and the other was for deaths among older children (age 29 days to 5 years at death). The verbal autopsy instruments were developed using the previous two BDHS verbal autopsy surveys, the WHO verbal autopsy questionnaire, and the instrument used since 2003 in the Matlab Health and Demographic Surveillance System. The Household Questionnaire was used to list all the usual members and visitors in the selected households. Some basic information was collected on the characteristics of each person listed, including his/her age, sex, education, and relationship to the head of the household. The main purpose of the Household Questionnaire was to identify women and men who were eligible for individual interview. In addition, information was collected about the dwelling itself, such as the source of water, type of toilet facilities, materials used to construct the house, and ownership of various consumer goods. The arsenic level of the water used by households for drinking was also tested. The Household Questionnaire was also used to record the heights and weights of all children under six years of age. The Women’s Questionnaire was used to collect information from ever-married women age 10- 49. These women were asked questions on the following topics: • Background characteristics (age, education, religion, etc.) • Reproductive history • Knowledge and use of family planning methods • Antenatal and delivery care • Breastfeeding and weaning practices • Vaccinations and health of children under age five • Marriage • Fertility preferences • Husband’s background and respondent’s work • Awareness of AIDS and other sexually transmitted diseases • Causes of deaths of children under five years of age The Men’s Questionnaire was used to collect information from men age 15-54 whether ever married or not. The men were asked questions on the following topics: • Background characteristics (including respondent’s work) • Health and life style (illness, use of tobacco) Introduction │ 7 • Marriage and sexual activity • Participation in reproductive health care • Awareness of AIDS and other sexually transmitted diseases • Attitudes on women’s decision making roles • Domestic violence The Community Questionnaire was completed for each sample cluster and included questions about the existence of development organizations in the community and the availability and accessibility of health and family planning services. The Verbal Autopsy Questionnaire was used for collection of open-ended information including narrative stories on the following topics: • Identification including detailed address of respondent • Informed consent • Detailed age description of deceased child • Information about caretaker or respondent of deceased child • Detailed birth and delivery information • Open-ended section allowing the respondent to provide a narrative history • Maternal history including questions on prenatal care, labor and delivery, and obstetrical complications • Information about accidental deaths • Detailed signs and symptoms preceding death • Treatment module and information on direct, underlying • Contributing causes of death from the death certificate, if available. The survey questionnaires, with the exception of the Verbal Autopsy Questionnaires, are included in Appendix E. 1.4.4 Training and Fieldwork The BDHS Women’s Questionnaire was pretested in September 2003 and the Men’s Questionnaire was pretested in December 2003. For the pretest, male and female interviewers were trained at the office of Mitra and Associates. After training, the teams conducted interviews in various locations in the field under the observation of staff from Mitra and Associates and members of the Technical Task Force (TTF). Altogether, 108 Women‘s and 45 Men’s Questionnaires were completed. Based on observations in the field and suggestions made by the pretest field teams, the TTF made revisions in the wording and translations of the questionnaires. In November 2003, candidates for field staff positions for the main survey were recruited. Recruitment criteria included educational attainment, maturity, ability to spend one month in training and at least four months in the field, and experience in other surveys. Training for the main survey was conducted for four weeks (1 December to 30 December 2003). Initially, training consisted of lectures on how to complete the questionnaires, with mock interviews between participants to gain practice in asking questions. Towards the end of the training course, the participants spent several days in practice interviewing in various places close to Dhaka. Trainees whose performance was considered superior were selected as supervisors and field editors. Fieldwork for the BDHS was carried out by 12 interviewing teams. Each consisted of one male supervisor, one female field editor, five female interviewers, two male interviewers, and one logistics 8 │ Introduction staff person, for a total of 120 field staff for the survey. Mitra and Associates also fielded four quality control teams of two persons each to check on the field teams. In addition to these field control teams, NIPORT monitored fieldwork by using their quality control teams. Additionally, USAID, ORC Macro, and NIPORT monitored the fieldwork by visiting teams in the field. Fieldwork commenced on 1 January 2004 and was completed on 25 May 2004. Fieldwork was implemented in five phases. 1.4.5 Data Processing All questionnaires for the BDHS were periodically returned to Dhaka for data processing at Mitra and Associates. The processing of the data collected began shortly after the fieldwork commenced. The processing operation consisted of office editing, coding of open-ended questions, data entry, and editing inconsistencies found by the computer programs. The data were processed on six microcomputers working in double shifts and carried out by 10 data entry operators and two data entry supervisors. The concurrent processing of the data was an advantage since the quality control teams were able to advise field teams of problems detected during the data entry. In particular, tables were generated to check various data quality parameters. Data processing commenced on 12 January 2004 and was completed by 24 June 2004. 1.4.6 Coverage of the Sample Table 1.1 shows response rates for the survey and reasons for nonresponse. A total of 10,811 households were selected for the sample; 10,523 were occupied, of which 10,500 were successfully interviewed. The shortfall is primarily due to dwellings that were vacant or destroyed or in which the inhabitants had left for an extended period at the time the interviewing teams visited them. Of the households occupied, 99.8 percent were successfully interviewed. In these households, 11,601 women Table 1.1 Results of the household and individual interviews Number of households, number of interviews, and response rates, according to residence, Bangladesh 2004 Residence Urban Rural Total Result Number Percent Number Percent Number Percent Household sample Households occupied 3,522 96.6 7,001 97.7 10,523 97.3 Households absent for extended period of time 46 1.3 76 1.1 122 1.1 Dwelling vacant or destroyed 68 1.9 74 1.0 142 1.3 Other 10 0.3 14 0.2 24 0.2 Total households selected 3,646 100.0 7,165 100.0 10,811 100.0 Household interviews Households interviewed 3,513 99.7 6,987 99.8 10,500 99.8 Household not interviewed 9 0.3 14 0.2 23 0.2 Total households occupied 3,522 100.0 7,001 100.0 10,523 100.0 Household response rate 99.7 99.8 99.8 Individual interviews: women Eligible women interviewed 3,904 98.3 7,536 98.8 11,440 98.6 Eligible women not interviewed 69 1.7 92 1.2 161 1.4 Total eligible women 3,973 100.0 7,628 100.0 11,601 100.0 Eligible woman response rate 98.3 98.8 98.6 Individual interviews: men Eligible men interviewed 1,514 95.6 2,783 95.7 4,297 95.7 Eligible men not interviewed 69 4.4 124 4.3 193 4.3 Total eligible men 1,583 100.0 2,907 100.0 4,490 100.0 Eligible man response rate 95.6 95.7 95.7 Introduction │ 9 were identified as eligible for the individual interview (i.e., ever-married and age 10-49) and interviews were completed for 11,440 or 98.6 percent of them. In households that were selected for inclusion in the man’s survey, 4,490 eligible men age 15-54 were identified, of which 4,297 or 95.7 percent were interviewed. The principal reason for nonresponse among eligible women and men was the failure to find them at home despite repeated visits to the household. The nonresponse rates for the current survey were lower than those for the 1999-2000 survey. Household Population and Housing Characteristics │ 11 HOUSEHOLD POPULATION AND HOUSING CHARACTERISTICS 2 This chapter provides information on the social, economic, and demographic characteristics of the households included in the 2004 BDHS. It presents information on household population characteristics, such as age, sex, educational attainment, and employment status. It also provides information on household characteristics, such as source of drinking water, electricity, sanitation facilities, housing construction materials, possession of durable goods, and ownership of a homestead and land. This information is intended to assist in the assessment of the representativeness of the survey. In the 2004 BDHS, information was also collected on the prevalence of arsenic in household drinking water, through testing of arsenic in drinking water at the time of the survey. For the purpose of the 2004 BDHS, a household is defined as a person or a group of related and/or unrelated persons who usually live in the same dwelling unit(s), who have common cooking and eating arrangements, and who acknowledge one adult member as a head of the household. A member of the household is any person who usually lives in the household; a visitor is someone who is not a member of the household, but stayed in the household the night before the interview. This survey collected information for all usual residents of the selected household (de jure population) and persons who stayed in the selected household the night before the date of interview (de facto population). This method of data collection allows the analysis of either de jure or de facto population. 2.1 HOUSEHOLD POPULATION BY AGE, SEX, AND RESIDENCE Age and sex are important demographic variables and are the primary basis of demographic classification in vital statistics, censuses, and surveys. They are also important variables in the study of mortality, fertility, and nuptiality. The distribution of the de facto household population in the 2004 BDHS is shown in Table 2.1 by five-year age groups, according to sex and urban-rural residence. The household population includes 51,255 persons, and the sample is 78 percent rural and 22 percent urban. Females constitute 51 percent of the population. Overall, the proportions of persons in the younger age groups are substantially larger than the proportions in the older age groups for each sex and in both urban and rural areas. The age-sex structure of the population is shown by a population pyramid in Figure 2.1. The pyramid is wide based and slightly narrower at the lowest base, a pattern that typically describes a high- fertility regime that has recently declined slightly. Figure 2.2 shows the distribution of the male and female household population by single year of age. The figure shows that noticeable heaping is observed at ages ending with 0 and 5. Ages ending with 1 and 9 are underreported. 12 │ Household Population and Housing Characteristics Table 2.1 Household population by age, sex, and residence Percent distribution of the de facto household population by five-year age groups, according to sex and residence, Bangladesh 2004 Urban Rural Total Age Male Female Total Male Female Total Male Female Total <5 11.3 11.5 11.4 13.8 12.7 13.2 13.2 12.4 12.8 5-9 12.5 10.8 11.7 13.6 12.8 13.2 13.3 12.4 12.8 10-14 11.7 12.6 12.2 13.0 12.3 12.7 12.7 12.4 12.6 15-19 10.5 13.5 12.0 9.5 12.5 11.0 9.8 12.7 11.3 20-24 8.6 11.1 9.9 7.1 9.7 8.4 7.4 10.0 8.7 25-29 7.4 9.1 8.3 6.6 7.9 7.2 6.8 8.1 7.5 30-34 7.7 7.3 7.5 6.6 6.7 6.7 6.8 6.9 6.8 35-39 6.6 6.2 6.4 6.0 5.5 5.7 6.1 5.7 5.9 40-44 6.7 4.8 5.7 5.5 4.3 4.9 5.8 4.4 5.1 45-49 5.0 4.1 4.5 4.4 4.2 4.3 4.5 4.2 4.4 50-54 3.5 2.5 3.0 3.5 2.6 3.1 3.5 2.6 3.0 55-59 2.3 1.9 2.1 2.2 2.8 2.5 2.2 2.6 2.4 60-64 2.4 1.7 2.1 2.7 2.2 2.4 2.6 2.1 2.3 65-69 1.4 0.9 1.1 1.7 1.3 1.5 1.7 1.2 1.4 70-74 1.2 0.7 1.0 1.8 1.1 1.4 1.7 1.0 1.3 75-79 0.5 0.4 0.4 0.7 0.5 0.6 0.7 0.5 0.6 80+ 0.7 0.8 0.7 1.3 0.9 1.1 1.2 0.8 1.0 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Number 5,530 5,721 11,251 19,570 20,435 40,004 25,099 26,156 51,255 Household Population and Housing Characteristics │ 13 Table 2.2 compares the broad age structure of the population from the 1989 Bangladesh Fertility Survey; the 1989 and 1991 Contraceptive Prevalence Surveys; and the 1993-1994, 1996-1997, 1999- 2000, and 2004 Bangladesh Demographic and Health Surveys. There has been a decline in the proportion of population less than 15 years of age and an increase in the proportion age 15-59; also, the median age of the population has been increasing. This pattern is consistent with a decline in fertility. Table 2.2 Population by age from selected sources Percent distribution of the de facto population by age group, selected sources, Bangladesh 1989-2004 Age group 1989 BFS 1989 CPS 1991 CPS 1993-1994 BDHS 1996-1997 BDHS 1999-2000 BDHS 2004 BDHS <15 43.2 43.2 42.7 42.6 41.0 39.2 38.2 15-59 50.9 50.9 51.2 51.2 53.1 54.4 55.1 60+ 5.9 5.9 6.0 6.2 5.9 6.4 6.6 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Median age u u u 18.4 18.8 19.5 20.2 u = Unknown (not available) Source: Huq and Cleland, 1990:38; Mitra et al., 1993:14; Mitra et al., 1997:9; NIPORT et al., 2001:11 14 │ Household Population and Housing Characteristics 2.2 HOUSEHOLD COMPOSITION Table 2.3 shows the distribution of households by sex of the head of household and by household size, according to urban-rural residence. At the national level and in rural areas, women head 10 percent of Bangladeshi households. The proportion of female-headed households is slightly lower in urban areas (9 percent). Fifty-eight percent of households in Bangladesh are composed of three to five members; the corresponding figure for urban areas is 60 percent. Average household size in Bangladesh is 5.0 members; urban household size is marginally smaller than rural household size. Table 2.3 Household composition Percent distribution of households by sex of head of household and household size, according to residence, Bangladesh 2004 Residence Characteristic Urban Rural Total Sex of head of household Male 90.6 89.7 89.9 Female 9.4 10.3 10.1 Total 100.0 100.0 100.0 Number of usual members 1 1.4 1.8 1.7 2 6.8 7.2 7.1 3 16.0 15.3 15.5 4 23.3 22.5 22.7 5 20.7 19.4 19.7 6 14.8 14.2 14.4 7 7.0 8.5 8.2 8 4.3 4.6 4.5 9+ 5.6 6.4 6.2 Total 100.0 100.0 100.0 Number of households 2,306 8,194 10,500 Mean size 4.9 5.0 5.0 Note: Table is based on de jure members, i.e., usual residents. 2.3 EDUCATIONAL ATTAINMENT OF HOUSEHOLD MEMBERS Education is a key determinant of the lifestyle and status an individual enjoys in a society. It affects many aspects of life, including demographic and health behavior. Studies have shown that educational attainment has strong effects on reproductive behavior, contraceptive use, fertility, mortality, morbidity, and attitudes and awareness related to family health and hygiene. In the 2004 BDHS, information on educational attainment was collected for each member of the household age six years and above. Tables 2.4.1 and 2.4.2 present the distribution of the de facto female and male household populations, respectively, by educational level, according to age, urban-rural residence, and administrative division. Generally, education has become more widespread over the years among both men and women, as is clear from a comparison of older and younger cohorts. Household Population and Housing Characteristics │ 15 Table 2.4.1 Level of education of household population: women Percent distribution of the de facto female household population age six and over by highest level of education attended or completed, according to background characteristics, Bangladesh 2004 Background characteristic No education Primary incomplete Primary complete1 Secondary incomplete Secondary complete or higher2 Total Number of women Median number of years Age 6-9 12.4 87.0 0.1 0.4 0.0 100.0 2,573 0.2 10-14 7.3 49.5 5.6 37.5 0.1 100.0 3,245 3.6 15-19 12.0 17.5 9.8 52.9 7.7 100.0 3,321 5.7 20-24 23.9 17.7 9.4 32.6 16.4 100.0 2,625 4.6 25-29 40.1 20.0 9.1 19.1 11.8 100.0 2,129 1.9 30-34 48.4 22.1 7.9 14.8 6.8 100.0 1,795 0.0 35-39 56.0 20.8 6.7 11.6 4.8 100.0 1,479 0.0 40-44 56.4 18.4 8.0 12.5 4.7 100.0 1,159 0.0 45-49 61.8 18.9 8.3 7.6 3.5 100.0 1,095 0.0 50-54 69.6 13.3 7.0 7.5 2.5 100.0 684 0.0 55-59 74.2 17.1 5.3 2.8 0.6 100.0 676 0.0 60-64 80.1 10.2 4.5 4.3 0.5 100.0 547 0.0 65+ 85.0 8.9 2.9 2.7 0.2 100.0 915 0.0 Residence Urban 28.2 27.1 6.8 25.5 12.3 100.0 4,932 3.2 Rural 36.2 31.5 6.8 21.8 3.7 100.0 17,317 1.2 Division Barisal 24.8 35.2 9.7 23.3 7.0 100.0 1,480 2.8 Chittagong 34.8 29.5 6.5 23.9 5.2 100.0 4,374 1.7 Dhaka 34.6 30.0 7.0 21.3 7.1 100.0 6,836 1.5 Khulna 29.0 32.4 6.0 27.2 5.4 100.0 2,544 2.6 Rajshahi 37.4 30.6 5.2 22.2 4.6 100.0 5,387 1.1 Sylhet 40.8 27.5 10.3 18.0 3.3 100.0 1,628 0.7 Total 34.4 30.5 6.8 22.6 5.6 100.0 22,249 1.6 Note: Total includes eight females missing information on age. Total includes a small proportion missing education level. 1 Primary complete is defined as completing grade 5. 2 Secondary complete or higher is defined as completing grade 10 or more. As expected, educational attainment is higher in urban areas than in rural areas. For example, 72 percent of females and 80 percent of males in urban areas have some education, compared with 64 and 71 percent of females and males, respectively, in rural areas. Differences in the median number of years of schooling between urban and rural persons are around two years for females and males. However, the urban-rural gap is narrowing: Figure 2.3 shows that since 1993-1994, education has improved more rapidly among women and men in rural areas than in urban areas. 16 │ Household Population and Housing Characteristics Table 2.4.2 Level of education of household population: men Percent distribution of the de facto male household population age six and over by highest level of education attended or completed, according to background characteristics, Bangladesh 2004 Background characteristic No education Primary incomplete Primary complete1 Secondary incomplete Secondary complete or higher2 Total Number of men Median number of years Age 6-9 15.5 84.0 0.0 0.4 0.0 100.0 2,726 0.1 10-14 8.7 56.4 5.9 29.0 0.0 100.0 3,194 3.1 15-19 13.3 20.0 12.1 45.7 8.9 100.0 2,449 5.0 20-24 15.8 17.4 12.1 30.9 23.8 100.0 1,858 5.3 25-29 27.9 17.1 9.3 25.1 20.6 100.0 1,701 4.4 30-34 34.7 16.5 8.7 20.5 19.5 100.0 1,712 3.8 35-39 40.2 17.2 7.9 17.6 16.8 100.0 1,535 2.3 40-44 43.0 19.2 7.9 18.6 11.2 100.0 1,445 1.3 45-49 41.3 17.5 6.9 19.4 14.8 100.0 1,135 2.1 50-54 42.3 16.4 7.5 16.5 17.3 100.0 879 2.2 55-59 39.4 17.0 7.8 16.9 18.7 100.0 564 2.5 60-64 45.7 20.4 8.4 15.3 10.0 100.0 652 0.3 65+ 50.0 19.0 9.2 13.8 7.9 100.0 1,303 0.0 Residence Urban 20.1 27.3 7.2 25.1 20.2 100.0 4,762 4.2 Rural 28.6 33.8 7.8 21.3 8.5 100.0 16,397 2.1 Division Barisal 18.0 39.3 7.1 22.3 13.0 100.0 1,312 3.2 Chittagong 25.9 33.2 7.4 23.2 10.2 100.0 3,910 2.7 Dhaka 26.9 32.0 7.9 20.8 12.3 100.0 6,640 2.6 Khulna 23.7 30.4 7.0 26.2 12.6 100.0 2,539 3.5 Rajshahi 29.6 31.3 6.6 22.0 10.4 100.0 5,236 2.3 Sylhet 29.8 32.3 12.2 19.0 6.8 100.0 1,522 2.0 Total 26.7 32.3 7.6 22.2 11.1 100.0 21,159 2.7 Note: Total includes seven males missing information on age. Total includes a small proportion missing education level. 1 Primary complete is defined as completing grade 5. 2 Secondary complete or higher is defined as completing grade 10. The proportion of the population that has achieved some education varies among administrative divisions. The proportions of women and men with some education are the highest in Barisal division (75 percent of women and 82 percent of men) and the lowest in Rajshahi and Sylhet divisions (around six in ten women and seven in ten men). The positive impact of different interventions to encourage women to attend school is also evident from the data available through the various surveys. In 2004, 70 percent of women age 15-19 had completed at least primary education, compared with 44 percent in 1993-1994, and 16 percent of women age 20-24 had completed at least secondary education, compared with 9 percent ten years ago. This steady increase in female education has narrowed the gap in education levels between males and females in younger cohorts. In fact, the 2004 BDHS indicates that a higher percentage of women age 15-19 have Household Population and Housing Characteristics │ 17 completed at least primary education, compared with men in the same age group. Nevertheless, men are still more likely than women to have completed secondary school. The difference, however, has narrowed in the last ten years. In 1993-1994, men age 20-24 were 2.3 times more likely than women in the same age group to have completed secondary school; in 2004 this ratio had declined to 1.5 times. 2.3.1 School Attendance Data for school attendance by age group, sex, and place of residence are presented in Table 2.5. In the whole country, more than eight in ten (84 percent) children age 6-10 years are in school. More than two-thirds of children age 11-15 years are attending school. School attendance of both age groups is higher for girls than for boys. Also, boys and girls age 6-15 years are more likely to attend school in rural areas than in urban areas. For those age 16-20, school attendance drops sharply from levels seen at younger ages, and attendance is higher for males than for females, and in urban areas than in rural areas. Table 2.5 School attendance Percentage of the de facto household population age 6-24 years attending school, by age, sex, and residence, Bangladesh 2004 Male Female Total Age Urban Rural Total Urban Rural Total Urban Rural Total 6-10 81.1 83.0 82.6 83.2 86.9 86.2 82.1 84.9 84.4 11-15 63.0 66.9 66.1 65.9 71.3 70.1 64.6 69.2 68.2 6-15 72.6 75.5 74.9 74.0 79.4 78.2 73.3 77.4 76.6 16-20 33.3 29.2 30.2 32.2 22.0 24.4 32.7 25.3 27.0 21-24 26.2 14.5 17.6 11.3 5.1 6.7 17.6 8.8 11.1 Over the last decade, school attendance has increased both for girls and boys in the younger age group (6-15 years), but the increases have been higher for girls than for boys. The data suggest that for 18 │ Household Population and Housing Characteristics those age 16-20, school attendance has actually fallen for males, while for females it has seen a slight increase. 2.4 EMPLOYMENT The 2004 BDHS Household Questionnaire included questions on whether each person age eight and above in the sampled households was working at the time of the survey and how he or she was compensated (or if at all) for work. Table 2.6 shows the percentage of women and men age eight and over working at the time of survey, by age and residence. Men are much more likely than women to be employed, regardless of age group or residence. Overall, 68 percent of men eight years or older are employed, compared with only 15 percent of female household members. Employment is higher in urban than in rural areas for males and females. Female employment rates are highest for women age 25-44; at least one in four women in this age group is working. Half of men age 65 and over are employed. Table 2.6 Employment status Percentage of female and male household population age eight and over working at the time of the survey, by age, sex, and residence, Bangladesh 2004 Female Male Age Urban Rural Total Urban Rural Total 8-9 2.3 0.4 0.8 2.6 2.3 2.3 10-14 10.4 3.1 4.7 19.7 17.6 18.0 15-19 17.7 7.4 9.7 54.5 58.0 57.2 20-24 22.0 15.4 17.0 71.2 79.6 77.5 25-29 26.1 24.9 25.2 89.6 92.5 91.8 30-34 30.7 26.9 27.8 97.8 96.8 97.0 35-39 32.3 26.8 28.1 98.4 98.6 98.6 40-44 31.5 24.5 26.2 97.3 98.2 98.0 45-49 23.1 21.5 21.8 98.1 98.3 98.3 50-54 19.1 14.7 15.5 94.3 94.8 94.7 55-59 12.9 8.7 9.4 86.2 93.8 92.1 60-64 11.3 7.7 8.4 71.6 79.4 77.8 65+ 2.2 4.2 3.9 49.0 53.2 52.5 Total 19.9 13.8 15.2 68.4 67.2 67.5 Number of persons 4,628 16,365 20,994 4,614 16,051 20,665 Table 2.7 presents the percent distribution of employed women and men by type of earnings, according to background characteristics. Almost 80 percent of working women and about 70 percent of the employed men work for cash only. The proportion of employed women who work for cash only is markedly higher for those who have completed secondary education. Employed men in urban areas and those who have completed secondary education are much more likely to be working for cash only. Household Population and Housing Characteristics │ 19 Table 2.7 Form of earnings Percent distribution of currently employed men and women age eight and over by type of earnings (cash, in kind, no payment), according to background characteristics, Bangladesh 2004 Women Men Background characteristic Earns cash only Earns kind only Both cash and kind Not paid Don’t know/ missing Total Number of employed women Earns cash only Earns kind only Both cash and kind Not paid Don’t know/ missing Total Number of employed men Age 8-9 * * * * * 100.0 9 (51.0) (14.0) (7.1) (27.9) (0.0) 100.0 31 10-14 51.7 7.6 35.2 4.5 1.0 100.0 154 58.3 13.3 7.9 20.3 0.2 100.0 589 15-19 76.7 5.0 12.1 5.5 0.7 100.0 319 67.3 16.9 6.7 9.1 0.0 100.0 1,488 20-24 84.6 3.3 8.6 3.5 0.0 100.0 440 75.8 14.7 5.9 3.6 0.0 100.0 1,503 25-29 84.3 5.4 6.9 3.1 0.3 100.0 537 76.4 17.5 5.2 0.8 0.0 100.0 1,644 30-34 82.7 5.5 10.0 1.8 0.0 100.0 500 78.3 16.9 4.4 0.2 0.2 100.0 1,739 35-39 77.1 8.7 11.3 2.4 0.5 100.0 416 76.0 17.4 6.2 0.4 0.0 100.0 1,610 40-44 81.1 6.4 10.1 2.3 0.0 100.0 306 75.8 18.7 5.4 0.0 0.0 100.0 1,494 45-49 76.4 9.2 11.1 3.3 0.0 100.0 242 72.7 20.8 6.4 0.0 0.2 100.0 1,194 50-54 73.8 10.0 14.9 1.3 0.0 100.0 107 69.1 24.5 6.3 0.1 0.0 100.0 876 55-59 64.8 19.1 16.2 0.0 0.0 100.0 65 60.5 32.8 6.7 0.0 0.0 100.0 549 60-64 (61.4) (10.8) (23.6) (4.1) (0.0) 100.0 50 56.2 34.3 9.3 0.1 0.1 100.0 529 65+ (54.7) (6.4 (34.9 (3.9) (0.0) 100.0 37 48.7 39.4 10.9 0.9 0.1 100.0 705 Residence Urban 80.7 2.8 14.9 1.7 0.0 100.0 922 89.3 5.2 3.8 1.6 0.1 100.0 3,157 Rural 77.2 8.1 10.7 3.6 0.3 100.0 2,262 65.8 24.4 7.0 2.8 0.1 100.0 10,793 Division Barisal 71.2 10.1 12.3 5.2 1.1 100.0 110 76.2 16.4 5.7 1.5 0.2 100.0 786 Chittagong 74.6 6.5 13.8 4.6 0.5 100.0 481 75.4 15.9 6.6 2.0 0.1 100.0 2,384 Dhaka 82.2 4.8 9.8 3.0 0.1 100.0 1,071 72.9 20.4 4.4 2.2 0.0 100.0 4,350 Khulna 79.3 6.4 8.8 5.6 0.0 100.0 374 71.7 19.4 5.8 3.0 0.1 100.0 1,799 Rajshahi 78.4 8.1 12.1 1.5 0.0 100.0 965 64.5 23.7 8.3 3.4 0.0 100.0 3,650 Sylhet 65.1 7.7 24.7 1.3 1.3 100.0 183 72.0 19.4 7.1 1.4 0.1 100.0 982 Education No education 73.4 7.8 16.1 2.6 0.1 100.0 1,649 69.6 21.6 7.9 0.8 0.0 100.0 4,858 Primary incomplete 74.8 8.1 12.4 4.2 0.4 100.0 642 69.6 21.0 5.9 3.4 0.1 100.0 3,090 Primary complete 84.3 5.1 7.8 2.8 0.0 100.0 193 67.8 23.6 6.4 2.1 0.0 100.0 1,348 Secondary incomplete 87.4 3.8 3.6 4.5 0.7 100.0 451 69.5 20.2 5.5 4.8 0.1 100.0 2,946 Secondary complete or higher 97.7 0.8 1.1 0.5 0.0 100.0 248 83.5 11.0 3.3 2.0 0.2 100.0 1,702 Wealth index Lowest 69.0 11.3 16.6 3.0 0.1 100.0 839 50.1 18.8 6.6 2.1 0.0 100.0 2,723 Second 82.0 6.0 7.6 4.4 0.1 100.0 688 64.9 24.4 7.8 2.8 0.1 100.0 2,827 Middle 80.7 7.3 7.9 3.7 0.4 100.0 540 65.3 25.9 5.6 3.1 0.0 100.0 2,858 Fourth 85.0 4.5 7.1 3.0 0.4 100.0 482 69.2 22.0 6.2 2.6 0.0 100.0 2,741 Highest 79.1 1.9 17.5 1.2 0.3 100.0 635 83.9 9.0 5.1 1.8 0.2 100.0 2,801 Total 78.2 6.6 11.9 3.1 0.2 100.0 3,184 71.1 20.1 6.3 2.5 0.1 100.0 13,951 Note: Total includes one woman and seven men missing information on education. Figures in parentheses are based on 25-49 unweighted cases; an asterisk indicates that a figure is based on fewer than 25 unweighted cases and has been suppressed. 20 │ Household Population and Housing Characteristics 2.5 HOUSING CHARACTERISTICS In the Household Questionnaire, respondents were asked about characteristics of their house- holds, including access to electricity, source of drinking water, type of sanitation facilities, fuel used for cooking, and main housing materials. Table 2.8 summarizes this information by residence. In Bangladesh, 41 percent of households have access to electricity. Coverage of electricity has increased by 9 percentage points since 1999-2000. There is a wide urban-rural gap with regard to access to electricity, with three in four households in urban areas and less than one in three households in rural areas having electricity. This gap, though wide, has narrowed since 1999-2000. Information on household source of drinking water is important because potentially fatal diseases, including typhoid, cholera, and dysentery, are prevalent in unprotected water sources. Tubewells are the predominant source of drinking water throughout Bangladesh. Almost seven in ten households in urban areas and more than nine in ten households in rural areas obtain drinking water from tubewells. Piped water is accessible only in urban areas; around one-third of urban households drink piped water, mainly piped inside the dwelling. The household drinking water was also tested for levels of arsenic. The results are presented in the last section of this chapter. Access to adequate sanitation facilities is an important determinant of health conditions. Table 2.8 shows that 86 percent of Bangladeshi households have some type of sanitation facility, including 59 percent that have hygienic toilets (septic tank/modern toilets, water-sealed/slab latrines, and pit latrines). As expected, sanitation facilities vary between rural and urban areas. Only 55 percent of rural households have hygienic toilets, compared with 71 percent of urban households. Moreover, 16 percent of rural households have no facility at all, compared with only 4 percent of urban households. A question about fuel used for cooking was also asked in the Household Questionnaire. Two types of fuel are predominantly used for cooking in Bangladesh: wood and crop residue. More than one- third of all households (35 percent) use wood; almost half use crop residue or straw for cooking. In urban areas, wood (40 percent) is the primary fuel used for cooking, followed by liquid gas or gas (31 percent), whereas the majority of rural residents depend on crop residue (56 percent), followed by wood (34 percent). In urban areas, one in five households uses crop residue as fuel for cooking. Tin is the most common roofing material in Bangladesh, accounting for 72 percent of rural households and 86 percent of urban households. However, urban and rural households vary widely in the use of cement or concrete for roofs. In urban areas, 23 percent of households live in dwellings with cement or concrete roofs, while in rural areas only 4 percent of household roofs are made of the same material. Since 1999-2000, tin roofing has become more popular in both urban and rural areas; in rural areas, the use of bamboo or thatch for roofing has declined dramatically. Almost half of all households in Bangladesh live in structures with walls made of natural materials, such as jute, bamboo, or mud. Seventeen percent live in houses with brick or cement walls, and 34 percent live in houses with tin walls. Tin as a material for walls has become almost twice as popular both in urban and rural areas in 2004 compared with 1999-2000. Urban households live in more solid dwellings than rural households. More than four in ten urban households live in structures with brick or cement walls, compared with only 9 percent of rural households. Household Population and Housing Characteristics │ 21 Table 2.8 Household characteristics Percent distribution of households by household characteristics, according to residence, Bangladesh 2004 Residence Household characteristic Urban Rural Total Electricity Yes 76.6 30.4 40.6 No 23.4 69.5 59.4 Total 100.0 100.0 100.0 Source of drinking water Piped inside dwelling 23.3 0.0 5.1 Piped outside dwelling 7.8 0.1 1.8 Tubewell 65.5 92.9 86.8 Shallow tubewell 0.1 0.3 0.3 Deep tubewell 2.4 3.2 3.0 Surface well/other well 0.1 1.2 1.0 Pond/tank/lake 0.8 1.9 1.6 River/stream 0.1 0.4 0.3 Total 100.0 100.0 100.0 Sanitation facility Septic tank/modern toilet 29.5 3.2 9.0 Water-sealed/slab latrine 18.5 12.8 14.1 Pit latrine 23.1 39.0 35.5 Open latrine 23.0 26.7 25.9 Hanging latrine 1.7 1.9 1.9 No facility/bush/field 4.2 16.4 13.7 Total 100.0 100.0 100.0 Cooking fuel Wood 40.2 34.0 35.3 Crop residue 20.2 56.3 48.3 Dung cakes 5.5 8.8 8.1 Liquid gas/gas 30.9 0.5 7.2 Other 3.2 0.4 1.0 Total 100.0 100.0 100.0 Main roof material Katcha (bamboo/thatch) 4.2 9.7 8.5 Tin 72.1 86.4 83.3 Cement, concrete, or tile 23.4 3.8 8.1 Total 100.0 100.0 100.0 Main wall material Jute/bamboo/mud (katcha) 30.1 52.5 47.6 Wood 1.2 1.8 1.6 Brick/cement 44.4 9.2 16.9 Tin 24.3 36.5 33.8 Total 100.0 100.0 100.0 Floor material Earth/bamboo (katcha) 52.5 93.7 84.6 Wood 1.5 0.2 0.5 Cement/concrete 45.9 5.9 14.7 Total 100.0 100.0 100.0 Food consumption Deficit in whole year 8.5 13.4 12.3 Sometimes deficit 29.5 38.6 36.6 Neither deficit nor surplus 42.9 35.2 36.9 Surplus 19.1 12.8 14.2 Total 100.0 100.0 100.0 Number of households 2,306 8,194 10,500 22 │ Household Population and Housing Characteristics The type of material used for flooring is an indicator of economic standing of the household as well as an indicator of potential exposure to disease-causing agents. The most commonly used floor materials in Bangladesh are earth or bamboo. The other floor materials used in the country are cement or concrete. Forty-six percent of urban households have cement floors; earth flooring is almost universal in rural areas (94 percent). In order to get a measure of a household’s economic vulnerability, the Household Questionnaire asked respondents whether they thought their household was a surplus or deficit household in terms of food consumption. One in eight households seems to experience food deficit throughout the year, while almost half of the households face food deficit at least some time in the year. Rural households are more likely than urban households to face food deficit at least sometimes in the year (52 versus 38 percent). Since 1999-2000, economic vulnerability, measured in terms of food deficit faced by the households, seems to have decreased. About 18 percent of respondents in 1999-2000 reported that their household faced food shortages throughout the year, compared with 12 percent in 2004. 2.6 HOUSEHOLD POSSESSIONS Information on possession of durable goods, a homestead, and land is presented in Table 2.9. Ownership of a radio or television is a measure of access to mass media; telephone ownership measures access to efficient communication; and bicycle and motorcycle ownership is a measure of access to means of transportation. In general, ownership of these items has a bearing on households’ access to information and health; along with other data, this information can also be used to rank households according to economic status. Table 2.9 Household durable goods and land ownership Percentage of households possessing various durable consumer goods, and ownership of land, according to residence, Bangladesh 2004 Residence Ownership Urban Rural Total Durable goods Almirah or wardrobe 46.9 24.1 29.1 Table 65.5 61.5 62.3 Chair/bench 67.9 63.7 64.7 Watch/clock 79.3 62.2 66.0 Cot/bed 93.5 91.1 91.7 Radio 36.5 28.6 30.4 Television 49.1 15.5 22.9 Bicycle 18.3 25.8 24.2 Motorcycle 4.0 1.3 1.9 Sewing machine 12.6 3.3 5.4 Telephone 15.6 1.8 4.8 None of the above 4.0 5.7 5.3 Land ownership Owns a homestead 90.2 95.5 94.3 Owns other land 38.5 56.3 52.4 Neither of the above 9.0 4.0 5.1 Number of households 2,306 8,194 10,500 Household Population and Housing Characteristics │ 23 More than 90 percent of households in urban and rural areas own a cot or bed. More than 60 percent of households own a watch or clock, a chair or bench, or a table. Almost 30 percent have an almirah or wardrobe. The urban-rural differentials in ownership of the above durable goods are small except for ownership of a watch or clock, or an almirah, which is markedly higher in urban areas. Thirty percent of households have radios, and 23 percent have a television. As for the two items of means of transportation, 24 percent of households own a bicycle, and 2 percent own a motorcycle. Only 1 in 20 households has a telephone or a sewing machine. Rural households are more likely to own bicycles than are urban households. Not surprisingly, urban households are more likely to own televisions, radios, telephones, motorcycles, and sewing machines. Ownership of almost all durable consumer goods has increased in rural areas since 1999-2000, in urban areas, there has been no increase in ownership of various items and for some items, a smaller proportion of households own the item. For example, ownership of radios in urban areas has decreased from 45 to 37 percent, while it has remained unchanged in rural areas at 29 percent. In rural areas, the proportion of households with television has increased from 10 to 16 percent, whereas television penetration in urban areas is at 49 percent, the same as in 1999-2000. Respondents were also asked about their household’s ownership of a house and other land. Ninety-four percent of households own a homestead, and slightly more than half own land other than a homestead. Ownership of a homestead or land is less common in urban areas than in rural areas. In the past four years, ownership of a homestead has increased by almost 10 percentage points from 81 to 90 percent in urban areas and from 88 to 96 percent in rural areas. 2.7 WEALTH INDEX In this report, an index of household economic status was created and used as a background characteristic with information on household ownership of assets and use of selected services.1 The economic status index used here was developed and tested in a large number of countries in relation to inequities in household income, use of health services, and health outcomes (Gwatkin et al., 2000). It is an indicator of the level of wealth that is consistent with expenditure and income measures (Rutstein, 1999). The wealth index was constructed using principal components analysis (Rutstein and Johnson, 2004). Asset information was collected with the 2004 BDHS Household Questionnaire and covered information on household ownership of a number of consumer items, ranging from a television to a bicycle, as well as dwelling characteristics, such as source of drinking water, sanitation facilities, and type of material used for flooring. Each asset was assigned a weight (factor score) generated through principal components analysis, and the resulting asset scores were standardized in relation to a normal distribution with a mean of zero and standard deviation of one (Gwatkin et al., 2000). Each household was then assigned a score for each asset, and the scores were summed for each household; individuals were ranked according to the total score of the household in which they resided. The sample was then divided into quintiles from one (lowest) to five (highest). A single asset index was developed for the whole sample; indexes were not prepared for urban and rural populations. 1 Variables include ownership of items listed in Table 2.9, except homestead; household characteristics in Table 2.8, except food consumption; and whether a household has at least one domestic worker. 24 │ Household Population and Housing Characteristics 2.8 ARSENIC IN HOUSEHOLD DRINKING WATER Arsenic in drinking water is a hazard to human health. It has attracted much attention since recognition in the 1990s of its wide occurrence in well water in Bangladesh. The main source of arsenic in drinking water is arsenic-rich rocks through which the water has filtered. It may also occur because of mining or industrial activity in some areas. In Bangladesh, water is considered to have an unsafe level of arsenic if the measured level of arsenic is equal to or greater than 50 parts per billion (ppb). A public health program has been undertaken throughout Bangladesh to test tubewell water for arsenic levels. The tubewells that have had their water tested are marked either red or green; a red mark means its water has an unsafe level of arsenic, and green implies that the water is not highly contaminated with arsenic. In addition, mass media communication strategies have been used to make communities aware of the importance of avoiding drinking arsenic- contaminated water. The Bangladesh Arsenic Mitigation Water Supply Project (BAMWSP) reports that 29 percent of the 4.9 million tubewells tested using Aqua kit had arsenic contamination at or greater than 50 ppb (BAMWSP). BAMWSP’s indicator, “proportion of tubewells having arsenic-contaminated water” provides only an indication of the proportion of the population exposed to arsenic-contaminated water—it is not an actual measure of the proportion of households consuming arsenic-contaminated water. To provide such an estimate for Bangladesh, specially trained interviewers tested the household drinking water using Hach’s EZ Arsenic Kit.2 Household drinking water was tested in 10,465 of the 10,500 households surveyed. In addition, the survey collected information on respondents’ awareness of the significance of arsenic in water and of the meaning of the color marking on the tubewells for those obtaining drinking water from a tubewell. The results show that, overall, 84 percent of household respondents had heard about the problem of arsenic in water. This included 75 to 78 percent of households in Barisal, Sylhet, and Rajshahi divisions and more than 90 percent in Dhaka and Khulna divisions (Figure 2.4). The awareness of the problem of arsenic in water increases from 69 percent among households in the lowest wealth quintile to 97 percent among those in the highest quintile. 2 Respondents to the Household Questionnaire were asked to provide a glass of the water that the household uses for drinking. Interviewers poured 50 milliliters of that water into a special testing vessel, added two reagents in the prescribed order, and quickly closed the vessel with a lid to which a testing strip was attached. Twenty minutes later, the testing strip was removed and matched with a color chart to determine the level of arsenic in the water. Household Population and Housing Characteristics │ 25 Table 2.10 shows that one in twelve households had elevated arsenic levels (equal to or greater than 50 ppb) in their drinking water on the day of the interview. Overall, households in Chittagong division, followed by Sylhet division, are most likely to have arsenic-contaminated drinking water (22 and 12 percent, respectively). Arsenic levels are not associated with wealth index quintile. Although households in the highest wealth quintile are least likely to drink water with high arsenic levels, those in the fourth quintile are most likely to have arsenic in their drinking water. According to Table 2.11, households with higher levels of arsenic in drinking water are more likely to be aware of the problem. 26 │ Household Population and Housing Characteristics Table 2.10 Level of arsenic in household drinking water Percent distribution of households by level of arsenic in drinking water, according to background characteristics, Bangladesh 2004 Level of arsenic (parts per billion) Background characteristic <50 ≥50 Total Number of households tested for arsenic Residence Urban 93.1 6.9 100.0 2,293 Rural 91.1 8.9 100.0 8,173 Division Barisal 98.7 1.3 100.0 631 Chittagong 78.0 22.0 100.0 1,811 Dhaka 92.3 7.7 100.0 3,357 Khulna 93.1 6.9 100.0 1,288 Rajshahi 97.7 2.3 100.0 2,748 Sylhet 88.5 11.5 100.0 630 Source of drinking water1 Piped into residence 99.2 0.8 100.0 535 Piped outside residence 98.1 1.9 100.0 189 Tubewell 90.4 9.6 100.0 9,089 Shallow tubewell (89.1) (10.9) 100.0 27 Deep tubewell 99.1 0.9 100.0 317 Surface well, pond, lake, river or stream 99.1 0.9 100.0 308 Wealth index Lowest 90.7 9.3 100.0 2,360 Second 92.1 7.9 100.0 2,203 Middle 90.6 9.4 100.0 2,021 Fourth 90.3 9.7 100.0 1,951 Highest 93.9 6.1 100.0 1,930 Heard of arsenic Yes 90.3 9.7 100.0 8,781 No 97.9 2.1 100.0 1,664 Total 91.5 8.5 100.0 10,465 Note: Total includes 20 households missing information on “heard of arsenic”; they are not shown separately. Figures in parentheses are based on 25-49 unweighted cases. 1 Water that was used for arsenic testing Table 2.11 presents data on households that obtain their drinking water from tubewells. Six out of ten of these households reported that the tubewells they use for drinking water are not marked. A puzzling finding was that almost half (45 percent) of the households that obtain drinking water from a red-marked tubewell did not have unsafe levels of arsenic when their drinking water was tested. This may indicate that arsenic levels can vary over time, or that there was some error in the original marking of the tubewell, or that there were measurement errors associated with the use of the test kits. Household Population and Housing Characteristics │ 27 Table 2.11 Arsenic levels and knowledge of arsenic and markings on tubewells Percent distribution of households that obtain their drinking water from tubewells by level of arsenic in water, according to whether they had heard of arsenic in drinking water and the reported marking on the tubewell used by the household for drinking water, Bangladesh 2004 Level of arsenic (ppb) Knowledge of arsenic/ marking on tubewell <50 ≥50 Total Number of households tested Heard of arsenic Yes 89.3 10.7 100.0 7,890 No 97.7 2.3 100.0 1,517 Marking on tubewell Red 45.2 54.8 100.0 614 Green 94.0 6.0 100.0 2,962 Not marked 93.6 6.4 100.0 5,611 Don’t know 95.8 4.2 100.0 241 Total 90.7 9.3 100.0 9,427 Note: Table includes 19 households with missing information on awareness of arsenic in drinking water. ppb = parts per billion The results of the 2004 BDHS and BAMWSP water testing efforts suggest a need to explore the following possibilities: (1) Awareness of unsafe levels of arsenic in drinking water may have caused some households to shift to safer sources of water. (2) Arsenic levels in water may be different when measured directly from the source and after storage of the same water. It is thought that iron in water may reduce the arsenic in stored water. (The International Centre for Diarrhoeal Disease Research, Bangladesh [ICDDR,B] is currently testing this hypothesis in Matlab.) (3) Measurements of arsenic levels in water vary over time. Tests of water from the same source can show different levels of arsenic when measured on different dates; results in Table 2.11 indicate some support for this statement. Characteristics of Survey Respondents │ 29 CHARACTERISTICS OF SURVEY RESPONDENTS 3 3.1 BACKGROUND CHARACTERISTICS OF RESPONDENTS Information on the basic characteristics of women and men interviewed in the survey is essential for the interpretation of findings presented later in the report. The 2004 BDHS interviewed 11,440 ever- married women under the age of 50 years, and 4,297 men in the age group 15-54 for information on individual respondents. Table 3.1.1 shows the distribution of ever-married and currently married women respondents by various background characteristics. More than nine in ten (93 percent) ever-married women under 50 years of age are currently married; almost three in four women reported having a marriage certificate. The age distribution of ever-married women is very similar to that found in the last three BDHS surveys; a little more than half are under 30 years of age. Twenty-three percent of the women live in urban areas. The distribution of female respondents by division is similar to that of the 1999-2000 BDHS. Nearly one-third live in Dhaka and about one-quarter in Rajshahi. Eighteen percent of the women respondents live in Chittagong and 12 percent in Khulna. The proportion of respondents from Sylhet and Barisal are smaller─6 percent from each division. There has been an increase in women’s educational attainment. The percentage of ever-married women with at least some secondary education has increased from 18 percent in 1996-1997 to 29 percent in 2004. On the other hand, the 2004 BDHS shows that about 40 percent of female respondents still have no education. Nine out of ten women are Muslim, with most of the remainder being Hindu. The 2004 BDHS interviewed both never-married and ever-married men in the age group 15-54 by randomly selecting one eligible man in every other household included in the sample for the 2004 survey. The 2004 survey is the first of the four BDHS surveys to include never-married men in the sample. About one-third of the male respondents in the 2004 BDHS have never married. The distribution of the male respondents by background characteristics is presented in Table 3.1.2. About 25 percent of the ever- married male respondents are in age group 15-29, compared with 95 percent of the never-married males. Less than 2 percent of the ever-married males are divorced, separated or widowed, compared to nearly 8 percent of the female respondents. The urban-rural distribution of ever-married males is similar to the urban-rural distribution of female respondents. However, the proportion of never-married men living in urban areas is slightly higher (26 percent) than the proportion of ever-married men (23 percent). The distribution of ever-married male respondents by administrative division shows that Dhaka (31 percent) and Rajshahi (29 percent) have the largest proportions of respondents, while Sylhet and Barisal have the smallest proportions. The distribution by division of never-married men is slightly different. The largest proportion of never- married male respondents is in Dhaka (31 percent), followed by Chittagong (21 percent) and Rajshahi (20 percent). About three in ten ever-married men have at least some secondary education, while another three in ten have no education. The educational attainment of never-married men is higher than ever-married men, perhaps because most of the never-married men belong to a relatively younger cohort; never- married men are almost twice as likely to have at least some secondary education compared to ever- married men. The distribution of male respondents by religion is similar to that of the women respondents; nine out of ten respondents are Muslims. 30 │ Characteristics of Survey Respondents Table 3.1.1 Background characteristics of respondents: women Percent distribution of ever-married women and currently married women age 10-49, by selected background characteristics, Bangladesh 2004 Ever-married women Currently married women Background characteristic Weighted percent Weighted number Unweighted number Weighted percent Weighted number Unweighted Number Age 10-14 1.3 150 140 1.4 145 136 15-19 14.0 1,598 1,563 14.5 1,536 1,498 20-24 19.2 2,202 2,202 20.0 2,121 2,112 25-29 17.6 2,013 2,012 18.3 1,935 1,929 30-34 15.7 1,793 1,783 15.9 1,683 1,673 35-39 12.7 1,457 1,480 12.4 1,309 1,330 40-44 10.1 1,160 1,185 9.3 982 1,005 45-49 9.3 1,066 1,075 8.2 870 870 Marital status Married 92.5 10,582 10,553 100.0 10,582 10,553 Divorced, separated or widowed 7.5 858 887 na na na Has marriage certificate Yes 73.1 8,360 8,234 73.8 7,810 7,675 No 26.9 3,073 3,198 26.2 2,768 2,874 Residence Urban 22.6 2,586 3,904 22.4 2,372 3,578 Rural 77.4 8,854 7,536 77.6 8,210 6,975 Division Barisal 6.3 719 1,360 6.4 674 1,271 Chittagong 17.8 2,041 2,069 17.7 1,877 1,898 Dhaka 31.2 3,570 2,589 31.3 3,315 2,399 Khulna 12.2 1,397 1,708 12.2 1,296 1,587 Rajshahi 26.2 2,994 2,564 26.3 2,782 2,376 Sylhet 6.3 719 1,150 6.0 638 1,022 Education No education 41.2 4,713 4,419 39.6 4,187 3,896 Primary incomplete 20.5 2,348 2,327 20.6 2,176 2,152 Primary complete 8.8 1,011 1,054 9.1 958 996 Secondary incomplete 22.2 2,541 2,648 23.2 2,457 2,547 Secondary complete or higher 7.2 827 992 7.6 804 962 Religion Islam 90.1 10,310 10,182 90.0 9,524 9,389 Hinduism 9.0 1,032 1,192 9.1 961 1,100 Buddhism 0.1 12 13 0.1 11 12 Christianity 0.7 80 47 0.8 80 46 Other/none/missing 0.1 6 6 0.1 6 6 Total 100.0 11,440 11,440 100.0 10,582 10,553 na = Not applicable Characteristics of Survey Respondents │ 31 Table 3.1.2 Background characteristics of respondents: men Percent distribution of ever-married men and never-married men age 15-54, by selected background characteristics, Bangladesh 2004 Ever-married men Never-married men Background characteristic Weighted percent Weighted number Unweighted number Weighted percent Weighted number Unweighted Number Age 15-19 1.0 28 30 53.9 794 586 20-24 8.0 227 237 29.4 433 336 25-29 14.7 416 431 11.7 173 150 30-34 17.7 499 604 3.6 53 46 35-39 16.1 456 582 1.3 19 11 40-44 17.4 491 579 0.1 1 1 45-49 16.4 464 474 0.1 1 2 50-54 8.6 242 228 0.0 0 0 Marital status Never married na na na 100.0 1,474 1,132 Married 98.5 2,780 3,133 na na na Divorced, separated or widowed 1.5 43 32 na na na Residence Urban 23.3 657 1,106 26.1 385 408 Rural 76.7 2,166 2,059 73.9 1,090 724 Division Barisal 5.3 148 329 5.2 77 123 Chittagong 15.1 426 474 21.2 313 236 Dhaka 31.2 880 752 31.4 462 260 Khulna 13.8 388 524 12.5 184 174 Rajshahi 29.0 819 803 20.0 295 196 Sylhet 5.7 160 283 9.7 143 143 Education No education 31.2 879 964 12.1 178 124 Primary incomplete 30.1 850 950 20.5 302 243 Primary complete 7.6 213 228 8.9 131 94 Secondary incomplete 18.0 508 576 36.4 537 391 Secondary complete or higher 13.2 372 447 22.2 327 280 Religion Islam 89.4 2,523 2,822 88.3 1,302 999 Hinduism 9.8 276 328 10.3 152 122 Buddhism 0.0 1 2 0.1 2 3 Christianity 0.7 21 11 1.1 16 7 Other/none/missing 0.8 24 15 1.3 20 11 Total 100.0 2,823 3,165 100.0 1,474 1,132 1 na = Not applicable Because the men interviewed in the BDHS survey were selected from a sub-sample of households in which ever-married women were interviewed, it is possible to match married male respondents with their wives to obtain a set of matched couples. Figure 3.1 shows the husband-wife differential in age for 2,533 couples. It is not surprising that for almost all couples, the husband is older than the wife, because it is the tradition in Bangladesh. However, it is noteworthy that the husband is at least ten years older than the wife among 44 percent of the couples, including 13 percent in which the wife is at least 15 years younger than the husband. 32 │ Characteristics of Survey Respondents 3.2 EDUCATIONAL ATTAINMENT Tables 3.2.1 and 3.2.2 present the distributions of female and male respondents, respectively, by the highest level of education completed or attended, according to age, urban-rural residence, division, and wealth index. A large majority of females and a majority of male respondents have not gone beyond the primary level of education. Generally, younger persons have reached higher levels of schooling than older people, as have urban residents. For example, 40 percent of women in urban areas have attended at least some secondary school, compared with 26 percent of rural women. Among the divisions, Barisal division has the largest and Sylhet division the lowest proportion of women and men who have completed secondary school. The educational level of women in Sylhet is worrisome, as a majority of women reported that they did not attend school at all, and only 17 percent of women had attended or completed secondary education. As expected, the level of education increases with the wealth index. For example, among women and men in the lowest quintile, only 6 and 12 percent, respectively, have at least some secondary education, compared with almost 60 percent of women and 70 percent of men in the highest quintile. Characteristics of Survey Respondents │ 33 Table 3.2.1 Level of education by background characteristics: women Percent distribution of ever-married women by highest level of schooling attended or completed, and median number of years of schooling, according to background characteristics, Bangladesh 2004 Highest level of schooling attended or completed Background characteristic No education Primary incomplete Primary complete1 Secondary incomplete Secondary complete or higher2 Total Number of women Median years of schooling Age 10-14 12.3 20.5 18.7 48.4 0.0 100.0 150 4.6 15-19 15.3 22.5 10.7 46.2 5.2 100.0 1,598 4.7 20-24 26.5 19.4 9.7 32.2 12.2 100.0 2,202 4.3 25-29 41.4 21.0 9.2 18.5 9.9 100.0 2,013 1.6 30-34 49.4 22.0 7.7 14.4 6.5 100.0 1,793 0.0 35-39 55.7 21.2 6.7 11.5 5.0 100.0 1,457 0.0 40-44 57.3 17.9 8.0 12.4 4.4 100.0 1,160 0.0 45-49 62.8 18.4 8.1 7.4 3.3 100.0 1,066 0.0 Residence Urban 33.0 18.8 8.0 25.0 15.2 100.0 2,586 3.7 Rural 43.6 21.0 9.1 21.4 4.9 100.0 8,854 1.1 Division Barisal 25.3 27.3 13.3 24.5 9.5 100.0 719 3.7 Chittagong 40.4 18.3 8.8 24.4 8.2 100.0 2,041 2.4 Dhaka 42.1 19.8 9.0 20.3 8.8 100.0 3,570 1.5 Khulna 32.9 23.7 7.7 28.9 6.7 100.0 1,397 3.2 Rajshahi 45.4 20.9 6.9 21.4 5.4 100.0 2,994 0.5 Sylhet 53.0 16.3 13.7 13.5 3.5 100.0 719 0.0 Wealth index Lowest 67.9 21.0 4.8 6.1 0.2 100.0 2,279 0.0 Second 51.9 24.4 7.5 14.9 1.3 100.0 2,290 0.0 Middle 40.1 23.9 9.1 23.6 3.2 100.0 2,267 1.9 Fourth 28.9 19.9 13.0 30.6 7.6 100.0 2,307 4.1 Highest 17.4 13.5 9.6 35.7 23.7 100.0 2,297 6.3 Total 41.2 20.5 8.8 22.2 7.2 100.0 11,440 1.8 1 Primary complete is defined as completing grade 5 2 Secondary complete is defined as completing grade 10 Regarding educational differences between spouses, in half of the couples both have some education, and in one in five couples the wife has no education and neither does the husband (Figure 3.2). As in 1999-2000, for three in ten couples, only one partner is educated. However, there is a slight decrease in the proportion of couples in which only the husband is educated. 34 │ Characteristics of Survey Respondents Table 3.2.2 Level of education by background characteristics: men Percent distribution of men by highest level of schooling attended or completed, and median number of years of schooling, according to background characteristics, Bangladesh 2004 Highest level of schooling attended or completed Background characteristic No education Primary incomplete Primary complete1 Secondary incomplete Secondary complete or higher2 Total Number of men Median years of schooling Age 15-19 12.0 21.6 10.7 43.4 12.5 100.0 822 6.5 20-24 15.0 28.9 8.8 23.4 23.8 100.0 660 4.7 25-29 24.0 24.6 6.5 24.6 20.3 100.0 589 4.2 30-34 29.8 26.6 7.1 21.8 14.8 100.0 552 3.0 35-39 34.6 28.2 7.4 13.8 16.0 100.0 475 1.4 40-44 31.6 31.5 8.1 17.9 10.9 100.0 491 2.1 45-49 33.8 27.7 5.2 18.1 15.2 100.0 465 1.7 50-54 31.9 30.7 9.2 13.0 15.2 100.0 242 2.2 Residence Urban 18.5 22.2 7.5 25.2 26.6 100.0 1,042 6.0 Rural 26.6 28.3 8.2 24.1 12.9 100.0 3,255 3.2 Division Barisal 13.2 31.4 5.7 26.3 23.4 100.0 225 4.9 Chittagong 25.2 26.7 7.6 23.7 16.8 100.0 739 3.6 Dhaka 27.2 23.7 8.7 24.7 15.7 100.0 1,342 3.8 Khulna 18.8 27.5 9.7 26.6 17.4 100.0 573 4.4 Rajshahi 27.6 27.4 6.5 22.4 16.1 100.0 1,114 3.2 Sylhet 20.4 33.4 10.6 25.6 10.0 100.0 304 3.6 Wealth index Lowest 49.5 34.0 4.4 10.5 1.6 100.0 717 0.0 Second 33.4 34.7 7.8 18.4 5.6 100.0 869 1.5 Middle 22.6 29.3 11.6 25.6 10.9 100.0 846 3.8 Fourth 16.5 24.0 8.0 32.9 18.6 100.0 892 6.1 Highest 7.5 14.9 7.8 30.8 38.9 100.0 973 9.3 Total 24.6 26.8 8.0 24.3 16.2 100.0 4,297 3.8 1 Primary complete is defined as completing grade 5 2 Secondary complete is defined as completing grade 10 Characteristics of Survey Respondents │ 35 3.3 EXPOSURE TO MASS MEDIA Information access is essential to increase people’s knowledge and awareness of what is taking place around them, which may affect their perceptions and behavior. In the survey, exposure to media was assessed by asking respondents how frequently they read a newspaper, watched television, or listened to a radio. It is important to know the subgroups that are more or less likely to be reached by the media for purposes of planning programs intended to spread information about health and family planning. Tables 3.3.1 and 3.3.2 show the percentage of female and male respondents, respectively, who are exposed to different types of mass communication media, by background variables. Table 3.3.1 Exposure to mass media: women Percentage of ever-married women who usually read a newspaper at least once a week, watch television at least once a week, and listen to the radio at least once a week, by background characteristics, Bangladesh 2004 Background characteristic No media Reads a newspaper at least once a week Watches television at least once a week Listens to the radio at least once a week All three media Number of women Age 10-14 32.4 4.8 49.0 48.1 2.9 150 15-19 31.8 7.7 52.7 44.6 3.6 1,598 20-24 33.7 9.3 53.7 38.7 4.9 2,202 25-29 38.9 7.5 49.3 32.9 3.1 2,013 30-34 44.5 5.8 43.7 30.4 2.1 1,793 35-39 46.6 5.5 40.4 28.7 1.9 1,457 40-44 50.5 6.2 38.0 27.0 2.2 1,160 45-49 59.4 3.7 29.1 22.8 1.2 1,066 Residence Urban 21.5 15.4 72.6 33.4 5.5 2,586 Rural 47.7 4.3 37.7 33.4 2.2 8,854 Division Barisal 49.4 7.4 29.2 36.7 3.3 719 Chittagong 41.5 7.4 45.5 35.1 3.4 2,041 Dhaka 36.2 8.1 53.3 33.4 3.0 3,570 Khulna 38.0 6.6 48.9 37.2 3.1 1,397 Rajshahi 46.2 4.8 40.9 31.4 2.2 2,994 Sylhet 51.4 7.4 37.3 25.6 3.8 719 Education No education 58.9 0.0 30.1 23.4 0.0 4,713 Primary incomplete 43.3 0.7 43.0 34.2 0.3 2,348 Primary complete 33.7 3.4 50.8 38.4 1.6 1,011 Secondary incomplete 22.8 13.3 63.0 45.5 6.7 2,541 Secondary complete or higher 7.4 47.7 81.1 44.8 17.2 827 Wealth index Lowest 71.7 0.3 19.4 16.2 0.1 2,279 Second 56.7 1.1 28.9 28.5 0.5 2,290 Middle 41.9 2.9 38.7 37.4 1.3 2,267 Fourth 27.2 6.5 58.2 44.7 3.3 2,307 Highest 11.7 23.4 82.3 39.9 9.5 2,297 Total 41.8 6.8 45.6 33.4 2.9 11,440 36 │ Characteristics of Survey Respondents Table 3.3.2 Exposure to mass media: men Percentage of men who usually read a newspaper at least once a week, watch television at least once a week, and listen to the radio at least once a week, by background characteristics, Bangladesh 2004 Background Characteristic No media Reads a newspaper at least once a week Watches television at least once a week Listens to the radio at least once a week All three media Number of men Age 15-19 8.1 37.8 82.6 62.7 20.3 822 20-24 8.9 42.2 83.2 58.1 21.1 660 25-29 13.8 34.0 75.1 52.9 16.7 589 30-34 18.9 32.7 69.5 49.2 17.3 552 35-39 19.5 28.9 67.9 46.8 14.7 475 40-44 22.5 30.8 63.7 44.3 14.4 491 45-49 31.5 25.3 55.4 43.7 12.9 465 50-54 32.4 28.8 48.8 40.1 12.9 242 Residence Urban 9.3 50.8 84.3 45.8 21.4 1,042 Rural 19.8 28.2 67.2 53.6 15.6 3,255 Division Barisal 20.8 34.6 59.4 58.3 20.5 225 Chittagong 11.9 35.7 79.1 53.8 15.9 739 Dhaka 19.2 30.9 68.9 49.3 15.0 1,342 Khulna 15.3 44.6 74.8 56.9 27.0 573 Rajshahi 19.2 29.3 69.0 49.6 14.2 1,114 Sylhet 15.0 35.8 74.2 50.6 17.8 304 Education No education 35.3 0.2 51.9 42.6 0.1 1,057 Primary incomplete 21.8 11.4 65.9 51.6 5.5 1,152 Primary complete 10.1 42.8 76.9 55.9 18.3 345 Secondary incomplete 5.9 53.3 84.7 57.6 28.7 1,045 Secondary complete or higher 2.7 87.0 86.9 55.0 43.7 698 Wealth index Lowest 35.8 9.0 48.9 43.9 4.2 717 Second 24.7 15.7 61.9 48.0 8.0 869 Middle 16.7 30.4 68.2 55.3 13.6 846 Fourth 10.4 39.2 79.9 57.4 23.5 892 Highest 3.5 65.7 91.1 52.5 31.5 973 Total 17.2 33.7 71.3 51.7 17.0 4,297 For both men and women, exposure to television is higher than exposure to radio or newspapers. Exposure to any media varies by gender; men are more likely to be exposed to any type of mass media than women (Figure 3.3). Seven percent of women and 34 percent of men read newspapers at least once a week, 46 percent of women and 71 percent of men watch television at least once a week, and 33 percent of women and 52 percent of men listen to the radio once a week. Only 3 percent of women and 17 percent of men are exposed to all three of these media sources. Forty-two percent of women and 17 percent of men have no access to mass media. Since 1999-2000, the proportion of women listening to the radio has increased moderately but it is still below the 1996-97 survey level. However, the exposure to television continues to increase substantially during the same period─around 30 percent during each interval between the surveys. On the other hand, in spite of progress in women’s education, the proportion of ever-married women who reported reading newspapers has declined. Characteristics of Survey Respondents │ 37 Although listening to the radio is more common among rural men than urban men, it is no surprise that urban residents are more likely to have access to television than rural residents. Exposure to television is found to be the least in Barisal division compared with the other administrative divisions. Listening to the radio and watching television are both negatively related to respondent’s age. Exposure to media is positively associated with educational attainment; the proportion with access to all three media outlets increases with increasing educational level of respondents. Similarly, access to all three media outlets increases as wealth increases for both sexes. 3.4 EMPLOYMENT 3.4.1 Employment Status The BDHS asked respondents whether they were employed at the time of the survey and, if not, whether they were employed in the 12 months preceding the survey. Tables 3.4.1 and 3.4.2 show that 22 percent of ever-married women and 87 percent of men are currently employed. Most men and women who are employed work year-round. 38 │ Characteristics of Survey Respondents Table 3.4.1 Employment status: women Percent distribution of ever-married women by employment status and continuity of employment, according to background characteristics, Bangladesh 2004 Currently employed Background characteristic Not employed Work all year Work seasonally Work occasionally Total Number of women Age 10-14 95.6 4.2 0.2 0.0 100.0 150 15-19 90.1 7.1 1.7 1.0 100.0 1,598 20-24 82.9 12.9 2.4 1.6 100.0 2,202 25-29 74.8 19.0 4.0 2.1 100.0 2,013 30-34 71.0 23.4 3.7 1.9 100.0 1,793 35-39 71.6 22.5 4.2 1.7 100.0 1,457 40-44 71.8 23.3 3.6 1.2 100.0 1,160 45-49 77.4 18.1 2.7 1.7 100.0 1,066 Residence Urban 74.4 21.6 2.2 1.9 100.0 2,586 Rural 78.7 16.3 3.4 1.6 100.0 8,854 Division Barisal 86.8 9.5 2.1 1.4 100.0 719 Chittagong 84.3 11.6 3.1 1.0 100.0 2,041 Dhaka 77.3 18.3 2.9 1.5 100.0 3,570 Khulna 76.3 17.2 3.7 2.8 100.0 1,397 Rajshahi 71.6 23.0 3.6 1.8 100.0 2,994 Sylhet 80.7 15.7 2.5 1.1 100.0 719 Education No education 70.7 22.8 4.6 1.9 100.0 4,713 Primary incomplete 78.5 15.8 3.1 2.5 100.0 2,348 Primary complete 85.4 11.3 2.2 1.2 100.0 1,011 Secondary incomplete 86.4 11.0 1.6 1.0 100.0 2,541 Secondary complete or higher 79.7 19.2 0.7 0.4 100.0 827 Wealth index Lowest 70.5 20.4 6.5 2.6 100.0 2,279 Second 73.6 20.6 3.7 2.1 100.0 2,290 Middle 78.3 17.7 2.8 1.2 100.0 2,267 Fourth 83.2 13.7 1.8 1.2 100.0 2,307 Highest 82.9 15.1 1.0 1.0 100.0 2,297 Total 77.7 17.5 3.1 1.6 100.0 11,440 The proportion currently employed among both men and women increases with age, and then plateaus after age 30. The lower employment among men under 25 is probably because many of them are enrolled in schools and colleges. Characteristics of Survey Respondents │ 39 Table 3.4.2 Employment status: men Percent distribution of men by employment status and continuity of employment, according to background characteristics, Bangladesh 2004 Currently employed Background characteristic Not employed Work all year Work seasonally Work occasionally Total Number of men Age 15-19 40.7 47.3 7.6 4.4 100.0 822 20-24 20.8 67.8 8.5 2.9 100.0 660 25-29 6.7 80.5 9.6 3.2 100.0 589 30-34 3.0 86.9 8.1 2.0 100.0 552 35-39 0.9 89.0 8.7 1.2 100.0 475 40-44 1.8 89.7 7.3 1.2 100.0 491 45-49 2.8 88.9 7.4 0.9 100.0 465 50-54 0.9 91.5 6.2 1.3 100.0 242 Residence Urban 15.5 77.7 3.8 2.9 100.0 1,042 Rural 12.1 76.2 9.4 2.3 100.0 3,255 Division Barisal 16.3 74.4 6.4 2.9 100.0 225 Chittagong 15.4 72.3 10.5 1.8 100.0 739 Dhaka 14.3 75.2 7.3 3.2 100.0 1,342 Khulna 12.4 79.0 7.7 0.8 100.0 573 Rajshahi 9.1 80.3 7.8 2.8 100.0 1,114 Sylhet 13.7 76.0 8.3 1.9 100.0 304 Education No education 2.3 87.2 8.4 2.2 100.0 1,057 Primary incomplete 3.0 83.4 10.3 3.2 100.0 1,152 Primary complete 5.5 82.7 11.0 0.9 100.0 345 Secondary incomplete 25.3 64.5 7.1 3.1 100.0 1,045 Secondary complete or higher 30.7 64.2 3.9 1.2 100.0 698 Wealth index Lowest 3.7 84.1 10.3 1.9 100.0 717 Second 6.4 79.9 9.1 4.6 100.0 869 Middle 10.4 77.3 9.5 2.6 100.0 846 Fourth 19.2 71.3 7.8 1.7 100.0 892 Highest 22.1 72.2 4.4 1.4 100.0 973 Total 12.9 76.5 8.1 2.4 100.0 4,297 There are some divisional variations in the proportion currently employed. Men and women in Rajshahi division are most likely to be employed (91 and 28 percent, respectively), while those in Barisal are least likely to be employed; women in Barisal are only half as likely to be employed as their counterparts in Rajshahi. Current employment for men shows a negative relation with educational level but for women the relationship of education to employment is not clear. Men with primary or lower education are generally fully employed compared with 75 percent or fewer of those who have attended secondary or higher education. The proportion currently employed decreases as the wealth status of respondents increases; however, the variations in employment by household economic status are much larger for men than for women. 40 │ Characteristics of Survey Respondents 3.4.2 Control over Women’s Earnings Women who were working and receiving cash earnings were asked who decides how their earnings are used. Out of 2,550 women who were currently employed, 2,291 earned cash. Table 3.5 shows that slightly less than half (47 percent) of working women decide by themselves how their cash earnings are used, while 40 percent make the decision jointly with someone else. One in eight women is not involved in making decisions on how her earnings are used. Table 3.5 Decision on use of earnings Percent distribution of currently employed, ever-married-women who receive cash earnings by person who decides how earnings are to be used, according to background characteristics, Bangladesh 2004 Person who decides how earnings are used Background characteristic Respondent only Respondent jointly with others1 Someone else only2 Missing Total Number of women Age 15-19 39.2 40.6 20.2 0.0 100.0 134 20-24 42.1 43.9 14.0 0.0 100.0 344 25-29 41.3 44.6 14.1 0.0 100.0 461 30-34 44.9 40.4 14.7 0.0 100.0 474 35-39 52.0 37.2 10.0 0.7 100.0 362 40-44 58.8 34.2 7.0 0.0 100.0 301 45-49 52.1 39.6 8.4 0.0 100.0 211 Marital status Married 39.3 46.5 14.2 0.0 100.0 1,910 Divorced/separated/widowed 85.7 9.5 4.2 0.7 100.0 382 Number of living children 0 47.5 36.4 16.1 0.0 100.0 193 1-2 48.4 40.6 10.7 0.2 100.0 1,035 3-4 44.7 41.2 14.0 0.0 100.0 773 5+ 47.5 40.0 12.5 0.0 100.0 290 Residence Urban 54.6 36.6 8.9 0.0 100.0 628 Rural 44.1 41.8 13.9 0.2 100.0 1,663 Division Barisal 48.4 33.5 18.1 0.0 100.0 85 Chittagong 52.5 35.5 12.0 0.0 100.0 285 Dhaka 51.7 37.3 10.6 0.4 100.0 719 Khulna 55.0 35.0 10.0 0.0 100.0 298 Rajshahi 38.3 49.1 12.6 0.0 100.0 776 Sylhet 41.0 32.9 26.1 0.0 100.0 127 Education No education 47.8 37.1 14.9 0.2 100.0 1,221 Primary incomplete 44.5 47.3 8.2 0.0 100.0 443 Primary complete 42.4 47.0 10.6 0.0 100.0 138 Secondary incomplete 46.5 40.5 13.0 0.0 100.0 322 Secondary complete or higher 52.7 40.0 7.3 0.0 100.0 167 Wealth index Lowest 46.4 37.5 16.2 0.0 100.0 568 Second 42.4 44.8 12.3 0.5 100.0 543 Middle 40.6 45.1 14.2 0.0 100.0 437 Fourth 54.1 36.4 9.6 0.0 100.0 360 Highest 55.0 36.7 8.2 0.0 100.0 383 Total 47.0 40.4 12.5 0.1 100.0 2,291 Note: Total includes five women age 10-14, who are not shown separately. 1 Husband or someone else 2 Includes husband Characteristics of Survey Respondents │ 41 Table 3.5 also shows how women’s degree of control over their earnings varies by background characteristics. Irrespective of age, most respondents make their own decisions (either alone or jointly with others) on how their cash earnings are used. However, the sole control over earnings generally increases and having no control at all over earnings declines, with increasing age. Currently married women, compared with other ever-married women, are much more likely to involve another person in making the decision or to have no say at all in decisions about the use of her earnings. Urban women are more independent than rural women in making their own decisions without involving anyone else (55 and 44 percent, respectively). Comparatively, few women in both rural (14 percent) and urban areas (9 percent) report that they are not involved in decisions as to the use of their earnings. There are variations by divisions in the way decisions are made on how women’s earnings are used. The percentage of women who make decisions on their earnings by themselves alone ranges from 52-55 percent in Dhaka, Chittagong, and Khulna divisions, to 38 percent in Rajshahi division. There are no clear patterns by education or wealth index, although the percentages of women deciding about their earnings are higher for the two highest wealth quintiles and also higher for women who have completed secondary education; 50 percent or more of these women have complete control over their earnings. In the 2004 BDHS, each man whose wife earned cash was asked who decides how her earnings are used. Table 3.6 shows that according to husbands, two-thirds of couples jointly decide how her Table 3.6 Decision on use of wife’s earnings Percent distribution of currently married men whose wives earn cash for work by person who decides how her earnings are to be used, according to background characteristics, Bangladesh 2004 Person who decides how wife’s earnings are used, according to husband Background characteristic Husband only Wife only Husband jointly with wife Total Number of men Age 20-24 (5.9) (19.1) (75.0) 100.0 22 25-29 (14.2) (26.5) (59.2) 100.0 41 30-34 16.0 16.2 67.8 100.0 76 35-39 11.9 30.0 58.1 100.0 70 40-44 15.2 28.3 56.5 100.0 84 45-49 8.0 14.5 77.5 100.0 85 50-54 6.0 25.2 68.8 100.0 50 Residence Urban 10.1 26.1 63.9 100.0 104 Rural 12.6 21.6 65.8 100.0 326 Division Barisal (9.7) (10.3) (80.1) 100.0 19 Chittagong 9.0 24.5 66.5 100.0 43 Dhaka 8.8 19.6 71.6 100.0 142 Khulna 7.4 22.7 69.9 100.0 58 Rajshahi 12.4 26.6 61.0 100.0 148 Sylhet (53.5) (23.3) (23.2) 100.0 20 Education No education 14.1 25.2 60.7 100.0 145 Primary incomplete 16.3 16.2 67.5 100.0 130 Primary complete (11.5) (19.9) (68.6) 100.0 38 Secondary incomplete 7.3 19.1 73.5 100.0 58 Secondary complete or higher 1.9 36.0 62.1 100.0 59 Total 12.0 22.7 65.4 100.0 430 Note: Based on data obtained from male respondents. Total includes one man age 15-19, who is not shown separately. Figures in parentheses are based on 25-49 men. 42 │ Characteristics of Survey Respondents earnings are used. One in eight husbands says that he has control over her earnings. Husbands age 35-49 and those most educated are more likely to report that their wives by themselves decide how their earnings are used. Some of these results are inconsistent with what married women report (see Table 3.5). For example, according to husbands one in four wives decides by herself how her earnings are used compared with almost four in ten married women who say that they decide without anyone else’s participation. 3.5 WOMEN’S EMPOWERMENT In addition to information on women’s education, employment status, and control over earnings, the 2004 BDHS survey collected information from women and men on other measures of women’s autonomy and status. In particular, questions were asked about women’s roles in making household decisions. Such information provides insight into women’s control over their environment and their attitudes towards gender roles, both of which are relevant to understanding women’s demographic and health behavior. 3.5.1 Women’s Participation in Decisionmaking: Women’s Perspective To assess women’s decisionmaking autonomy, the 2004 BDHS sought information on women’s participation in six different types of household decisions: on the respondents’ own health care; on child health care; on making large household purchases; on making household purchase for daily needs; on visits to family or relatives; and on what food to cook each day. Table 3.7 shows the percent distribution of women according to who in the household usually has the final say on each aspect, according to current marital status. Table 3.7 Women’s participation in household decisionmaking Percent distribution of women by person who has the final say in making specific decisions, according to current marital status and type of decision, Bangladesh 2004 Currently married women Formerly married women1 Household decision Respondent only Respondent jointly with husband Respondent jointly with someone else Husband only Someone else only Total Respondent Respondent jointly with someone else Someone else only Missing Total Own health care 17.6 23.5 4.6 47.9 6.4 100.0 58.3 13.5 28.2 0.0 100.0 Child health care 17.4 32.7 4.9 31.8 5.6 100.0 54.1 11.9 19.9 14.2 100.0 Large household purchases 8.8 38.8 8.0 34.2 10.3 100.0 46.7 23.1 30.1 0.0 100.0 Daily household purchases 18.1 30.4 7.0 34.0 10.4 100.0 50.0 18.8 31.0 0.1 100.0 Visits to family or relatives 12.4 36.6 6.8 35.7 8.5 100.0 55.6 21.1 23.2 0.1 100.0 What food to cook each day 74.2 4.6 8.8 3.4 8.9 100.0 60.5 16.9 22.4 0.1 100.0 Note: Table is based on 10,582 currently married women and 858 formerly married women. 1 Divorced, separated or widowed women Characteristics of Survey Respondents │ 43 Among currently married women, few women independently make decisions on any of the aspects asked about, except on the decision on what food to cook. In fact, a large proportion of women have no final say even on decisions that concern their own health (54 percent) or their child’s health (37 percent). Close to half of the women also said that they have no say on large or daily household purchases, and on whether they could visit family members. Ever-married women who are currently not married have slightly less control over what food to cook than married women. However, they have much more independence in other decisionmaking. More than 70 percent are involved in decisions on their own health care, with 58 percent deciding alone and 14 percent deciding with someone else. The other decisions are also made mostly by either the respondents themselves (47 to 56 percent) or jointly with someone else. Interestingly, each of the six decisions is made by someone else alone for 20-30 percent of formerly married women. In Table 3.8, the autonomy of currently married women can be gauged either by their independently making such decisions or jointly deciding on such issues. The table shows that 28 percent of women have a say in all the six areas of decisionmaking, while 7 percent have no say at all in any of the specified areas. Generally, women’s participation in making all of the specified decisions is higher in urban areas than in rural areas (35 percent as compared to 26 percent) and increases with age and also by wealth. For example, the proportion of women who participate in all decisionmaking increases from 12 percent among women age 10-14 years to a peak of 34 percent among those in age group 40-44 years. Women without children, most of whom are young, are least likely to participate in all the six household decisionmaking factors. Women in Sylhet are least likely to have the final say (alone or jointly) on decisions regarding their child’s health, large household purchases or daily household purchases, and visiting relatives, compared with the other divisions. Less than half of the Sylheti women are involved in these decisions. Women in Rajshahi are least likely to make decisions about their own health care—only four in ten women reported that they are involved in the final say. Thirty-one percent of women who are employed for cash participate in making all decisions, compared with 27 percent of unemployed women. Except for deciding about what food to cook, urban women, the highest educated women, women employed for cash, and women in the wealthiest households are generally more likely to participate in final decisions on household matters. Nevertheless, between 43 and 49 percent of these women do not participate in final decisions concerning their own health care. 44 │ Characteristics of Survey Respondents Table 3.8 Women’s participation in decisionmaking by background characteristics Percentage of currently married women who say that they alone or jointly have the final say in specific decisions, by background characteristics, Bangladesh 2004 Alone or jointly have final say in: Background characteristic Own health care Child health care Large household purchases Daily household purchases Visits to family or relatives What food to cook each day All specified decisions No specified decisions Number of women Age 10-14 31.9 22.6 32.4 31.4 31.4 42.3 12.3 44.5 145 15-19 37.0 36.9 45.4 45.0 43.4 70.3 18.1 18.8 1,536 20-24 41.9 52.1 52.0 51.4 51.8 84.3 23.4 8.8 2,121 25-29 47.4 59.6 59.0 59.2 57.7 91.5 30.2 4.8 1,935 30-34 49.3 62.1 59.4 61.1 61.1 93.6 31.3 4.1 1,683 35-39 50.1 63.6 62.8 64.5 64.5 95.1 33.5 2.5 1,309 40-44 52.9 61.5 60.2 58.7 61.8 95.3 34.1 2.1 982 45-49 47.0 55.0 54.1 52.8 56.6 94.3 30.0 3.8 870 Number of living children 0 36.5 19.5 42.2 42.7 43.8 64.0 11.4 23.8 1,246 1-2 45.7 59.4 56.7 56.4 56.2 87.2 28.4 6.9 4,542 3-4 48.4 61.0 60.1 60.2 59.8 94.0 31.6 3.6 3,303 5+ 47.5 57.7 52.9 53.4 55.5 94.6 31.1 3.8 1,490 Residence Urban 53.2 61.7 64.1 64.3 65.2 87.3 34.5 6.6 2,372 Rural 43.5 53.0 53.0 53.0 53.0 87.7 25.8 7.7 8,210 Division Barisal 47.7 56.4 54.2 55.5 57.2 84.0 29.7 9.7 674 Chittagong 50.7 59.4 58.1 59.0 59.7 87.4 31.7 7.3 1,877 Dhaka 46.9 54.5 55.7 55.9 58.0 87.6 29.3 7.8 3,315 Khulna 49.6 58.5 57.0 55.7 54.3 86.0 29.1 8.2 1,296 Rajshahi 39.3 52.0 54.8 54.3 52.1 91.0 22.7 5.0 2,782 Sylhet 42.6 48.5 48.1 48.9 49.9 81.3 25.4 12.2 638 Education No education 43.3 52.8 53.4 53.7 53.5 92.1 26.5 5.5 4,187 Primary incomplete 46.5 57.6 56.7 56.5 55.3 89.4 29.1 6.4 2,176 Primary complete 47.2 56.0 60.4 60.6 60.0 86.3 29.6 7.6 958 Secondary incomplete 45.0 53.6 52.7 52.4 54.2 80.9 25.9 11.2 2,457 Secondary complete or higher 56.6 62.5 66.2 66.6 68.4 81.9 34.4 8.5 804 Current employment Not employed 44.4 53.6 53.7 53.5 54.2 86.4 26.9 8.3 8,455 Employed for cash 51.2 61.5 63.2 64.2 61.8 93.5 31.1 3.2 1,910 Employed not for cash 45.4 51.1 59.3 58.9 62.2 84.7 30.2 9.0 213 Wealth index Lowest 42.6 50.1 51.9 52.0 50.8 90.4 24.7 6.7 2,042 Second 42.4 53.3 53.5 52.8 51.9 89.5 23.9 6.1 2,112 Middle 43.9 54.2 53.1 53.7 53.7 87.3 27.0 8.2 2,112 Fourth 46.2 54.7 55.3 54.7 57.0 86.4 28.0 7.8 2,168 Highest 53.1 62.2 63.5 64.4 65.0 84.7 34.9 8.2 2,148 Total 45.7 55.0 55.5 55.6 55.7 87.6 27.8 7.4 10,582 Note: Total includes three women with missing information on employment, who are not shown separately. Characteristics of Survey Respondents │ 45 3.5.2 Wife’s Participation in Decisionmaking: Husbands’ Perspective All married men interviewed in the 2004 BDHS survey were asked whether they take their wife’s opinion into account when making decisions on large household expenses, minor daily household expenses, and when they wish to visit family, friends, or relatives. An overwhelming majority of currently married men—almost nine in ten—assert that they take their wife’s opinion into account in deciding about large household expenses and also about visiting family, friends, and relatives (Table 3.9). The wife’s opinion is not as widely sought for making daily household expenses; seven in ten men say that they seek their spouse’s opinion on minor household expenses. Only 5 percent of men said that they do not seek their wife’s opinion when making any of the above decisions. Differentials by all background characteristics for specific decisions are minimal. Table 3.9 Wife’s participation in decisionmaking Percentage of currently married men who take their wife’s opinion into account in making specific decisions, according to background characteristics, Bangladesh 2004 Decision Background characteristic Large household expenses Daily household expenses Visiting family, friends or relatives None of the specified decisions Number of currently married men Age 15-19 (72.1) (64.6) (77.9) (15.6) 28 20-24 81.7 64.9 85.3 7.0 221 25-29 81.3 65.3 87.3 7.5 401 30-34 88.0 68.9 90.4 4.7 493 35-39 89.5 71.6 89.4 5.2 446 40-44 90.9 70.1 91.3 4.8 490 45-49 92.4 79.6 94.9 1.6 462 50-54 90.8 75.4 91.2 3.1 240 Residence Urban 87.1 70.5 89.5 5.7 655 Rural 88.4 71.1 90.5 4.6 2,125 Division Barisal 93.7 64.6 89.9 3.1 146 Chittagong 86.0 68.3 88.8 6.3 422 Dhaka 87.7 74.1 91.0 4.4 866 Khulna 84.5 62.0 90.3 5.9 384 Rajshahi 90.6 73.0 90.3 4.0 805 Sylhet 86.4 78.5 90.0 6.1 157 Education No education 85.8 72.4 86.4 6.8 866 Primary incomplete 87.2 71.1 92.4 4.0 831 Primary complete 93.2 68.3 93.2 2.8 209 Secondary incomplete 87.7 68.3 90.7 4.9 504 Secondary complete or higher 93.0 72.8 91.8 3.2 370 Total 88.1 71.0 90.2 4.8 2,780 Note: Figures in parentheses are based on 25-49 men. 3.5.3 Freedom of Movement Table 3.10 gives information on another dimension of women’s autonomy measured in this BDHS survey: women’s freedom of movement. Currently married women were asked whether they go alone or can go alone outside the village/town/city and to a health center or hospital. The data show that 46 │ Characteristics of Survey Respondents 31 percent of women say that they go or can go alone outside village/town/city and the exactly the same proportion state that they go or can go alone to the hospital or health center. Seventeen percent of women say that they go or can go alone to both these places. By age, the proportion of women who have no freedom (those who can go to neither place alone) varies little except that the teenagers are slightly more restricted. Overall, urban women, women in Khulna division, the best-educated women, and the wealthiest women have more freedom of movement. On the other hand, women in Sylhet and Barisal divisions have the least freedom of all subgroups shown in Table 3.10. Table 3.10 Freedom of movement Percentage of currently married women who say they go or can go outside the village or town, or to a hospital or health center, according to background characteristics, Bangladesh 2004 Background characteristic Goes or can go alone outside village Goes or can go alone to a health center or hospital Goes or can go to both places Goes or can go to at least one place Goes or can go to neither place Number of women Age 10-14 35.4 16.1 12.4 39.0 61.0 145 15-19 34.7 19.2 12.1 41.8 58.2 1,536 20-24 31.8 26.9 14.0 44.8 55.2 2,121 25-29 27.0 33.7 15.8 44.9 55.1 1,935 30-34 29.4 36.8 18.6 47.6 52.4 1,683 35-39 28.7 35.9 18.4 46.1 53.9 1,309 40-44 33.8 36.7 23.5 47.0 53.0 982 45-49 34.0 33.0 20.9 46.0 54.0 870 Residence Urban 37.6 42.6 24.8 55.4 44.6 2,372 Rural 29.1 27.6 14.4 42.2 57.8 8,210 Division Barisal 21.5 30.4 13.1 38.7 61.3 674 Chittagong 32.3 30.0 17.4 44.9 55.1 1,877 Dhaka 31.3 30.6 16.8 45.0 55.0 3,315 Khulna 34.3 40.0 22.4 51.9 48.1 1,296 Rajshahi 32.7 30.1 16.0 46.8 53.2 2,782 Sylhet 21.6 21.5 10.3 32.9 67.1 638 Education No education 27.7 29.9 16.0 41.6 58.4 4,187 Primary incomplete 26.9 29.3 14.2 42.0 58.0 2,176 Primary complete 27.9 29.4 14.7 42.6 57.4 958 Secondary incomplete 34.2 31.3 17.2 48.3 51.7 2,457 Secondary complete or higher 52.9 41.6 28.9 65.6 34.4 804 Wealth index Lowest 27.0 27.2 13.9 40.4 59.6 2,042 Second 27.1 27.6 13.8 40.9 59.1 2,112 Middle 29.9 28.4 15.5 42.9 57.1 2,112 Fourth 31.9 31.0 16.2 46.7 53.3 2,168 Highest 38.6 40.3 24.3 54.7 45.3 2,148 Total 31.0 31.0 16.8 45.2 54.8 10,582 3.6 MEN’S ATTITUDES TOWARDS WIFE-BEATING Domestic violence is not uncommon in Bangladesh. Although questions on domestic violence were not asked of female respondents, currently married men were asked whether they thought it was Characteristics of Survey Respondents │ 47 justified for a husband to beat his wife in the following four situations: if she goes out without telling him, if she neglects the children, if she argues with him, and if she fails to provide food on time. Table 3.11 shows that many currently married men find wife-beating to be justified in certain circumstances. Overall, more than half (55 percent) of Bangladeshi men agree that at least one of these factors is sufficient justification for wife-beating. This is not unexpected because many traditional customs in Bangladesh teach and expect women to accept, tolerate, and even rationalize wife-beating. This custom impedes women’s empowerment and has serious health consequences. The most widely accepted reason for wife-beating in Bangladesh is a wife going out without informing her husband; almost half of currently married men agree with this reason for wife-beating. Nearly three in ten currently married men think that if a wife argues with her husband it is a justifiable reason for beating. Only one in six men think that neglecting the children is a justification for wife- beating. An even smaller proportion believes that failing to provide food on time is a justifiable reason to hit or beat the wife. Table 3.11 Men’s attitude towards spousal violence Percentage of currently married men who agree that a husband is justified in hitting or beating his wife for specific reasons, by background characteristics, Bangladesh 2004 Husband is justified in hitting or beating his wife if she: Background characteristic Goes out without telling him Neglects the children Argues with him Fails to provide food on time Percentage who agree with at least one of the specified reasons Percentage who agree with none of the specified reasons Number of currently married men Age 15-19 (55.0) (24.2) (23.6) (6.0) (56.2) (43.8) 28 20-24 54.1 20.7 32.8 6.2 60.8 39.2 221 25-29 48.5 18.5 32.5 9.5 57.9 42.1 401 30-34 49.4 13.4 25.4 6.6 56.0 44.0 493 35-39 47.2 17.1 27.0 8.5 54.2 45.8 446 40-44 46.3 14.8 28.8 10.4 53.2 46.8 490 45-49 47.2 18.3 24.5 6.6 52.6 47.4 462 50-54 51.5 14.5 25.1 9.4 56.9 43.1 240 Residence Urban 39.1 11.4 19.8 4.6 45.8 54.2 655 Rural 51.6 18.2 30.1 9.3 58.3 41.7 2,125 Division Barisal 55.0 23.0 34.7 12.3 59.7 40.3 146 Chittagong 42.1 12.7 20.9 6.3 47.6 52.4 422 Dhaka 51.5 17.0 27.6 6.4 59.3 40.7 866 Khulna 48.1 13.3 25.6 6.1 54.6 45.4 384 Rajshahi 50.1 19.8 32.7 11.4 56.9 43.1 805 Sylhet 38.0 10.7 18.8 8.4 44.7 55.3 157 Education No education 56.0 20.5 34.9 11.9 64.2 35.8 866 Primary incomplete 53.1 19.3 33.1 9.3 60.7 39.3 831 Primary complete 50.1 15.5 29.2 7.9 56.4 43.6 209 Secondary incomplete 40.3 14.4 21.1 3.8 46.6 53.4 504 Secondary complete or higher 31.8 4.9 6.6 3.1 34.0 66.0 370 Wealth index Lowest 54.3 22.5 35.1 13.1 64.4 35.6 552 Second 56.0 17.6 32.6 9.3 62.9 37.1 599 Middle 53.0 19.1 33.4 9.3 60.0 40.0 555 Fourth 44.8 15.3 24.7 6.3 51.1 48.9 520 Highest 34.3 8.3 11.9 2.6 37.6 62.4 554 Total 48.6 16.6 27.7 8.2 55.4 44.6 2,780 Note: Figures in parentheses are based on 25-49 men. 48 │ Characteristics of Survey Respondents The table also shows attitudes towards wife-beating by background characteristics. Acceptance of wife-beating for each specified reason is lower among urban men and generally among those in Sylhet and Chittagong divisions. Men’s acceptance of wife-beating declines as their level of education increases. The acceptance of wife-beating is almost one-half as much among the most educated men as among the uneducated men. For each reason, poorer men (in the two lowest quintiles) are more likely than their wealthier counterparts to hold the attitude that wife-beating is justified. Since 1999-2000, attitudes toward wife-beating have changed for the worse. More than half of currently married men (55 percent) in 2004, compared with 37 percent earlier, agree with at least one specified reason justifying a husband beating his wife. The most noticeable change in men’s attitudes since 1999-2000 is a greater restriction on women’s movement. The percentage of men who consider wife-beating justified if a wife goes out without telling her husband has doubled from 25 percent to 49 percent. On the other hand, in the same period, the proportion of men who agree that a husband may beat his wife if she neglects her children has dropped from 23 to 17 percent. Fertility │ 49 FERTILITY 4 4.1 INTRODUCTION A major objective of the 2004 Bangladesh Demographic and Health Survey (BDHS) is to examine fertility levels, trends, and differentials in Bangladesh. The focus on fertility is due to its important role in determining Bangladesh’s population growth rate and its impact on economic development. This chapter presents a description of current and past fertility, cumulative fertility and family size, birth intervals, age at first birth, and reproductive behavior of adolescents. Most of the fertility measures are based on the birth histories collected from ever-married women age 15-49 interviewed during the survey. Each woman was asked a series of questions to give a retrospective history of all of her births. To encourage complete reporting, the interviewer asked the respondent about the number of sons and daughters living with her, the number living elsewhere, and the number who had died. She was then asked for a history of all of her births, including the month and year of birth, name, sex, and survival status of each birth. For children who had died, age at death was recorded. Interviewers were given extensive training in probing techniques designed to help respondents report this information accurately. The following measures of current fertility are derived from the birth history data: • Age-specific fertility rates1 (ASFR) are expressed as the number of births per thousand women in the age group and represent a valuable measure for assessing the current age pattern of childbearing. They are defined in terms of the number of live births during a specific period to women in the particular age group divided by the number of woman-years lived in that age group during the specified period. • Total fertility rate (TFR) is defined as the total number of births a woman would have by the end of her childbearing period if she were to pass through those years bearing children at the currently observed rates of age-specific fertility. The TFR is obtained by summing the age-specific fertility rates and multiplying by five. • General fertility rate (GFR) is the number of live births occurring during a specified period per 1,000 women of reproductive age. • Crude birth rate (CBR) is the number of births per 1,000 population during a specified period. 1 Numerators for age-specific fertility rates are calculated by summing the number of live births that occurred in the period 1-36 months preceding the survey (determined by the date of interview and the date of birth of the child) and classifying them by the age (in five-year groups) of the mother at the time of birth (determined by the mother’s date of birth). The denominators for the rates are the number of woman-years lived in each of the specified five-year age groups during the period 1-36 months preceding the survey. Since only women who had ever married were interviewed in the BDHS, the number of women in the denominator of the rates was inflated by factors calculated from information in the Household Questionnaire on the proportions ever married to produce a count of all women. Never-married women are presumed not to have given birth. 50 │ Fertility The various measures of current fertility are calculated for the three-year period preceding the survey, which roughly corresponds to the calendar period 2001-2003. A three-year period was chosen because it reflects the current situation, without unduly increasing sampling error. Despite measures to improve the data quality, BDHS information is subject to the same types of error that are inherent in all retrospective sample surveys, namely, possible omission of some births (especially births of children who died at a very young age) and the difficulty of determining the date of birth of each child accurately. These difficulties can bias estimates of fertility trends. A brief discussion of the quality of the BDHS fertility data appears in Appendix C.3 and shows that such errors are minimal. 4.2 CURRENT FERTILITY LEVELS Table 4.1 presents information on the current fertility levels for Bangladesh as a whole and for urban and rural areas. According to the 2004 BDHS results, the total fertility rate for women age 15-49 is 3.0. This means that a Bangladeshi woman would have, on average, 3.0 children in her lifetime if the current age- specific fertility rates remained constant. Table 4.1 Current fertility rates Age-specific fertility rates, the total fertility rate, the general fertility rate, and the crude birth rate for the three years preceding the survey, by urban-rural residence, Bangladesh 2004 Residence Age group Urban Rural Total 15-19 110 142 135 20-24 154 204 192 25-29 112 143 135 30-34 82 84 83 35-39 31 44 41 40-44 12 18 16 45-49 1 3 3 TFR 15-49 2.5 3.2 3.0 GFR 97 122 117 CBR 25.8 29.5 28.7 Note: Rates are for the period 1-36 months preceding the survey. Rates for age group 45-49 may be slightly biased because of truncation. TFR: Total fertility rate for women age 15-49, expressed per woman GFR: General fertility rate (births divided by the number of women age 15-44), expressed per 1,000 women CBR: Crude birth rate, expressed per 1,000 population The general fertility rate in Bangladesh is 117. This means that there were 117 births for every 1,000 women of reproductive age during the three-year period preceding the survey. Data also show a crude birth rate of 29 births per 1,000 population for the period under review. Bangladeshi women have a pattern of early childbearing. According to current fertility rates, on average, women will have 22 percent of their births before reaching age 20 and will complete 76 percent of their childbearing before age 30. Fertility │ 51 4.3 FERTILITY DIFFERENTIALS Table 4.2 shows differentials in fertility by residence, administrative division, educational attainment, and wealth index. The total fertility rate for rural women (3.2 births) is almost one child higher than for urban women. The urban-rural difference is especially large at younger ages, which probably reflects longer education and later marriage of women in urban areas (Figure 4.1). Table 4.2 Fertility by background characteristics Total fertility rate for the three years preceding the survey and mean number of children ever born to women age 40-49 years, by background characteristics, Bangladesh 2004 Background characteristic Total fertility rate1 Mean number of children ever born to women age 40-49 Residence Urban 2.5 4.4 Rural 3.2 5.3 Division Barisal 2.9 5.5 Chittagong 3.7 5.6 Dhaka 2.9 4.9 Khulna 2.8 4.5 Rajshahi 2.6 4.9 Sylhet 4.2 5.8 Education No education 3.6 5.3 Primary incomplete 3.3 5.4 Primary complete 2.9 5.0 Secondary incomplete 2.7 4.3 Secondary complete or higher 2.2 2.8 Wealth index Lowest 4.0 5.4 Second 3.2 5.6 Middle 3.0 5.3 Fourth 2.3 5.1 Highest 2.5 4.3 Total 3.0 5.1 1 Women age 15-49 years 52 │ Fertility Fertility varies widely by administrative divisions (Figure 4.2). Sylhet and Chittagong divisions have the highest TFRs (4.2 and 3.7, respectively), while the lowest TFRs are in Rajshahi and Khulna (2.6 and 2.8, respectively). As expected, the educational attainment of women is strongly associated with fertility. The TFR decreases steeply from 3.6 for women with no education to 2.2 for women with secondary complete or higher Fertility │ 53 education. Fertility is also negatively related with wealth; the disparity in fertility between women in the lowest and highest wealth quintiles is 1.5 children per woman. Table 4.2 also presents a crude assessment of trends in fertility in the various subgroups by comparing the TFRs for the three years preceding the survey with the average number of children ever born to women who are now at the end of their childbearing period, age 40-49. The former is a measure of current fertility, while the latter is a measure of past or completed fertility. Overall, comparison of past and present fertility indicators suggests a decline of two children per woman, from 5.1 to 3.0 children. There has been a substantial decline in fertility in urban and rural areas, and in all administrative divisions. The fertility declined by two or more children in three of the six divisions: Barisal, Dhaka, and Rajshahi. 4.4 FERTILITY TRENDS Trends in fertility in Bangladesh since the early 1970s can be examined by observing a time series of estimates produced from demographic surveys fielded over the last two and half decades, beginning with the 1975 Bangladesh Fertility Survey (BFS). The estimates shown in Table 4.3 describe the ongoing Bangladeshi fertility transition. Fertility has declined sharply, from 6.3 in 1971-1975 to 3.0 in 2001-2003 (Figure 4.3). During this period, fertility declined rapidly in the late 1980s and early 1990s, and plateaued at around 3.3 for most of the 1990s. The 2004 BDHS data indicate that after almost a decade-long stagnation, the TFR declined slightly from 3.3 to 3.0 between 1997-1999 and 2001-2003. Since 1997-1999, a small decline in fertility was observed in all divisions except Sylhet and Khulna. Investigation of the age pattern of fertility shows no anomalies; the decline since the mid-1980s has been fairly uniform over all age groups of women. Table 4.3 Trends in current fertility rates Age-specific fertility rates (per 1,000 women) and total fertility rates (TFRs) among women age 15-49, selected sources, Bangladesh, 1975- 2004 Survey and approximate time period 1975 BFS 1989 BFS 1991 CPS 1993-1994 BDHS 1996-1997 BDHS 1999-2000 BDHS 2004 BDHS Age group 1971-1975 1984-1988 1989-1991 1991-1993 1994-1996 1997-1999 2001-2003 15-19 109 182 179 140 147 144 135 20-24 289 260 230 196 192 188 192 25-29 291 225 188 158 150 165 135 30-34 250 169 129 105 96 99 83 35-39 185 114 78 56 44 44 41 40-44 107 56 36 19 18 18 16 45-49 35 18 13 14 6 3 3 TFR 15-49 6.3 5.1 4.3 3.4 3.3 3.3 3.0 Note: For the 1975 Bangladesh Fertility Survey (BFS) and 1989 BFS, the rates refer to the five-year period preceding the survey; for the other surveys, the rates refer to the three-year period preceding the survey. The BFS and Bangladesh Demographic and Health Survey (BDHS) utilized full birth histories, while the 1991 Contraceptive Prevalence Survey (CPS) used an eight-year truncated birth history. Sources: 1975 BFS (MHPC, 1978:73), 1989 BFS (Huq and Cleland, 1990:103), 1991 CPS (Mitra et al., 1993 :34), 1993-1994 BDHS (Mitra et al., 1994:24), 1996-1997 BDHS (Mitra et al., 1997:30), and 1999-2000 BDHS (NIPORT et al., 2001:34) 54 │ Fertility Table 4.4 shows trends in the percentage of currently married women who reported that they were pregnant at the time of the survey, according to age group. Reports on current pregnancy are almost surely underestimates, since many women may be pregnant but not yet aware of their status. However, the data are useful because, while fertility rates depend to some extent on accurate reporting of dates of events, the proportion pregnant is a “current status” indicator. Change over time in the percent pregnant is an independent indicator of fertility change. In Bangladesh, the percent pregnant has generally declined over time, from 13 percent in 1975 to 7 percent in 2004. During this period, the percent pregnant declined in the late 1980s and early 1990s, and stalled around 8 percent for most of the 1990s. The 2004 BDHS results show that the percent pregnant has declined to 7 percent from 8 percent in 1996-1997 and 1999-2000. Table 4.4 Percent pregnant Percentage of currently married women who were pregnant at the time of interview, by age group, selected sources, Bangladesh 1975-2004 Age group 1975 BFS 1989 BFS 1991 CPS 1993-1994 BDHS 1996-1997 BDHS 1999-2000 BDHS 2004 BDHS 15-19 15.2a 14.7a 19.6 17.1 14.7 15.9 14.3 20-24 15.5 13.3 16.2 13.0 10.3 11.8 9.9 25-29 14.9 10.4 11.2 9.0 8.9 8.5 7.8 30-34 11.2 8.3 7.1 7.0 5.1 4.8 4.8 35-39 10.7 4.8 4.2 2.7 3.4 2.4 2.3 40-44 u u 1.5 0.8 1.3 1.0 0.1 45-49 u u 0.2 0.0 0.0 0.4 0.2 Total 12.5 9.3 10.7 8.7 7.7 7.8 6.6 u = Unknown (not available) a Currently married women under age 20 Source: 1975 Bangladesh Fertility Survey (BFS) and 1989 BFS (Cleland et al., 1994:21), 1991 Contraceptive Prevalence Survey (CPS) (Mitra et al., 1993:39), 1993-1994 Bangladesh Demographic and Health Survey (BDHS) (Mitra et al., 1994:31), 1996-1997 BDHS (Mitra et al.1997:34), and 1999-2000 BDHS (NIPORT et al., 2001:36) Fertility │ 55 Table 4.5 Trends in fertility by marital duration Fertility rates for ever-married women by duration (years) since first marriage, for five-year periods preceding the survey, Bangladesh 2004 Number of years preceding the survey Marriage duration 0-4 5-9 10-14 15-19 0-4 253 279 260 266 5-9 205 229 249 282 10-14 143 175 202 242 15-19 86 109 166 [207] 20-24 45 75 [114] - 25-29 22 [45] - - Note: Duration-specific fertility rates are per 1,000 women. Estimates in brackets are truncated. Table 4.5 presents fertility rates for ever-married women by marital duration, the number of years since the first marriage, for five-year periods preceding the survey. The data confirm a sharp decline in fertility and indicate that fertility has declined at all marital durations. 4.5 CHILDREN EVER BORN AND LIVING Table 4.6 shows the distribution of all women and currently married women by age and number of children ever born. It also shows the mean number of children ever born to women in each five-year age group, an indicator of the momentum of childbearing. The mean number of children ever born for all women is 2.5, which means that, on average, Bangladeshi women age 15-49 have had fewer than three births; currently married women have 3.0 births on average. Allowing for mortality of children, Bangladeshi women have, on average, 2.2 living children, while currently married women have an average of 2.6 living children. Figures for currently married women do not differ greatly from those for all women at older ages; however, at younger ages, the percentage of currently married women who have had children is much higher than the percentage among all women. Currently married women age 45-49 have given birth to an average of 5.8 children, of whom 4.7 survived. Among all women age 15-49, the average number of children who have died per woman is 0.35. Among currently married women, it is 0.41; that is, 14 percent of children born to currently married women had died. The proportion of children ever born who have died increases with women’s age. Among currently married women, for example, the proportion of children ever born who have died increases from 8 percent for women age 20-24 to 20 percent for women age 45-49. 56 │ Fertility Table 4.6 Children ever born and living Percent distribution of all women and currently married women age 15-49 by number of children ever born, and mean number of children ever born and mean number of living children, according to age group, Bangladesh 2004 Number of children ever born Age 0 1 2 3 4 5 6 7 8 9 10+ Total Number of women Mean number of children ever born Mean number of living children ALL WOMEN 15-19 72.1 21.5 5.9 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 3,337 0.35 0.32 20-24 24.4 30.0 28.5 13.3 3.2 0.5 0.0 0.0 0.0 0.0 0.0 100.0 2,597 1.42 1.31 25-29 8.4 12.0 30.8 25.9 14.8 5.7 1.6 0.7 0.1 0.1 0.0 100.0 2,100 2.55 2.28 30-34 4.4 6.3 21.2 23.5 19.9 12.9 6.5 3.1 1.8 0.5 0.1 100.0 1,815 3.43 3.02 35-39 2.7 4.8 14.0 19.7 20.3 14.9 11.3 5.9 3.7 1.7 1.0 100.0 1,463 4.12 3.48 40-44 3.0 3.0 8.5 17.6 21.1 14.8 11.2 10.0 5.1 2.9 2.7 100.0 1,164 4.65 3.84 45-49 1.4 1.6 6.0 10.9 15.3 15.0 15.2 12.1 9.3 6.4 6.8 100.0 1,066 5.63 4.48 Total 25.0 14.7 17.3 14.3 10.8 6.8 4.5 3.0 1.8 1.0 0.9 100.0 13,542 2.50 2.15 CURRENTLY MARRIED WOMEN 15-19 40.8 45.3 12.9 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 100.0 1,536 0.74 0.68 20-24 10.3 35.0 34.0 16.2 3.9 0.6 0.0 0.0 0.0 0.0 0.0 100.0 2,121 1.70 1.56 25-29 3.8 11.9 32.2 27.7 15.8 6.1 1.6 0.7 0.1 0.1 0.0 100.0 1,935 2.69 2.41 30-34 2.6 5.3 20.9 24.4 20.7 13.7 6.8 3.2 1.9 0.5 0.1 100.0 1,683 3.55 3.14 35-39 1.8 3.7 13.2 19.7 20.7 16.0 11.7 6.4 3.8 1.8 1.0 100.0 1,309 4.25 3.61 40-44 1.8 2.5 8.1 16.5 21.4 15.9 11.8 10.3 5.1 3.5 3.0 100.0 982 4.81 4.01 45-49 1.1 0.8 5.2 9.7 15.4 15.4 15.7 12.7 9.8 6.8 7.6 100.0 870 5.82 4.67 Total 9.7 17.6 21.0 17.3 12.9 8.3 5.3 3.5 2.1 1.2 1.1 100.0 10,436 3.00 2.59 Nearly three-fourths of women age 15-19 have never given birth. However, this proportion declines to 4 percent for women age 30-34 years and rapidly decreases further for older women, indicating that childbearing among Bangladeshi women is nearly universal. The percentage of women in their forties who have never had children provides an indicator of the level of primary infertility—the proportion of women who are unable to bear children at all. Since voluntary childlessness is rare in Bangladesh, it is likely that married women with no births are unable to have children. The BDHS results suggest that primary infertility is low, less than 2 percent. (This estimate of primary infertility does not include women who may have had one or more births but who are unable to have more [secondary infertility].) A comparison of the mean number of children ever born reported in the 2004 BDHS and various other surveys is presented in Table 4.7. The comparison does not highlight recent changes in fertility, but rather is an indication of the cumulative changes in fertility over the decades preceding the 2004 BDHS. Despite the fluctuations between surveys, the data generally show only modest declines until the late 1980s. Between 1985 and 1989, the decline in mean number of children ever born was substantial in all but the youngest and oldest age groups. Although this was followed by little change between 1989 and 1991, the mean number of children again declined considerably between 1991 and 1993-1994, especially among women age 25 and above, and showed further decline between 1993-1994 and 1999-2000 at all ages except 15-19. The most recent data showed a decline in the mean number of children between 1999-2000 and 2004 among women age 30 and above. Fertility │ 57 Table 4.7 Trends in children ever born Mean number of children ever born by age group, selected sources, Bangladesh, 1975-2004 Age group 1975 BFS 1981 CPS 1983 CPS 1985 CPS 1989 BFS 1989 CPS 1991 CPS 1993- 1994 BDHS 1996- 1997 BDHS 1999- 2000 BDHS 2004 BDHS 15-19 0.6 0.5 0.6 0.4 0.4 0.4 0.4 0.3 0.4 0.4 0.4 20-24 2.3 2.0 2.2 2.0 1.7 1.8 1.7 1.6 1.5 1.4 1.4 25-29 4.2 3.7 3.8 3.6 3.1 3.3 3.2 2.9 2.8 2.6 2.6 30-34 5.7 5.4 5.5 5.1 4.7 4.7 4.5 4.1 3.9 3.6 3.4 35-39 6.7 6.4 6.5 6.5 5.9 5.9 5.7 5.2 4.8 4.3 4.1 40-44 7.1 7.3 7.4 7.4 6.6 7.0 6.7 6.4 5.6 5.1 4.7 45-49 6.7 7.6 7.5 7.2 7.3 7.5 7.4 6.9 6.4 6.1 5.6 Total u u u u u u 3.5 3.0 2.8 2.6 2.5 u = Unknown (not available) Source: 1975 Bangladesh Fertility Survey (BFS), 1983 and 1985 CPSs (Kantner and Frankenberg, 1988:21); 1991 CPS (Mitra et al., 1993:31); 1993-1994 BDHS (Mitra et al., 1994:33); 1996-1997 BDHS (Mitra et al., 1997: 36); 1999-2000 BDHS (NIPORT et al., 2001:39) all others (Cleland et al., 1994:11) 4.6 BIRTH INTERVALS Examination of birth intervals, defined as the length of time between two successive live births, is important in providing insights into birth spacing patterns and, subsequently, maternal and child health. Short birth intervals are associated with an increased risk of death for mother and child. Studies have shown that children born less than 24 months after a previous sibling risk poorer health and also threaten maternal health. Table 4.8 shows the percent distribution of non-first births that occurred in the five years preceding the BDHS by the number of months since the previous birth. Birth intervals are generally long in Bangladesh (the median birth interval is 39 months). The long period of breastfeeding in Bangladesh (an average of 32 months [Chapter 11]) and the corresponding long period of postpartum amenorrhea (an average of 9 months [Chapter 6]) are likely to contribute to the relatively high percentage of births occurring after an interval of 24 months or more. Almost six in ten non- first births occur three or more years after the previous birth, while one-fourth of births take place 24-35 months after the previous birth (Table 4.8). Nearly one in six children (16 percent) is born after a “too short” interval (less than 24 months). The median birth interval in the 2004 BDHS (39 months) is almost the same as in the 1999-2000 BDHS. However, it is longer than the median reported in the 1993-1994 BDHS (35 months) and in the 1996- 1997 BDHS (37 months). The median birth interval is substantially shorter for teenage mothers (27 months). More than one in three births to teenage mothers age 15-19 occurs after a “too short” interval of less than 24 months. 58 │ Fertility Table 4.8 Birth intervals Percent distribution of non-first births in the five years preceding the survey, by number of months since preceding birth, according to background characteristics, Bangladesh 2004 Months since preceding birth Background characteristic 7-17 18-23 24-35 36-47 48+ Total Number of non-first births Median number of months since preceding birth Age of mother 15-19 14.5 20.0 40.6 21.0 3.9 100.0 224 26.5 20-29 6.9 10.1 27.4 23.1 32.5 100.0 3,028 37.8 30-39 4.6 8.4 20.7 18.9 47.4 100.0 1,485 46.0 40-49 4.0 8.1 17.7 14.1 56.1 100.0 213 52.2 Birth order 2-3 6.0 9.4 24.6 21.9 38.0 100.0 3,018 40.3 4-6 6.9 10.2 26.9 20.7 35.3 100.0 1,516 38.5 7+ 8.1 12.8 27.4 19.6 32.0 100.0 417 36.7 Sex of prior birth Male 6.7 9.6 24.9 21.7 37.0 100.0 2,484 39.5 Female 6.2 10.3 26.2 21.0 36.3 100.0 2,467 39.1 Survival of prior birth Living 4.3 9.2 24.8 22.5 39.2 100.0 4,409 41.2 Dead 24.4 15.7 31.6 12.2 16.1 100.0 542 26.0 Residence Urban 6.5 10.2 21.3 20.1 41.9 100.0 939 42.8 Rural 6.5 9.9 26.5 21.7 35.5 100.0 4,012 38.7 Division Barisal 4.6 7.2 26.3 22.1 39.8 100.0 294 41.6 Chittagong 6.8 10.4 30.1 24.1 28.7 100.0 1,106 36.8 Dhaka 6.1 9.7 26.4 20.7 37.1 100.0 1,580 39.7 Khulna 6.8 7.8 14.8 18.1 52.5 100.0 480 49.4 Rajshahi 5.2 9.9 21.7 20.2 43.0 100.0 1,034 43.5 Sylhet 10.7 13.9 30.9 22.7 21.9 100.0 456 33.7 Education No education 5.7 10.2 26.8 21.5 35.7 100.0 2,454 38.8 Primary incomplete 5.1 9.3 25.6 25.2 34.8 100.0 965 39.5 Primary complete 8.5 9.4 24.1 19.8 38.1 100.0 455 39.5 Secondary incomplete 8.3 9.8 23.7 19.5 38.7 100.0 863 40.2 Secondary complete or higher 9.1 10.6 20.9 13.4 46.0 100.0 213 44.1 Wealth index Lowest 6.5 11.1 30.7 22.8 28.9 100.0 1,422 36.5 Second 5.7 9.4 25.4 23.4 36.1 100.0 1,056 39.1 Middle 6.8 9.3 25.2 20.3 38.3 100.0 947 39.5 Fourth 5.1 10.6 23.2 19.8 41.2 100.0 817 41.3 Highest 8.6 8.4 18.7 18.6 45.7 100.0 708 45.5 Total 6.5 9.9 25.5 21.4 36.7 100.0 4,951 39.3 Note: First-order births are excluded. The interval for multiple births is the number of months since the preceding pregnancy that ended in a live birth. Fertility │ 59 The length of the birth interval is closely associated with the survival status of the previous sibling. The median birth interval is 15 months shorter for children whose previous sibling died than for children whose previous sibling is alive (26 and 41 months, respectively). The percentage of births occurring within a very short interval (less than 18 months) is six times higher for children whose prior sibling died than for children whose prior sibling survived (24 and 4 percent, respectively). The shorter intervals for the former group are partly due to a shortened period of breastfeeding (or no breastfeeding) for the preceding child, leading to an earlier return of ovulation and hence increased chance of pregnancy. Minimal use of contraception, presumably because of a desire to “replace” the dead child as soon as possible, could also be one of the factors responsible for the shorter birth interval in these cases. The median number of months since the preceding birth increases with household economic status; from 37 months in the lowest wealth quintile to 46 months for households in the highest wealth quintile. 4.7 AGE AT FIRST BIRTH The onset of childbearing has a direct effect on fertility. Early initiation into childbearing lengthens the reproductive period and subsequently increases fertility. In many countries, postponement of first births— reflecting an increase in the age at marriage—has contributed greatly to overall fertility decline. Moreover, bearing children at a young age involves substantial risks to the health of both the mother and child. Early childbearing also tends to restrict educational and economic opportunities for women. Table 4.9 presents the percent distribution of women by age at first birth according to current age.2 The median age at first birth is not shown for women age 15-19 because the large majority of these young girls had not become mothers before age 15. The median age at first birth is about 18 years across all age cohorts, except for women age 45-49 years, whose median age at first birth is 17 years, indicating a slight change in the age at first birth. This slight increase in age at first birth is reflected in the smaller proportion of younger women whose first birth occurred before age 15; 18 percent of women in their late forties report having had their first birth before age 15, compared with only 6 percent of women age 15-19. Comparison of data from the 1999-2000 BDHS and the 2004 BDHS shows little change in the median age at first birth for women age 20-49. Table 4.9 Age at first birth Percent distribution of women by age at first birth, according to current age, Bangladesh 2004 Age at first birth Current age Percent- age of women with no births <15 15-17 18-19 20-21 22-24 25+ Total Number of women Median age at first birth 15-19 72.1 6.4 na na na na na 100.0 3,337 a 20-24 24.4 10.2 35.8 17.7 na na na 100.0 2,597 18.4 25-29 8.4 13.4 40.7 19.5 9.3 6.7 2.0 100.0 2,100 17.7 30-34 4.4 13.2 40.7 19.9 10.4 7.2 4.2 100.0 1,815 17.7 35-39 2.7 11.5 43.0 19.5 12.3 5.6 5.5 100.0 1,463 17.7 40-44 3.0 13.7 42.2 19.0 11.4 5.9 4.8 100.0 1,164 17.6 45-49 1.4 17.7 47.6 14.6 8.4 6.0 4.4 100.0 1,066 16.9 na = Not applicable a = Omitted because less than 50 percent of the women had a birth before reaching age 15 2 The data are based on all women, including those who have never married (see footnote 1 for a description of the inflation factors used to estimate the total number of women). 60 │ Fertility Table 4.10 summarizes the median age at first birth for different age cohorts by background characteristics of respondents. For women age 20-49, the median is slightly higher in urban areas than in rural areas, and is highest in Sylhet division, compared with other divisions. Median age at first birth is about two years higher for women in the highest wealth quintile (19 years), compared with those in the lowest wealth quintile (17 years). Women with secondary or more education start childbearing later than those with less or no education. Among women age 25-49, the median age at first birth is 17 years for women with no education and 23 years for women who have completed secondary education. Table 4.10 Median age at first birth Median age at first birth among women age 20-49 years, by current age and background characteristics, Bangladesh 2004 Current age Women age Women age Background characteristic 20-24 25-29 30-34 35-39 40-44 45-49 20-49 25-49 Residence Urban 19.5 18.3 18.1 18.2 18.1 17.5 18.4 18.1 Rural 18.1 17.5 17.6 17.6 17.5 16.8 17.6 17.4 Division Barisal 18.9 17.0 18.4 17.5 17.7 17.7 17.8 17.7 Chittagong 19.2 18.0 18.0 18.1 17.5 17.4 18.2 17.8 Dhaka 18.3 17.9 17.5 17.5 17.7 17.1 17.7 17.6 Khulna 17.9 17.4 17.4 17.7 17.1 16.4 17.4 17.3 Rajshahi 17.7 17.1 17.5 17.5 17.5 16.4 17.3 17.2 Sylhet 19.9 19.0 18.9 18.7 18.6 17.3 18.9 18.6 Education No education 16.6 16.8 17.2 17.6 17.4 16.5 17.0 17.1 Primary incomplete 17.1 17.2 17.6 17.3 17.4 17.1 17.3 17.3 Primary complete 17.4 17.6 17.6 17.2 17.8 17.9 17.6 17.6 Secondary incomplete 19.0 18.3 18.6 18.0 18.0 17.7 18.6 18.3 Secondary complete or higher a 23.4 23.0 21.0 21.1 21.4 a 22.6 Wealth index Lowest 16.9 16.8 17.5 17.7 17.9 16.4 17.1 17.2 Second 17.4 17.2 17.3 17.6 17.7 16.7 17.3 17.3 Middle 18.6 17.4 17.7 17.3 17.1 17.0 17.6 17.3 Fourth 19.7 18.4 17.5 17.5 17.5 17.0 18.1 17.7 Highest 19.8 19.3 18.8 18.4 18.1 17.5 18.9 18.5 Total 18.4 17.7 17.7 17.7 17.6 16.9 17.7 17.6 Note: The medians for cohort 15-19 could not be determined because more than half of these women have not yet had a birth. a = Omitted because less than 50 percent of the women had a birth before reaching age 20 4.8 ADOLESCENT FERTILITY Adolescent fertility is a major social and health concern. Teenage mothers are more likely to suffer from severe complications during delivery, which result in higher morbidity and mortality for both themselves and their children. In addition, young mothers may not be sufficiently emotionally mature to bear the burden Fertility │ 61 of childbearing and rearing. Early entry into reproduction denies them the opportunity to pursue academic goals. This is detrimental to their prospects for good careers, which often lowers their status in society. Table 4.11 shows that one-third of adolescents age 15-19 have begun childbearing. Twenty-eight percent of these teenagers in Bangladesh have given birth, and another 5 percent are pregnant with their first child. As expected, the proportion of women age 15-19 who have begun childbearing rises rapidly with age. Early childbearing among teenagers is more prominent in rural areas, compared with urban areas, and in Rajshahi and Khulna divisions, compared with other divisions. Childbearing begins later in Sylhet, compared with the rest of the divisions, mainly because of relatively late marriage in Sylhet. Delayed childbearing is strongly related to education among women age 15-19. Only 16 percent of the teenagers who had completed secondary education had begun childbearing, compared with almost half of those with primary incomplete or no education. Childbearing begins earlier among adolescents in the poorest 40 percent of the households; in these households, four out of ten adolescents have begun childbearing. In contrast, three out of ten of the adolescents in the richest 20 percent of the households either have had births or are pregnant with their first child. Table 4.11 Teenage pregnancy and motherhood Percentage of women age 15-19 who are mothers or pregnant with their first child, by background characteristics, Bangladesh 2004 Percentage who are: Background characteristic Mothers Pregnant with first child Percentage who have begun childbearing Number of women Age 15 7.5 4.0 11.5 778 16 18.3 3.9 22.2 706 17 30.8 6.4 37.2 660 18 36.9 5.9 42.8 629 19 54.8 4.0 58.8 563 Residence Urban 21.9 4.3 26.1 758 Rural 29.7 5.0 34.7 2,576 Division Barisal 25.9 3.6 29.5 215 Chittagong 23.7 4.0 27.7 725 Dhaka 27.2 4.2 31.5 956 Khulna 31.2 6.4 37.7 367 Rajshahi 35.7 6.6 42.3 797 Sylhet 16.2 2.8 19.0 271 Education No education 42.5 4.0 46.5 435 Primary incomplete 40.2 5.6 45.8 541 Primary complete 32.8 4.3 37.1 334 Secondary incomplete 22.3 5.0 27.2 1,760 Secondary complete or higher 10.9 4.6 15.5 253 Wealth index Lowest 37.7 3.2 40.9 559 Second 33.7 5.6 39.2 665 Middle 21.7 3.6 25.2 1,006 Fourth 17.4 4.5 21.9 949 Highest 23.0 5.6 28.5 492 Total 27.9 4.8 32.7 3,337 62 │ Fertility Between the 1999-2000 BDHS and the 2004 BDHS, the proportion of adolescents age 15-19 who had begun childbearing declined slightly, from 35 to 33 percent. Fertility Regulation │ 63 FERTILITY REGULATION 5 5.1 KNOWLEDGE OF FAMILY PLANNING METHODS Information on knowledge of family planning methods was collected by asking female respondents to name ways or methods by which a couple could delay or avoid pregnancy. If the respondent did not mention a particular method spontaneously, the interviewer described the method and asked whether the respondent had heard about the method. In this manner, knowledge was assessed for seven modern methods of family planning (the pill, IUD, injectables, Norplant, condoms, female sterilization, and male sterilization) and two traditional methods of family planning (periodic abstinence and withdrawal). Any other methods, if spontaneously mentioned by the respondent, were also recorded. Knowledge of family planning methods is widespread in Bangladesh (Table 5.1). All ever-married women know of at least one modern method of family planning, and eight out of every ten women know of at least one traditional method. On average, a woman has heard of 7.6 methods of family planning. There is virtually no difference in knowledge between ever-married and currently married women. Table 5.1 Knowledge of contraceptive methods Percentage of ever-married and currently married women who know any contraceptive methods, by specific method, Bangladesh 2004 Contraceptive method Ever-married women Currently married women Any method 100.0 100.0 Any modern method 100.0 100.0 Pill 99.9 99.9 IUD 84.9 85.0 Injectables 98.5 98.6 Norplant 76.0 76.7 Condom 91.6 92.2 Female sterilization 96.1 96.1 Male sterilization 72.7 72.9 Any traditional method 80.3 81.0 Periodic abstinence 70.4 71.0 Withdrawal 57.8 58.7 Other 7.3 7.6 Mean number of methods known 7.6 7.6 Number of women 11,440 10,582 Almost all respondents have heard of pills, injectables, and female sterilization. More than nine out of ten women know of condoms. Knowledge of other modern methods is also widespread; a majority of currently married women have heard of the IUD (85 percent), Norplant (77 percent), and male sterilization (73 percent). Knowledge of traditional methods is lower than modern methods. The data show similar levels of knowledge of specific methods for both currently married and ever-married women. 64 │ Fertility Regulation Since the 1999-2000 Bangladesh Demographic and Health Survey (BDHS), there has been little overall change in knowledge of contraceptives. However, knowledge of some specific methods has changed. For example, knowledge of implants has increased from 56 to 77 percent of currently married women and knowledge of condoms increased from 90 to 92 percent. 5.2 EVER USE OF CONTRACEPTION Respondents who said that they had heard of a method of family planning were asked whether they had ever used that method. Ever use of family planning methods in the 2004 BDHS thus refers to use of a method at any time, without making a distinction between past and current use. Collection and analysis of ever-use data have special significance for family planning programs. These data indicate the proportion of the population who were exposed to contraceptive use at least once. Therefore, data on ever use reflect the success of programs in promoting use of family planning methods among eligible couples. In addition, data on ever use—together with data on current use—are valuable for studying couples who discontinue use. Among ever-married women, four-fifths have used a contraceptive method at some time, three- fourths have used a modern method, and more than one-fourth have used a traditional method (Table 5.2). The pill is by far the most commonly used method; more than six out of ten ever-married women say they have used it. The next most commonly used method is injectables (26 percent). Very few women report having ever used male sterilization (less than 1 percent). As expected, currently married women are somewhat more likely than ever-married women to report ever use of a family planning method. Table 5.2 Ever use of contraception Percentage of ever-married and currently married women who have ever used any contraceptive method, by specific method and age, Bangladesh 2004 Modern method Traditional method Age Any method Any modern method Pill IUD Inject- ables Norplant Condom Female sterili- zation Male sterili- zation Any tradi- tional method Periodic absti- nence With- drawal Other Number of women EVER-MARRIED WOMEN 10-14 54.0 48.1 42.7 0.0 0.4 0.0 14.3 0.0 0.0 10.2 5.2 5.0 0.0 150 15-19 67.0 60.3 49.9 0.5 10.4 0.3 21.9 0.0 0.0 18.8 7.5 13.1 0.4 1,598 20-24 81.1 76.0 67.4 1.6 22.1 1.5 26.9 0.3 0.1 25.3 14.0 14.6 1.1 2,202 25-29 87.1 83.8 73.8 4.3 33.8 2.3 23.4 2.1 0.4 29.0 17.6 15.5 3.0 2,013 30-34 86.8 82.0 70.7 8.2 39.2 1.3 20.5 5.2 0.4 33.4 23.4 14.0 3.9 1,793 35-39 84.7 80.4 65.7 11.4 32.3 1.6 18.8 9.3 1.8 34.9 25.8 14.5 3.5 1,457 40-44 80.5 73.1 56.1 10.7 26.0 1.0 14.9 13.2 1.9 37.8 30.3 15.0 4.2 1,160 45-49 70.9 59.9 40.0 6.7 15.7 0.2 8.8 15.9 1.4 32.8 25.3 10.6 3.7 1,066 Total 80.2 74.5 62.4 5.6 26.0 1.3 20.5 5.2 0.7 29.3 19.3 14.0 2.6 11,440 CURRENTLY MARRIED WOMEN 10-14 54.6 48.6 43.0 0.0 0.4 0.0 14.7 0.0 0.0 10.5 5.4 5.1 0.0 145 15-19 68.5 61.8 51.2 0.5 10.8 0.3 22.5 0.0 0.0 19.1 7.6 13.3 0.4 1,536 20-24 82.0 77.1 68.4 1.7 22.5 1.5 27.5 0.3 0.1 25.8 14.3 14.8 1.1 2,121 25-29 88.9 85.6 75.5 4.4 34.5 2.4 23.8 2.2 0.4 29.4 17.6 16.0 3.1 1,935 30-34 88.8 84.5 73.1 8.6 40.8 1.4 21.4 5.4 0.5 34.4 24.0 14.5 4.2 1,683 35-39 89.2 84.7 69.9 12.0 34.7 1.8 19.8 9.6 1.8 37.1 27.4 15.6 3.8 1,309 40-44 86.1 79.4 61.3 12.0 29.5 1.2 16.4 14.1 1.9 40.7 32.7 16.2 4.8 982 45-49 75.5 65.4 44.3 7.9 18.1 0.2 10.2 16.7 1.5 35.2 27.1 11.3 4.1 870 Total 82.8 77.4 65.1 5.8 27.4 1.4 21.5 5.2 0.7 30.2 19.7 14.6 2.8 10,582 Fertility Regulation │ 65 The level of ever use of family planning has increased steadily in Bangladesh (Table 5.3). In 2004, 80 percent of ever

View the publication