Pakistan - Demographic and Health Survey - 1992

Publication date: 1992

Pakistan Pakistan Demographic and Health Survey 1990/1991 Nationat Institute of Population Studies ~DHS Demographic and Health Surveys IRD/Macro International inc. Pakistan Demographic and Health Survey 1990/1991 National Institute of Population Studies Islamabad, Pakistan IRD/Macro International Inc. Columbia, Maryland USA July 1992 This report summarises the findings of the 1990-91 Pakistan Demographic and Health Survey (PDHS) conducted by the National Institute of Population Studies, in collaboration with the Federal Bureau of Statistics. IRD/Macro International Inc. provided technical assistance. Funding was provided by the U.S. Agency for International Development and the Govemment of Pakistan. The PDHS is part of the worldwide Demographic and Health Surveys (DHS) programme, which is designed to collect data on fetaility, family planning, and maternal and child health. Additional information on the Pakistan survey may be obtained from the National Institute of Population Studies, No. 8, Street 70, F-8/3, Islamabad, Pakistan (Telephone 850205; Fax 851977; Telex 54139 NIPS PK). Additional information about the DHS programme may be obtained by writing to: DHS, IRD/Macro lntemational Inc., 8850 Stanford Boulevard, Suite 4000, Columbia, MD 21045, USA (Telephone 410-290-2800; Fax 410-290-2999; Telex 198116). CONTENTS Page Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii Summary of Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xix Map of Pakistan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiv 3 INTRODUCTION by Abdul Razzaque Rukanuddin and Tauseef Ahmed 1.1 Physical Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Climate, Rainfall, and Seasons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.3 Administrative Divisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.4 People, Culture, Religion, and Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.5 Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.6 Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.7 Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.8 Literacy and Educational Attainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.9 Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.10 Population and Family Planning Policies and Programmes . . . . . . . . . . . . . . . . . . . . . . 5 1.11 Health Policies and Programmes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 SURVEY DESIGN AND IMPLEMENTATION by Tauseef Ahmed, M.D. Mallick and Alfredo Aliaga 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 Objectives of the Pakistan Demographic and Health Survey . . . . . . . . . . . . . . . . . . . . 9 Organisation of the Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Sample Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Recruitment, Training and Fieldwork . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Data Entry and Processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Field Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Coverage of the Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 CHARACTERISTICS OF HOUSEHOLDS AND RESPONDENTS by Tauseef Ahrned and Syed Mubashir Ali 3.1 3.2 3.3 3.4 3.5 3.6 Household Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Age-sex Distribution of the Household Population . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Marital Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Educational Attainment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Housing Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Presence of Household Durable Goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 i i i 4 6 7 8 Page 3.7 3.8 Background Characteristics of Female Respondents . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Exposure to Mass Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 FERTILITY by Fred Arnold and Mehboob Sultan 4.1 4.2 4.3 4.4 4.5 Fertility Levels and Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Children Ever Born . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Birth Intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Age at First Birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Teenage Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 KNOWLEDGE AND USE OF FAMILY PLANNING by Nasra M. Shah and Syed Mubashir Ali 5.1 5.2 5.3 5.4 5.5 5.6 Knowledge of Family Planning Methods and Sources . . . . . . . . . . . . . . . . . . . . . . . . . 53 Contraceptive Use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Knowledge of the Fertile Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Age at Sterilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Source of Supply and Accessibility of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . 68 Cost of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 FAMILY PLANNING ATTITUDES by Abdul Razzaque Rukanuddin and Mehboob Sultan 6.1 6.2 6.3 6.4 6.5 6.6 6.7 Discussion of Family Planning Among Couples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Attitudes of Couples Toward Family Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Family Planning Messages on Electronic Mass Media . . . . . . . . . . . . . . . . . . . . . . . . . 77 Acceptability of Family Planning Messages on Electronic Mass Media . . . . . . . . . . . 79 Intentions About Future Use of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Reasons for Nonuse of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Preferred Future Method of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 PROXIMATE DETERMINANTS OF FERTILITY by Zeba Sathar and Tauseef Ahmed 7.1 7.2 Marriage Patterns and Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 Breastfeeding and Postpartum Infecundibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 FAMILY SIZE PREFERENCES by Syed Mubashir Ali and Abdul Razzaque Rukanuddin 8.1 8.2 8.3 8.4 8.5 Desire for Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Sex Preference for Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Need for Family Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Ideal Family Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Fertility Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 iv Page INFANT AND CHILD MORTALITY by Tauseef Ahmed, Mansoor-ul-Hassan Bhatti and George Bicego 9.1 9.2 9,3 9.4 9.5 9.6 Definitions of Infant and Child Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Assessment of Data Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Levels and Trends in Infant and Child Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Differentials in Infant and Child Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 Causes of Death in Early Childhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 High-risk Fertility Behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121 10 MATERNAL AND CHILD HEALTH by Abdul Razzaque Rukanuddin and K, Zaki Hasan 10.1 10.2 Maternal Care Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Child Care Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 11 FEEDING PATTERNS AND THE NUTRITIONAL STATUS OF CHILDREN by Tauseef Ahmed and Mohammad Ayub 11.1 11.2 Breastfeeding and Nutritional Intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Nutritional Status of Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 12 HUSBANDS' SURVEY by Sultan S. Hashmi 12.1 12.2 12.3 12.4 12.5 12.6 12.7 Background Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Knowledge and Use of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Prospective Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Approval of Family Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Acceptability of Media Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Fertility Desires and Sex Preference for Childrcn . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Ideal Number of Children for Husbands and Wives . . . . . . . . . . . . . . . . . . . . . . . . . . 186 APPENDICES Appendix A Appendix B Appendix C Appendix D Pakistan Demographic and Health Survey Staff . . . . . . . . . . . . . . . . . . . . . . . . 189 Estimates of Sampling Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197 Data Quality Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215 Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225 V Table 1.1 Table 2.1 Table 2.2 Table 2.3 Table 2A I able 2.5 Table 3.1 Table 3.2 Table 3.3 Table 3.4 Table 3.5 Table 3.6 Table 3.7 Tab'e 3.8 Table 3.9 Table 3.10 Table 3.11 Table 4.1 Table 4.2 "Iable 4.3 Table 4.4 Table 4.5 Table 4.6 Table 4.7 TABLES Page Population size and distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Sample coverage for urban and rural areas combined . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Sample coverage for urban areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Sample coverage for rural areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Results of the household and individual interviews for the women's sample . . . . . . . . . 16 Results of the household and individual interviews for the husbands' sample . . . . . . . . 17 Household composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Household population by age, residence and sex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Population by age from selected sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Household population by age, sex and marital status . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Educational level of household population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 School enrolment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Housing characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Household durable goods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Background characteristics o f female respondents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Level of education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Exposure to mass media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Crude birth rates according to selected surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Current fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 Current fertility according to selected surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Fertility by background characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 Fertility trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Fertility by marital duration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Children ever born and living . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 vii Table 4.8 Table 4.9 Table 4.10 Table 4.11 Table 4.12 Table 4.13 Table 4.14 Table 5.1 Table 5.2 Table 5.3 Table 5.4 Table 5.5 Table 5.6 Table 5.7 Table 5.8 Table 5.9 Table 5.10 Table 5.11 Table 5.12 Table 5.13 Table 5.14 Table 5.15 Table 6.1 Table 6.2 Page Children ever born according to selected surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Mean number of children ever born . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Birth intervals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Age at first birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 Age at first birth by background characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 Teenage fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 Children ever 10ore to teenagers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 Knowledge and source of contraceptive methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Trends in contraceptive knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Knowledge of modem contraceptive methods and source o f methods . . . . . . . . . . . . . . 56 Ever use of contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 Current use of contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Use of contraception by non-pregnant women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Trends in contraceptive use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Current use of contraception by background characteristics . . . . . . . . . . . . . . . . . . . . . . 63 Number of children at first use of contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Brand names of condoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Knowledge of fertile period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 Timing of sterilisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 Source of supply . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 T ime to source of supply for modem contraceptive methods . . . . . . . . . . . . . . . . . . . . . 70 Costs of contraceptive methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Discussion of family planning by couples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Attitudes of couples toward family planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 viii Table 6.3 Table 6.4 Table 6.5 Table 6.6 Table 6.7 Table 6.8 Table 7.1 Table 7.2 Table 7.3 Table 7.4 Table 7.5 Table 7.6 Table 7.7 Table 7.8 Table 7.9 Table 7.10 Table 8.1 Table 8.2 Table 8.3 Table 8.4 Table 8.5 Table 8.6 Table 8.7 Table 8.8 Page Family planning messages on radio and television . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Perceived effectiveness of mass media messages on use of family planning . . . . . . . . . 78 Acceptability of media messages on family planning . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Future contraceptive use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Reasons for not intending tu use contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 Preferred method of contraception for future use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Current marital status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Mean age at marriage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 Age at first marriage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Median age at first marriage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Ideal age at marriage for women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Marriage between relatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 Polygyny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Breasffeeding, postpartum amenorrhoea, abstinence and insusceptibility . . . . . . . . . . . . 93 Median duration of postpartum insusceptibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Termination of exposure to the risk of pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Fertility preferences by number of living children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Fertility preferences by age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Desire to limit childbearing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Preferred sex of next child . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Need for family planning services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Ideal number of children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Mean ideal number of children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 Planning status of births . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 ix Table 8.9 Table 9.1 Table 9.2 Table 9.3 Table 9.4 Table 9.5 Table 9.6 Table 9.7 Table 9.8 Table 10.1 Table 10.2 Table 10.3 Table 10.4 Table 10.5 Table 10.6 Table 10.7 Table 10.8 Table 10.9 Page Wanted fertility rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 Indices of early infant deaths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Sex differentials in infant and neonatal mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Reporting of age at death . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Infant and child mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Infant and child mortality by background characteristics . . . . . . . . . . . . . . . . . . . . . . . 116 Infant and child mortality by demographic characteristics . . . . . . . . . . . . . . . . . . . . . . 118 Causes of death in early childhood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 High-risk fertility behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 Antenatal care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Number of antenatal care visits and stage of pregnancy . . . . . . . . . . . . . . . . . . . . . . . . 128 Tetanus toxoid vaccination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129 Place of delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 Assistance during delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131 Characteristics of delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 Vaccinations by source of information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 Vacc inat ionsbybackgroundchamcter i s t i cs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Vaccinations in first year of life by current age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Table 10.10 Prevalence and treatment of acute respiratory infection . . . . . . . . . . . . . . . . . . . . . . . . 138 Table 10.11 Prevalence and treatment of fever . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Table 10.12 Prevalence o f diarrhoea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 Table 10.13 Knowledge and use of ORS packets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 Table 10.14 Treatment of diarrhoea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 Table 10.15 Feeding practices during diarrhoea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 X Table 11.1 Table 11.2 Table 11.3 Table 11.4 Table 11.5 Table 11.6 Table 11.7 Table 11,8 Table 11.9 )'age Initial breastfeeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150 Liquids and food items given before breastfeeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 Nutritional intake . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 Breastfeeding and supplementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Type of supplementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Median duration and frequency of breastfeeding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 Age at which liquids and foods were introduced . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Reasons fors topp ingbreast feed ing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Nutritional status by demographic characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 Table 11.10 Nutritional status by background characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Table 12.1 Husbands' background characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Table 12.2 Age difference between spouses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171 Table 12.3 Husband's level of education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 Table 12.4 Knowledge of contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Table 12.5 Knowledge of contraception among couples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 Table 12.6 Knowledge and use of modem methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Table 12.7 Current use of contraception reported by couples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178 Table 12.8 Reasons for not intending to use contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Table 12.9 Preferred method of contraception for future use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180 Table 12.10 Wife's perception of husband's attitude about family planning . . . . . . . . . . . . . . . . . . 181 Table 12.11 Husband's perception ofwife's attitude about family planning . . . . . . . . . . . . . . . . . . 181 Table 12.12 Acceptability of mass media messages on family planning . . . . . . . . . . . . . . . . . . . . . 182 Table 12.13 Perceived effectiveness of mass media messages on family planning . . . . . . . . . . . . . 183 Table 12.14 Reproduction intentions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 xi Page Table 12.15 Desire to limit future births . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 Table 12.16 Desire for more children by wives and their husbands . . . . . . . . . . . . . . . . . . . . . . . . . 185 Table 12.17 Ideal number of children of wives and their husbands . . . . . . . . . . . . . . . . . . . . . . . . . 187 Appendix B Table B.1 List of selected variables for sampling errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 Table B.2 Sampling errors: Entire sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Table B.3 Sampling errors: Urban areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 TableB.4 Sampling errors: Major cities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204 Table B.5 Sampling errors: Other urban areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205 Table B.6 Sampling errors: Rural areas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 206 Table B.7 Sampling errors: Punjab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207 Table B.8 Sampling errors: Sindh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208 Table B.9 Sampling errors: NWFP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 Table B.10 Sampling errors: Balochistan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210 Table B.I1 Sampling errors: Age group 15-24 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211 Table B.12 Sampling errors: Age group 25-34 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212 Table B.13 Sampling errors: Age group 35-49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213 Appendix C Table C.1 Age distribution of household population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217 Table C.2 Age distribution of eligible and interviewed women . . . . . . . . . . . . . . . . . . . . . . . . . . . 219 Table C.3 Completeness of reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220 Table C.4 Births by calendar year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 221 Table C.5 Reporting of age at death in days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222 Table C.6 Reporting of age at death in months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223 xii N~3.1 Ngure3.2 Figure3.3 Figure4.1 ~gure4 .2 ~gure5.1 Figure 5.2 Figure 5.3 Figure 5.4 Figure 6.1 ~gure7.1 Figure 8.1 Figure 8.2 Figure 9.1 Figure 9.2 Figure 9.3 Figure 10.1 Figure 10.2 FIGURES Population pyramid of Pakistan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 School enrolment by age and place of residence . . . . . . . . . . . . . . . . . . . . . . . . 26 Distribution of currently married women by age, Pakistan 1975-1991 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Age-Specific fertility rates, Pakistan 1970-1991 . . . . . . . . . . . . . . . . . . . . . . . . 40 Total fertility rate (TFR) and mean number of children ever bom (CEB) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Knowledge of modem contraceptive methods and sources among currently married women by residence . . . . . . . . . . . . . . . . . . . . . . . . . 57 Current use of modem contraceptive methods among currently married women 15-49 by residence and province . . . . . . . . . . . . . . . . 64 Current use of modem contraceptive methods among currently married women 15-49 by education . . . . . . . . . . . . . . . . . . . . . . . . . . 65 Sources of family planning among current users of modem contraceptive methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 Approval of family planning among currently married women by residence and education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 Marriage between relatives and between nonrelatives among ever-married women 15-49, Pakistan and major cities . . . . . . . . . . . . . . . 91 Fertility preferences among currently married women 15-49 . . . . . . . . . . . . . . . . 99 Fertility preferences among currently married women 15-49 by number of living children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 Distribution of deaths under five by age at death . . . . . . . . . . . . . . . . . . . . . . 115 Infant and child mortality by place of residence . . . . . . . . . . . . . . . . . . . . . . . 117 Infant mortality by mother's age at first birth . . . . . . . . . . . . . . . . . . . . . . . . . 119 Antenatal care, place of delivery, and assistance during delivery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127 Percentage of children 12-23 months who are fully vaccinated . . . . . . . . . . . . . 136 xiii Figure 10.3 Figure 10.4 Figure 10.5 Figure 11.1 Figure 11.2 Figure 11.3 Figure 11.4 Figure 11.5 Figure 11.6 Figure 12.1 Figure 12.2 Figure 12.3 Figure 12.4 Page Prevalence of ARI symptoms and fever in the two weeks preceding the survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Knowledge and use of ORS by residence and province . . . . . . . . . . . . . . . . . . 143 Percentage of children receiving treatment for diarrhoea by type of treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 Percentage of children given selected liquids before being put to the breast by province . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Percentage of children given water, milk, or solid/mushy food the day before the interview . . . . . . . . . . . . . . . . . . . . . . . . 154 Breastfeeding among children age 0-23 months . . . . . . . . . . . . . . . . . . . . . . . 156 Age at which liquids and solid/mushy food were first given to children age 24-59 months . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 Percentage of children under five who are underweight by age . . . . . . . . . . . . . 165 Undemutrition among children under five years of age . . . . . . . . . . . . . . . . . . 166 Age difference between husbands and wives . . . . . . . . . . . . . . . . . . . . . . . . . 171 Percentage of couples in which both the husband and wife know specific contraceptive methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 175 Knowledge of modem contraceptive methods, knowledge of sources, and current use among husbands by province . . . . . . . . . . . . . . . . . 177 Desire for more children among husbands and wives by number of living children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 xiv PREFACE One of the major objectives of the National Institute of Population Studies (NIPS) is to assist the Ministry of Population Welfare by assessing and evaluating the Population Welfare Programme. The Pakistan Demographic and Health Survey (PDHS) has been one of the major activities in this context. This survey is a part of the worldwide exercise to assess the changing demographic and health situation through 60 surveys in different countries. This effort is being coordinated by the Demographic and Health Surveys programme of IRD/Macro International Inc., Columbia, Maryland. Planning for the PDHS began in June 1990 and the survey was executed between December 1990 and May 1991. While the data processing was done simultaneously with the fieldwork, the tabulation was done during June and July 1991 and the preliminary report was released in August 1991. This survey, which is a follow-up of the Pakistan Contraceptive Prevalence Survey undertaken during 1984-85, was given high priority in the work plan of the Institute. The survey provides us with an up-to-date set of relevant data useful to evaluate population, health and family planning programmes and to assess the overall demographic situation in the country. The results produced in this report provide social scientists, policy makers, and planners with a clear picture about thc current level of demographic and health indicators and trends in the recent past and illuminate the likely direction for the future. The importance of the PDHS lies in the fact that it provides basic resource material for the Eighth Five-Year Plan presently being formulated. For undertaking this arduous task, the Institute cooperated with IRD/Macro International Inc., which provided technical assistance and data processing equipment. The Federal Bureau of Statistics provided assistance in the selection of the sample and the fieldwork. The United States Agency lor International Development (USAID) and the Government of Pakistan provided financial assistance and staff for the execution of this activity. To all these agencies, NIPS is highly indebted. Since this project is a research exercise, there could be differences in the findings of this survey and other data available from different sources. This is an understandable situation; questions have especially been raised about the data on immunisations, the contraceptive method mix and fertility estimates. The users of these data may use caution while interpreting these differences and may draw their own conclusions. Those who actually worked on the project from its inception to its completion deserve special appreciation. I am also thankful to Mr. K. U. Faruqui and Mr. K. M. Chima for their interest, support and guidance during their stay at NIPS as Project Director. We also acknowledge the services of Dr. A. Ghayur who worked as Principal Investigator for the initial part of the project. Islamabad 1992 M. S. Jillani, Ph.D. Executive Director XV ACKNOWLEDGMENTS The Pakistan Demographic and Health Survey (PDHS) was completed as a part of an international exercise undertaken in collaboration with IRD/Macro Intemational Inc. (IRD), Columbia, Maryland. The PDHS was conducted in order to update information on human reproduction, infant and child mortality, contraceptive use, maternal and child care, and the nutritional status of children. The major objective of the PDHS was to provide a data base for evaluating programme efforts and developing strategies and plans for the furore programme. As in the case of all previous demographic surveys, the successful completion depended on a joint effort of a number of organizations and individuals. The National Institute of Population Studies undertook the responsibility of implementing the project only when IRD provided assistance in terms of technical knowhow and equipment, USAID/Islamabad fully financed it, and the Federal Bureau of Statistics (FBS) provided the sampling frame, field supervisory personnel and almost all logistical support for the fieldwork. Conducting the fieldwork was a huge task and all activities were accomplished on time only with the dedicated, relentless and devoted efforts of the PDHS staff at the headquarters, data entry staff, the office editors, able FBS supervisors and drivers and all our field teams. Many thanks to each one of them for undertaking an enormous amount of hardship during the winter and maintaining the speed and efficiency of work during the month of Ramadan and during unexpected rains from February to April 1991. We are deeply indebted to Dr. M. S. JiUani, former Secretary of the Ministry of Population Welfare, for the guidance and personal interest needed to maintain the speed of the project. The timely release of funds by him was definitely a big boost for all. Later, his support as a Project Director worked as a catalyst for finalizing and releasing the preliminary and final report of the PDHS. The PDHS was undertaken in several stages: establishment of the office, questionnaire design and modification, sample design, pretesting of the questionnaire, training of field teams, fieldwork, questionnaire editing, and data entry. As an integral part of the survey activities, every activity was appraised and guided by the Technical Advisory Committee. This is to acknowledge the efforts and timely advice of the members of the Committee in the successful completion of the survey. Many thanks are due to the late Mr. S. M. Ishaque, former Director Genera/, FBS for his professional association and help to the PDHS. We also acknowledge with deep gratitude all the moral and logistic support that we received from the offices of all provincial Director Generals, Population Welfare Departments and a large number of District Population Welfare Officers in conducting our field survey in very remote areas of Pakistan. Special thanks are also due to the reviewers of various chapters which include Ms. Anne R. Cross, Dr. Elisabeth Sommerfelt, Dr. Kate Stewart and Dr. Fred Arnold. We would also like m thank the following IRD personnel for their assistance in producing tabulations, designing the survey and the sample, and editing and typing the report: Mr. Noureddine Abdermhim, Dr. Alfredo Aliaga, Dr. Ties Boerma, Mr. Trevor Croft, Ms. Jeanne Cushing, Ms. Thanh Le, Dr. A.M. Marckwardt, Ms. Kaye Mitchell, Dr. Sidney Moore, Mr. Luis Ochoa, Mr. Guillermo Rojas, Dr. Jeremiah Sullivan, and Ms. Jane Weymouth. The amount of work put in by Dr. S. S. Hashmi and Dr. A. Razzaque Rukanuddin to review and finalize this report needs to be fully xvii acknowledged and appreciated. Finally, we would like to acknowledge with deep gratitude and thanks the relentless and committed efforts of Dr. Fred Arnold who provided immense moral support and technical assistance at each stage of the project. Tauseef Ahmed, Ph.D. Principal Investigator xviii SUMMARY OF FINDINGS The Pakistan Demographic and Health Survey (PDHS) was fielded on a national basis between the months of December 1990 and May 1991. The survey was carried out by the National Institute of Population Studies with the objective of assisting the Ministry of Population Welfare to evaluate the Population Welfare Programme and maternal and child health services. The PDHS is the latest in a series of surveys, making it possible to evaluate changes in the demographic status of the population and in health conditions nationwide. Earlier surveys include the Pakistan Contraceptive Prevalence Survey of 1984-85 and the Pakistan Fertility Survey of 1975. Until recently, fertility rates had remained high with little evidence of any sustained fertility decline. In recent years, however, fertility has begun to decline due to a rapid increase in the age at marriage and to a modest rise in the prevalence of contraceptive use. The lotal fertility rate is estimated to have fallen from a level of approximately 6.4 children in the early 1980s to 6.0 children in the mid-1980s, to 5.4 children in the late 1980s. The exact magnitude of the change is in dispute and will be the subject of further research. Important differentials of fertility include the degree ofurbanisation and the level of women's education. The total fertility rate is estimated to be nearly one child lower in major cities (4.7) than in rural areas (5.6). Women with at least some secondary schooling have a rate of 3.6, compared to a rate of 5.7 children for women with no formal education. There is a wide disparity between women's knowledge and use of contraceptives in Pakistan. While 78 percent of currently married women report knowing at least one method of contraception, only 21 percent have ever used a method, and only 12 percent are currently doing so. Three-fourths of current users are using a modem method and one-fourth a traditional method. The two most commonly used methods are female sterilisation (4 percent) and the condom (3 percent). Despite the relatively low level of contraceptive use, the gain over time has been significant. Among married non-pregnant women, contraceptive use has almost tripled in 15 years, from 5 percent in 1975 to 14 percent in 1990-91. The contraceptive prevalence among women with secondary education is 38 percent, and among women with no schooling it is only 8 percent. Nearly one-third of women in major cities arc current users of contraception, but contraceptive use is still rare in rural areas (6 percent). The Government of Pakistan plays a major role in providing family planning services. Eighty-five percent of sterilised women and 81 percent of IUD users obtained services from the public sector. Condoms, however, were supplied primarily through the social marketing programme. The use of contraceptives depends on many factors, including the degree of acceptability of the concept of family planning. Among currently married women who know of a contraceptive method, 62 percent approve of family planning. There appears to be a considerable amount of consensus between husbands and wives about family planning use: one-third of female respondents reported that both they and their husbands approve of family planning, while slightly more than one-fifth said they both disapprove. The latter couples constitute a group for which family planning acceptance will require concerted motivational efforts. The educational levels attained by Pakistani women remain low: 79 percent of women have had no formal education, 14 percent have studied at the primary or middle school level, and only 7 percent have attended at least some secondary schooling. The traditional social structure of Pakistan supports a natural fertility pattern in which the majority of women do not use any means of fertility regulation. In such populations, the proximate determinants of fertility (other than contraception) are crucial in determining xix fertility levels. These include age at marriage, breastfeeding, and the duration of postpartum amenorrhoea and abstinence. The mean age at marriage has risen sharply over the past few decades, from under 17 years in the 1950s to 21.7 years in 1991. Despite this rise, marriage remains virtually universal: among women over the age of 35, only 2 percent have never married. Marriage patterns in Pakistan are characterised by an unusually high degree of consangninity. Half of all women are married to their first cousin and an additional 11 percent are married to their second cousin. Breasffeeding is important because of the natural immune protection it provides to babies, and the protection against pregnancy it gives to mothers. Women in Pakistan breastfeed their children for an average of20months. Themeandurationofpostpartumamenorrhoeais slightly more than 9 months. After tbebirth of a child, women abstain from sexual relations for an average of 5 months. As a result, the mean duration of postpartum insusceptibility (the period immediately following a birth during which the mother is protected from the risk of pregnancy) is 11 months, and the median is 8 months. Because of differentials in the duration of breastfeeding and abstinence, the median duration of insusceptibility varies widely: from 4 months for women with at least some secondary education to 9 months for women with no schooling; and from 5 months for women residing in major cities to 9 months for women in rural areas. In the PDHS, women were asked about their desire for additional sons and daughters. Overall, 40 percent of currently married women do not want to have any more children. This figure increases rapidly depending on the number of children a woman has: from 17 percent for women with two living children, to 52 percent for women with four children, to 71 percent for women with six children. The desire to stop childbearing varies widely across cultural groupings. For example, among women with four living children, the percentage who want no more varies from 47 percent for women with no education to 84 percent for those with at least some secondary education. Gender preference continues to be widespread in Pakistan. Among currently married non-pregnant women who want another child, 49 percent would prefer to have a boy and only 5 percent would prefer a girl, while 46 percent say it would make no difference. The need for family planning services, as measured in the PDHS, takes into account women's statements concerning recent and future intended childbearing and their use of contraceptives. It is estimated that 25 percent of currently married women have a need for family planning to stop childbearing and an additional 12 percent are in need of family planning for spacing children. Thus, the total need for family planning equals 37 percent, while only 12 percent of women are currently using contraception. The result is an unmet need for family planning services consisting of 25 percent of currently married women. This gap presents both an opportunity and a challenge to the Population Welfare Programme. Nearly one-tenth of children in Pakistan die before reaching their first birthday. The infant mortality rate during the six years preceding the survey is estimaled to be 91 per thousand live births; the under-five mortality rate is 117 per thousand. The under-five mortality rates vary from 92 per thousand for major cities to 132 for rural areas; and from 50 per thousand for women with at least some secondary education to 128 for those with no education. The level of infant mortality is influenced by biological factors such as mother's age at birth, birth order and, most importantly, the length of the preceding birth interval. Children born less than two years after their next oldest sibling are subject to an infant mortality rate of 133 per thousand, compared to 65 for those spaced two to three years apart, and 30 for those born at least four years after their older brother or sister. XX One of the priorities of the Government of Pakistan is to provide medical care during pregnancy and at the time of delivery, both of which are essential for infant and child survival and safe motherhood. Looking at children born in the five years preceding the survey, antenatal care was received during pregnancy for only 30 percent of these births. In rural areas, only 17 percent of births benefited from antenatal care, compared to 71 percent in major cities. Educational differentials in antenatal care are also striking: 22 percent of births of mothers with no education received antenatal care, compared to 85 percent of births of mothers with at least some secondary education. Tetanus, a major cause of neonatal death in Pakistan, can be prevented by immunisation of the mother during pregnancy. For 30 percent of all births in the five years prior to the survey, the mother received a tetanus toxoid vaccination. The differentials are about the same as those for antenatal care generally. Eighty-five percent of the births occurring during the five years preceding the survey were delivered at home. Sixty-nine percent of all births were attended by traditional or trained birth attendants, while 19 percent were assisted by a doctor or nurse. The Expanded Programme on Immunisatlon in Pakistan has met with considerable success. Among children 12 to 23 months of age, 70 percent had received a BCG vaccination, 50 percent a measles vacci- nation, and 43 percent had received all three doses of DPT and polio vaccine. Only 35 percent, however, had received all of the recommended vaccinations, while 28 percent had received none at all. Thirty-nine percent of boys were fully protected, compared to 31 percent of girls. Sixteen percent of children under the age of five had been ill with a cough accompanied by rapid breathing during the two weeks preceding the survey. Children 6-11 months old were most prone to acute respiratory infections (23 percent). Two-thirds (66 percent) of children who were sick were taken to a health facility or provider. All but 15 percent of the sick children received some kind of treatment. About the same proportion of children (15 percent) had suffered from diarrhoea in the two weeks preceding the survey, with the highest incidence among children under two years of age. Nearly half (48 percent) were taken to a health facility or provider. About two of five (39 percent) children with diarrhoea were treated with oral rehydration solution prepared from ORS packets. Knowledge of oral rehydration therapy is widespread: 90 percent of mothers recognise ORS packets. Nearly two-thirds (63 percent) of mothers have used ORS packets at some time, and among these, three-quarters had mixed the solution correctly the last time they prepared it. Thirty percent of children had suffered from fever in the two weeks preceding the survey. Those most prone to illness were age 6 to 11 months. Two-thirds of children with fever were taken to a health facility or provider. Inadequate nutrition continues to be a serious problem in Pakistan. Fifty percent of children under five years of age suffer from stunting (an indicator of chronic undemutrition), as measured by height for age. The prevalence of stunting increases with age, from 16 percent for children under 6 months to 63 percent of four-year olds. The lowest prevalence is found in Punjab (44 percent), and the highest in Balochistan (71 percent). The mother's level of education is an important factor; the prevalence of stunting varies from 18 percent for mothers with some secondary education to 56 percent for mothers with no education. Acute undemutrition, low weight for height, is less of a problem in Pakistan than chronic undemutrition. Nine percent of children suffer from acute undemutrition (wasting). The prevalence of wasting does not vary substantially between geographic groupings. The largest differential is for mother's xxi education: 4 percent of children of mothers with some secondary school or higher education are wasted, compared to 10 percent of children of mothers with no schooling. A systematic subsample of households in the women's survey was selected to obtain information from the husbands of currently married women. The focus was on obtaining information about attitudes, behaviour, and the role of husbands regarding family planning. Husbands' responses concerning knowledge and use of contraception were remarkably similar to women's responses: about four-fifths knew of at least one method, two-thirds knew of a source of supply, one-fourth reported that they and their spouses had used contraception sometime in the past, and about one-seventh were current users. Although a majority of husbands (56 percent) approve of family planning, wives are more likely to favour family planning than their husbands. Since husbands usually have a predominant role in family decision making, the family planning programme should increase efforts to educate and motivate husbands. xxii PAKISTAN ~-- ~ CHINA GILGIT AFGHANISTAN PE$HAWAR ISLAMABAD JAMMU & KASHMIR ~SPUT~ AREA) LAHORE • QUETTA PUNJAB IRAN BALOCHISTAN INDIA SINDH [11 KARACHI ARABIAN SEA "~ xxiv CHAPTER 1 INTRODUCTION Abdul Razzaque Rukanuddin and Tauseef Ahmed This report gives the major findings of the Pakistan Demographic and Health Survey (PDHS) conducted from December 1990 to May 1991 on a nationwide basis. After the preliminary report, published in August 1991, this is the first in a series of reports on the PDHS findings with the objective of improving the Population Welfare Programme and health services in Pakistan. Besides presenting results at the national level, this report presents information by urban-rural areas and by province. Before presenting the major findings, this chapter discusses the physical features, people, culture, religion, language, population distri- bution and size, fertility and mortality levels, literacy and educational attainment, economy, population and family planning and health policies and programmes of the country. The objective of this presentation is to make the reader familiar with the historical, geographic, socioeconomic and demographic features of the country. 1.1 Physical Features Pakistan, situated in the northwestern part of the South Asian subcontinent, obtained independence from the British on August 14, 1947 after the subdivision of the Indian subcontinent. It is a land mass of diversified relief with vast plains in the Indus basin, a rocky expanse of plateaus in the southwest and majestic mountains in the north with beautiful valleys, snow-covered peaks and glaciers. Pakistan extends from 24 ° to 37°N latitude and from 61 ° to 75°E longitude. On its east and southeast lies India, to the north and northwest is Afghanistan, to the west is Iran and in the south, the Arabian Sea. It has a common frontier with China on the border of its Gilgit Agency. Tajikistan, formerly in the USSR, is separated from Pakistan by a narrow strip of Afghan territory called Wakhan. This variety of landscape divides Pakistan into six major regions: the Northern High Mountainous Region, the Western Low Mountainous Region, the Balochistan Plateau, the Potohar Uplands, and the Punjab and Sindh fertile plains. Pakistan is a land of great rivers like the Indus and its tributaries, large dams like Tarbela, and high mountain peaks like K2 (Mount Goodwin Austin - 8,611 metres) and Nanga Parbat (8,126 metres). 1.2 Climate, Rainfall, and Seasons Pakistan has a continental type of climate, characterized by extreme variations of temperature depending on the topography of the country. Pakistan experiences a general deficiency of rainfall. Although it is in the monsoon region, it is arid, except for the southem slopes of the Himalayas and the submountalnous tract where the annual rainfall varies between 76 and 127 cm. Balochistan is the driest part of the country with an average rainfall of 21 cm. There are four well-marked seasons in Pakistan, namely: 1. Cold season (December to March) 2. Hot season (April to June) 3. Monsoon season (July to September) 4. Post-monsoon season (October to November). 1.3 Administrative Divisions The total land area of Pakistan is about 796,000 square kilometres. Pakistan is comprised of the provinces of Punjab, Noah West Frontier, Balochistan and Sindh and the Federally Administered Tribal Areas (FATA) of the north and northwest (see map, page xxiv). Each province is divided into administrative divisions, districts, tehsils and talukas. There were 16 divisions and 72 districts in the country in 1991. Islamabad, the capital of Pakistan, which lies in the northem part of the country at the bottom of the Margala hills near Rawalpindi, is a well-planned city which was constructed beginning in the 1960s. 1.4 People, Culture, Religion, and Language Pakistan historically attracted migrants from many nations in the northwest and the northeast. These include Dravidians, Aryans, G reeks, Turks, Persians, Afghans, Arabs and Mughals. The dominant racial type in Pakistan is Indo-Aryans. In the cultural arena, Pakistan has inherited a rich heritage. A highly developed way of life was attained by the people of Pakistan in the Indus Valley Civilization about 5000 years ago which came to an end around 1500 B.C. About 500 B.C., the northern city of Taxila emerged as a famous centre of Buddhist leaming and culture which existed for a thousand years. Pakistan is an ideological state which came into existence as a result of the demand for a separate homeland for the Muslims of the Indian subcontinent. The Muslim majority areas were mostly carved out into Pakistan. Therefore, the large majority of the population of Pakistan is comprised of Muslims. A negligible minority of Hindus is settled mainly in the border districts of Sindh. Christians are widely spread throughout the country and form about three percent of the total population. The Parsis (Zoroastrians), whn number about 20,000, are an economically notable minority, mostly settled in Karachi. The constitution of Pakistan guarantees the right of minorities to profess, practice and propagate their religion and every administrative position is open to them with the exception of the Head of State and the Prime Minister. Urdu is the language most commonly spoken throughout the country. Balochi and Brohi are spoken in most of Balochistan, Pushto in North West Frontier Province (NWFP) and also in some parts of Balochistan, Punjabi in Punjab, and Sindhi in the Province of Sindh. Saraiki is widely spoken in southern Punjab in the districts of Multan, Bahawalpur and Dera Ghazi Khan and adjoining areas in Balochistan, NWFP and Sindh. The medium of education is Urdu but English continues to be used in higher education and professional colleges, particularly in scientific and technical fields. English is widely used for commercial, legal and other official business in the country. 1.5 Population Population Size Pakistan is the ninth most populous country in the world after China, India, the former USSR, USA, Indonesia, Brazil, Japan, and Nigeria. The population of Pakistan was 16.6 million at the beginning of the twentieth century (in 1901). By the time of independence in 1947, the population was estimated to have doubled to 32.5 million. In the first decennial census (1951), the population of Pakistan was reported to be 33.8 million while in the last decennial census in 1981 the population size was 84.3 million (see Table 1.1). In 1991, the population was estimated to be around 115 million with males comprising 52.5 percent of the population. The sex ratio of the population is estimated to be 111 males per 100 females. Since independence, the population has increased at an average growth rate of 2.9 percent per annum. The present growth rate of the population is estimated to be around three percent. 2 Table 1.1 Population size and distribution Distribution of population, intercensal change and average annual growth ra~ of population by residence. Pakistan, 1951-1981 Census year Area 1951 1961 1972 1981 Urban 6,019 Rural 27,798 Total 33,817 Urban 17.8 Rural 82.2 Total 100.0 Urban Rural Total Urban Rural Total Population (in OOOs) 9,655 16,594 23,841 33,324 48,727 60,412 42,978 65,321 84,254 Percent distribution 22.5 25.4 28,3 77.5 74.6 71.7 100.0 100.0 100.0 Intercensal percent change 60.4 71.9 43.7 19.9 46.2 24.0 27.1 52.0 29.0 Average annual growth rate 4.88 4.77 4.38 1.84 3.32 2.58 2.45 3.67 3.06 Source: Population Census Orgadisation (1985). Population Distribution The population of Pakistan is unevenly distributed among its various provinces. Punjab is the most densely populated province with about one-quarter (26 percent) of the total land area of the country and more than half (56 percent) of the total population. The next most densely populated provinces are Sindh, with less than one-fifth (18 percent) of the land area and 23 percent of the total population and North West Frontier Province (NWFP) and the Federally Administered Tribal Area (FATA) with 13 percent of the land area and 16 percent of the total population. Balochistan, which is the largest province by area (with 44 percent of the total land area), has the lowest proportion of Pakistan's total population (5 percent). The population density in the country increased from 43 persons per square kilometre in 1951 to 106 persons per square kilometre in 1981 and further to around 145 persons per square kilometre in 1991. Urban-Rural Distribution Pakistan is predominantly an agricultural country with just over 50 percent of the work force employed in occupations related to agriculture. The 1981 Census reported that 72 percent of the total population lived in rural areas. However, urban growth over the years has been dramatic. The proportion 3 urban increased from 18 percent in 1951 to 28 percent in 1981. In terms of absolute numbers, the urban population nearly quadrupled from 6.0 million in 1951 to 23.8 million in 1981. However, the intercensal average annual growth rate of the urban population declined from 4.9 percent for the period 1951-61 to 4.4 percent for the period 1972-81, primarily due to a change in the definition of urban areas (see Table 1.1). 1.6 Fertility Several attempts have been made in Pakistan to estimate fertility rates through direct as well as indirect techniques. A number of estimates have been made based on different sets of data, methods and assumptions. Given the trend in population growth, the inevitability of fertility as an important focus of population studies cannot be overemphasized. But a major problem in Pakistan is the wide variations in fertility estimates derived from different sets of data (Rukanuddin and Farooqui 1988), reflecting problems in data inconsistency due to methodological and procedural differences. For instance, the direct fertility estimates based on the 1975 Pakistan Fertility Survey and the 1984-85 Pakistan Contraceptive Prevalence Survey are lower than the indirect estimates based on the Population Growth Surveys (conducted between 1968 and 1979) and the Pakistan Demographic Surveys (conducted annually since 1984). However, prior demographic surveys confirm the persistence of a high level of fertility in Pakistan but with a gradual decline over time. The principal decline has been observed for the younger age groups and is attributed primarily to an increase in the age at marriage. Changes over time in other proximate determinants of fertility in Pakistan such as contraceptive use and breastfeeding are less conducive to lower fertility. Since 1974, surveys have estimated the crude birth rate to vary from 37 to 43 per thousand population and the total fertility rate to range between 5.9 and 6.9 children per woman. 1.7 Mortality In Pakistan, the systematic study of trends, levels and differentials in mortality is impeded by a lack of reliable data. Although a system of vital registration has been in existence in the country since the last quarter of the 19th century, the recorded data suffer from errors in coverage and inaccuracies in the information provided. It is estimated that at the time of independence, the crude death rate (CDR) was around 25 to 30 per thousand population. The decline in mortality after the Second World War has been very rapid, with the CDR falling to about 10 to 12 deaths per thousand in the 1980s. This has been due inter alia to improvements in the availability of food through higher levels of production, the effective control of procurement and distribution of food grains, and the increasing pace of socioeconomic development. Epidemics have also been eliminated and diseases brought under control with the development of effective public health measures and medical services such as inoculation and vaccination programmes. The infant mortality rate was around 150 to 180 deaths per thousand live births at the time of independence in 1947. This has declined to less than 100 in 1991, mainly due to improved health services and a successful immunisation programme. Available evidence suggests that slightly more than one-third (36 percent) of all deaths occur during infancy in Pakistan. Moreover, one-third of all infant deaths occur within one week of birth. An additional 22 percent of deaths occur in the second to fourth week. In other words, more than half of infant deaths are neonatal deaths that occur within four weeks of birth. Much could be done to eliminate some of the causes of neonatal deaths such as short birth intervals and high parity births. Maternal deaths, associated with complications of pregnancy and childbirth, are quite high. Four of five deliveries are attended by traditional birth attendants or elderly women. Repeated and closely spaced pregnancies and births coupled with high parity pregnancies are found to result in a high incidence of maternal deaths. In Pakistan it is estimated that around 500 matemal deaths occur per hundred thousand live births. Although a gradual decline in mortality has been taking place in the country, health care coverage is still insufficient. Only 55 percent of the population has access to health services. A significant augmentation of services is necessary in order to reduce mortality, especially in rural areas. The life expectancy at bin2h has increased from 35-38 years at the time of independence to close to 60 years around 1990. The single largest increase in longevity occurred after the 1960s. In the past, males in Pakistan, on the whole, enjoyed a longer life expectancy (3-4 years longer than females) because of higher female mortality at younger ages and during the reproductive years (although this result might have been affected by differential underreporting of mortality by sex). Recently this difference has been reduced. 1.8 Literacy and Educational Attainment Pakistan has one of the lowest literacy rates (31 percent) in the world. Moreover, in 1985 there was a wide gap between male (43 percent) and female (18 percent) literacy rates. The lowest femaie literacy rate (4 percent), as of the 1981 Census, was observed for Balochistan. The literacy rate among rural females was only 2 percent in Balochistan and 4 percent in NWFP (Rukanuddin and Farooqui 1988). The primary school enmlment ratio is also very low (49 percent). The corresponding figures for males and females are 63 percent and 35 percent, respectively. Primary education in Pakistan is further characterized by drop-out and repeater rates which are considered to be among the highest in the world. Only 50 percent of the students who enter primary school complete the five years of primary school. Students, on the average, go to school for 1.7 years, which is very low compared to the average years of schooling in other developing countries (United Nations Development Programme 1991). 1.9 Economy Pakistan is intrinsically an agricultural country with more than 70 percent of its population living in rural areas. Agriculture is the largest single sector of the economy, employing more than 50 percent of the labour force. Agriculture accounts for 24 percent of the gross domestic product (GDP) and 70 percent of export earnings (Rukanuddin and Farooqui 1988). Development in agriculture and industry has transformed the economy of Pakistan and moved the country toward self sufficiency in meeting its basic needs. In 1990-91, the average per capita income in Pakistan was about Rs 9000 (US$400). The average rural monthly income per household in Pakistan is around one-third lower than the per household urban income. Moreover, it has been estimated that about 30 percent of the population in Pakistan live below the poverty line. Pakistan also has a low gross domestic savings rate of 13 percent of the GDP. The average annual growth rate of the GDP during the period 1985-90 was about 5.8 percent. 1.10 Population and Family Planning Policies and Programmes Pakistan was a pioneer among the most populated developing countries in supporting and implementing family planning activities start'rag in the 1950s. Concern has been expressed in successive Five-Year Development Plans (1955-60 to 1988-93) about rapid population growth and provisions have been made to support a family planning programme to deal with this burgeoning problem. Different approaches and strategies have been adopted during each plan period to promote the concept of a small family norm and to encourage the use of modem methods of family planning. These strategies have varied in design, coverage, outreach, supervision and guidance. However, due to a lack of consistent government commitment and social and cultural constraints, the programme has not been adequately effective in providing family planning services or generating widespread demand for the adoption of contraceptives. Financial and operational obstacles have also hindered the coverage of the programme, which is in the range of 25-30 5 percent of the total population. Family planning facilities are more concentrated in urban areas than in rural areas. The fertility inhibiting effect of the family planning programme has been low in Pakistan and contraceptive use has remained low despite the existence of the programme for the last three decades. The environment for family planning in Pakistan has been quite difficult. Factors which are generally associated with high fertility rates worldwide also pertain to Pakistan: high illiteracy and low educational attainment (particularly among females), poverty, high infant and child mortality, high maternal mortality, a preference for sons, poor access to health facilities, low socioeconomic status of women, ignorance, conservatism, fatalism and religiosity. These factors reinforced one another in maintaining high and stable fertility rates in the country. After many years of effort, the coverage of family planning services does not exceed one-third of the population. Various fertility surveys have found a wide gap between knowledge and the use of contraception in Pakistan. These surveys, however, have also indicated the existence of a potential demand for family planning expressed by Pakistani women (Population Welfare Division 1986). I.II Health Policies and Programmes The Ministry of Health provides health care services through government hospitals and other health outlets. The objective of the health policy is to reduce the incidence of morbidity and mortality by providing preventive and curative care to the whole population. Specific attention is given to reducing infant and child mortality, curtailing severe undemutrifion among children and mothers, and improving child survival and safe motherhood. In order to combat high childhood morbidity and mortality due to infectious and communicable diseases, an immunisatlon programme was initiated in 1978 to protect infants and young children against six common diseases and pregnant mothers against tetanus. This programme was greatly accelerated in 1982 with the collaboration of the World Health Organisation and UNICEF. The Expanded Programme on lmmu- nisation (EPI) is a major component of this scheme to provide universal immunisation. High maternal mortality is a priority area for health policy and coverage is provided to mothers through ante- and postnatal services performed at maternal and child health centres. These efforts are complemented by projects focusing on child survival and nutritional status through growth monitoring, adequate food supplementation and the promotion of breastfeeding. The government is committed to improving the quality of health services and the coverage of primary health care services, especially in the rural areas, through its Basic Health Units and Rural Health Centres. The provincial Health Departments of the respective provinces provide these services through their outlets. It was only in 1991 that the new health policy provided for family planning services to be offered through all health outlets as an integral part of health services. 6 REFERENCES Population Census Organisation [Pakistan]. 1985. Hand Book of Population Census Data. Islamabad: Statistics Division. Population Welfare Division [Pakistan]. 1986. Pakistan Contraceptive Prevalence Survey, 1984-85. Islamabad: Ministry of Planning and Development. Rukanuddin, Abdul Razzaque, and M. Naseem lqbal Farooqui. 1988. The State of Population in Pakistan, 1987. Islamabad: National Institute of Population Studies. United Nations Development Programme. 1991. Human Development Report, 1991. New York: Oxford University Press. 7 CHAPTER 2 SURVEY DESIGN AND IMPLEMENTATION Tauseef Ahmed, M. D. Mallick and Alfredo Aliaga This chapter outlines various aspects of the design and implementation of the Pakistan Demographic and Health Survey--namely, the objectives and organisation of the survey, the sample design, the ques- tionnaire design, training and fieldwork, data processing, and implementation. 2.1 Objectives of the Pakistan Demographic and Health Survey The primary objective of the Pakistan Demographic and Health Survey (PDHS) was to provide national- and provincial-level data on population and health in Pakistan. The primary emphasis was on the following topics: fertility, nuptiality, family size preferences, knowledge and use of family planning, the potential demand for contraception, the level of unwanted fertility, infant and child mortality, breastfeeding and food supplementation practices, maternal care, child nutrition and health, immunisations and child morbidity. This information is intended to assist policy makers, administrators and researchers in assessing and evaluating population and health programmes and strategies. The PDHS is further intended to serve as a source of demographic data for comparison with earlier surveys, particularly the 1975 Pakistan Fertility Survey (PFS) and the 1984-85 Pakistan Contraceptive Prevalence Survey (PCPS). 2.2 Organisation of the Survey In April, 1990, the National Institute of Population Studies (NIPS), on behalf of the Government of Pakistan, signed a contract with the United States Agency for International Development (USAID) and IRD/Macro International Inc. (IRD), Columbia, Maryland, to carry out the Pakistan Demographic and Health Survey in collaboration with the Federal Bureau of Statistics, Statistics Division, Government of Pakistan. Technical assistance was provided by IRD for all phases of the survey through the Demographic and Health Surveys programme. The survey was funded by the United States Agency for Intemationai Development (USAID) and the Govemment of Pakistan. 2.3 Sample Design The sample design adopted for the Pakistan Demographic and Health Survey is a stratified, clustered and systematic sample of households. The universe consists ofaU urban and rural areas of the four provinces of Pakistan as defined in the 1981 Population Census, excluding the Federally Administered Tribal Areas (FATA), military restricted areas, the districts of Kohistan, Chitral and Malakand, and protected areas of North West Frontier Province (NWFP). The population of excluded areas constitutes about 4 percent of the total population. For the urban sample, the sampling frame used was the master sample prepared by the Federal Bureau of Statistics. This frame was developed by dividing each city/t0wn into enumeration blocks of approximately 200-250 households with detailed and clearly recognizable boundary particulars and maps. The updating of the frame was done on the basis of the information obtained from the 1988 Census of Establishmants. For the rural sample, the sampling frame used was the village list published by the 1980 Housing Census. The primary sampling units in the urban domain were enumeration blocks; in the rural domain they were mouzas/dehs/villages. 9 Sample Size and Allocation The PDHS sample is a subsample of the Federal Bureau of Statistics master sample, which includes 7,420 primary sampling units (PSUs). Consideration in the selection of the PDHS sample was given to the population parameters and geographic levels for which estimates were required, the resources available, and the expected rate of nonresponse. A sample of 8,019 households (secondary sampling units) was selected for coverage from 408 sample areas (PSUs). The distribution of primary sampling units, secondary sampling units (SSUs), eligible women and eligible husbands and their actual coverage in the four provinces is given in Tables 2.1, 2.2 and 2.3. Stratification Plan Cities having a population of 500,000 and above (Faisalabad, Gujranwala, Hyderabad, Karachi, Lahore, Multan, Peshawar, and Rawalpindi) were included to form the domain for the major cities. Quetta, which had a population of less than 500,000 but is the capital of Balochistan, was also included as a major city. For the selection of the sample, each of these cities constituted a separate stratum which was further stratified into low, middle, and high income areas, based on information collected in each enumeration block at the time the urban sampling frame was updated. For the remaining urban cities/towns, divisions of NWFP, Sindh, Punjab and Balochistan were grouped together to form a stratum. For the rural domain, each district in each province was considered a stratum, except in Balochistan where each division constituted a stratum. A two-stage stratified sample design was adopted for the survey. The sample PSUs from each urban stratum were selected with probability proportional to the number of households. The sample PSUs from each rural stratum were selected with probability proportional to the population enumerated in the 1981 census. Table 2.1 Sample coverage for urban and rural mess combined Coverage of primary sampling units (PSUs), secondary sampling units (SSUs), efigible women and eligible husbands, Pakistan 1990-91 Number of SSUs Number of Number of Number of PSUs (households) eligible women eligible husbands Not Not Not Not Province Covered covered Total Covered covered Total Covered covered Total Covered covered Total Pu~ab 155 0 155 2598 192 2790 2207 124 2331 461 103 564 Sindh II0 0 110 2071 189 2260 1798 102 1900 364 175 539 NWFP 82 0 82 1609 147 1756 1665 24 1689 313 81 394 B~ochistan 60 1 61 915 298 1213 941 43 984 216 44 260 Total 407 1 408 7193 826 8019 6611 293 6904 1354 403 1757 10 Table 2.2 Sample coveral~e for urban areas Coverage of urban primary sampling units (PSUs), secondary sampling units (SSUs), eligible women and eligible husbands, Palfistan 1990-91 Province Number of SSUs Number of Number of Number of PSUs (households) eligible women eligible husbands Not Not Not Not Covered covered Total Covered covered Total Covered covered Total Covered covered Total Pu~ab T 72 0 72 1178 118 1296 995 71 1066 212 56 268 M 39 0 39 667 71 738 558 51 609 116 36 152 O 33 0 33 511 47 55g 437 20 457 96 20 116 Sindh T 70 0 70 1167 93 1260 1059 72 1131 206 120 326 M 49 0 49 835 65 900 755 57 812 147 85 232 O 21 0 21 332 28 360 304 15 319 59 35 94 NWFP T 42 0 42 699 57 756 749 12 761 144 51 195 M 14 0 14 247 23 270 261 5 266 56 25 81 S 28 0 28 452 34 486 488 7 495 88 26 114 B~ochistan T 41 0 41 558 180 738 581 28 609 134 32 166 M 18 0 18 234 90 324 246 12 258 61 11 72 O 23 0 23 324 90 414 335 16 351 73 21 94 Totsl T 225 0 225 3602 448 4050 3384 183 3567 696 259 955 M 120 0 120 1983 249 2232 1820 125 1945 380 157 537 O 105 0 105 1619 199 1818 1564 58 1622 316 102 418 T ~- ToUd M = Major city 0 = Other urban Table 2.3 Sample coverage for rural areas Coverage of rural primary sampling units (PSUs), secondary sampling units (SSUs), eligible women and eligible husbands, pakistan 1990-91 Ptovino~ Number of PSUs Number of SSUs Number of Number of (households) eligible women eligible husbands Not Not Not Not Covered cove~v.d Total Covered covered Total Covered covered Total Covered covered Total Punjab 83 0 83 1420 74 1494 1212 53 1265 249 47 296 Sindh 40 0 40 904 96 1000 739 30 769 158 55 213 NWFP 40 0 40 910 90 1000 916 12 928 169 30 199 B~ochistan 19 1 20 357 118 475 360 15 375 82 12 94 Total 182 1 183 3591 378 3969 3227 110 3337 658 144 802 11 Households within each sample PSU were considered secondary sampling units (SSUs). A fixed number of SSUs were selected systematically with equal probability using a random start and a sampling interval: 18 SSUs from each PSU in the urban domain in the four provinces and in the rural domain of Punjab Province and 25 SSUs from each PSU in the rural domain of the remaining three provinces of Sindh, NWFP and Baiochistan. Unlike previous surveys in Pakistan, the PDHS did not allow the substitution of households in the case of nonresponse. From the selected sample of SSUs, a systematic subsample of one in three households was chosen for inclusion in the husbands' sample. The husbands of eligible women in these households were eligible to be interviewed, provided that they slept in the household the night before the interview. The sample was designed to produce reliable estimates of population and health indicators separately for Karachi and for urban and rural areas ofPnnjab, Sindh, NWFP and Balochistan. This objective required an oversampling of all urban areas as well as the provinces of NWFP, Balochistan and Sindh. Because of the nature of the PDHS sample, a separate weighting factor was required for every PSU. The weighting procedure has two major components: the design component and the response differential component, with the design component being the major one. The weights were standardized so that the weighted number of completed cases at the national level is equal to the unweighted total. After data entry, weights were applied to the households and individuals in each PSU, to insure that the weighted sample would properly represent the actual geographic distribution of the population of Pakistan. Weights for husbands followed the same methodology as weights for women, except that the husbands' nonresponse rates were used in the calculations. The target was to interview 8,019 ever-married women age 15-49. The size of the target sample was based on an assumption of 1.1 eligible women per household and a nonresponse rate of 10 percent. A summary of the distribution of eligible women and eligible husbands by province and urban-rurai residence is presented in Tables 2.1,2.2 and 2.3. In general, the sample was adequate in size and sufficiently representative of the population to provide reliable estimates for the country as a whole, for urban areas, for rural areas, and for each province. However, for smaller groups, the sampling errors are generally higher. The calculated sampling errors for selected variables are shown in Appendix B. 2.4 Questionnaires Three types of questionnaires were used in the PDHS: the Household Questionnaire, the Woman's Questionnaire and the Husband's Questionnaire (see Appendix D). The contents of the questionnaires were based on the DHS Model B Questionnaire, which is designed for use in countries with low contraceptive prevalence. Additions and modifications to the model questionnaire were made after extensive consultatious with related ministries and interested organisations and with members of the PDHS Technical Advisory Committee. The questionnaires were translated from the original English version into the national language (Urdu) and three regional languages (Panjabi, Sindhi and Pushto). The Household Questionnaire listed all usual residents of a sampled household, plus all visitors who slept in the household the night before the interview. Some basic information was collected on the charac- teristics of each person listed, including their age, sex, marital status, education and relationship to the head of the household. The main purpose of this section oftbe Household Questionnaire was to identify women and men who were eligible for the Women's Questiotmaire and the Husband's Questionnaire. In addition, the Household Questionnaire collected information on the household itself, such as the source of water, type 12 oftullet facilities, materials used in the construction of the house, and ownership of various durable consumer goods. The Woman's Questionnaire was used to collect information from eligible wometr--that is, all ever-marriedwomenage 15-49whosleptinthehouseholdthenightbeforethehonseholdinterview. Eligible women were asked questions about the following topics: Background characteristics Reproductive history Knowledge and use of contraeeption Pregnancy and breasffeeding Vaccinations and the health of children Marriage Family size preferences Husband's background In addition, interviewing teams measured the height, weight and arm circumference of all respond- ents' children under age five. The PDHS was the first national survey that collected demographic, health and anthropometric data simultaneously. The questionnaire was designed to be completed in an average interview time of about 60 minutes. The actual mean time for the individual interview was 53 minutes. The interview time ranged from 47 minutes for women with no children born since January 1986 to 60 minutes for women who had three or more children during that period. Interviews were also conducted with a subsample of husbands of eligible women who were married at the time of the survey. The Husband's Questionnaire consists of a subset of the questions on the Woman's Questionnaire, with particular emphasis on family planning, marriage, and family size preferences. 2.5 Recruitment, Training and Fieldwork The selection of field teams was done at the regional level in order to insure that interviewers were accustomed to local dialects and cultural norms and were acquainted with localities in adjacent areas. The majority of field interviewers had received either a bachelor's or a master's degree. In September-October 1990, prior to the main survey, a pretest of the questionnaires and field procedures was carried out. A two-week training session for interviewers and supervisors was conducted at Punjab University, Lahore. The training session was followed by two weeks of fieldwork. A total of 309 pretest interviews were completed in urban and rural areas of all four provinces in Pakistan (Punjab, Sindh, North West Frontier Province, and Balochistan). Training for the main survey took place in November-December 1990. Training was held simul- taneonsly at the Regional Training institutes of the Ministry of Population Welfare in three cities--Karachi, Lahore and Peshawar. Staff members from the National Institute of Population Studies, the Federal Bureau of Statistics, the Regional Training Institutes and IRD/Macro Intematinnal conducted the training sessions. Participants in the training course included 16 statistical officers from the Federal Bureau of Statistics (FBS) and more than 80 female and male interviewers. The four-week training course consisted of instruction in general interviewing techniques and field procedures, a detailed review of the questionnaires, practice in weighing and measuring children, and practice interviews in the field. Trainees who performed satisfactorily in the training programme were selected as interviewers for the main survey. The female interviewers whose performance was rated as superior were selectexl as field editors. 13 The fieldwork for the PDHS was carried out by 15 interviewing teams. Each team consisted of one field supervisor from FBS, one field editor, three female interviewers, one male interviewer and one driver (see Appendix A for a complete list of survey staff). The fieldwork started in December 1990 and was completed by May 1991. Transportation for the field teams was provided by FBS, provincial Population Welfare Departments, and NIPS. Assignment of PSUs to the teams and various logistic decisions were made by the PDHS staff. Each team was allowed a fixed period of time to complete fieldwork in a PSU before moving to the next PSU. All the teams started their fieldwork close to or adjacent to their headquarters. The main duty of the field editors was to examine the completed questionnaires in the field and ensure that all necessary corrections were made. An additional duty was to examine the on-going interviews and verify the accuracy of information collected on the eligibility of respondents. Throughout the survey, PDHS staff maintained close contact with all 15 teams through direct communication and spot-checking. The objective was to provide support in the field and advice to enhance data quality and the efficiency of interviewers. This objective was accomplished by communicating data problems and possible solutions to the interviewing teams, reminding interviewers about proper probing techniques, and examining the fieldwork of the supervisors. Each team supervisor was provided by FBS with the original household listing and the household sample selected by computer for each designated PSU. In case of any error in the sample information, the supervisors contacted FBS headquarters to resolve the problem. 2.6 Data Entry and Processing All completed questionnaires for the PDHS were sent to the National Institute of Population Studies for data entry and processing. The data entry operation consisted of office editing, coding, data entry and machine editing. Although field editors examined the completed questionnaires in the field, these were re-edited at the PDHS headquarters by specially trained office editors. This re-examination covered: checking all skip sequences, checking circled response codes, and checking the information recorded in the filter questions. Special attention was paid to the consistency of responses to age questions and the accurate completion of the birth history. A second stage of office editing comprised the assignment of appropriate occupational codes and the addition of commonly mentioned "other" responses to the coding scheme. One supervisor and five data entry operators were responsible for the data entry and computer editing operations. The data were processed using five microcomputers and the DHS data entry and editing programmes written in ISSA (the Integrated System for Survey Analysis). The data entry started in the first week of January 1991, within one week of the receipt of the first set of completed questionnaires. The data entry was done directly from the precoded questionnaires. All data entry and editing operations were completed by July 1991. A series of computer-based checks were done to clean the data and remove inconsistencies. Age imputation was also completed at this stage. As in all DHS surveys, age variables such as current age, age at first marriage, and the ages of all living or dead children were imputed for those cases in which information was missing or incorrect entries were detected. The PDHS followed the DHS tabulation plan, in order to maintain comparability with other countries where DHS surveys have been conducted. Some additional tables were inchided to examine special topics included on the modified PDHS questionnaire. 2.7 Field Problems Every survey is subject to a variety of field problems, which cannot be fully anticipated. The major problems encountered in the PDHS are highlighted below, with a discussion of their possible effects. 14 Transportation: Each field team was assigned a vehicle to visit dispersed PSUs and to move quickly from one sample area to the next. Unexpected heavy rains during the months of March, April, and May brought landslides and flooding in Punjab, NWFP and Balochistan, causing substantial delays. Tube-boats were, therefore, hired in some areas. Several attempts were made to reach engulfed PSUs. At times, travel on foot for several miles was necessary to reach the designare, xl PSUs. Security of Teams: The law and order situation in Sindh was at its worst from January through April, Teams in Sindh were advised to take full precautions before going to any disturbed rural PSU. In addition, local security officers had to accompany interviewing teams to several PSUs in Balochistan. One PSU in Balochistan could not be reached by a Sindh team because of the insecure situation and a lack of police protection for the PDHS team. Supervision: In some instances, the work of certain supervisors was found to be weak: they were not moving to new PSUs as planned; they lacked coordination among team members; they did not dispatch the questionnaires from completed PSUs on time; they gave unauthorized leave to interviewers; they sent in an incomplete set of questionnaires; and at times they did not help female interviewers to locate sample households. Funds: Funds for the fieldwork were often not released by responsible agencies in a timely fashion. These delays caused frustration for interviewers as they had to rely on borrowed money rather than their own salaries. A loan of Rs. 2 million from the Ministry of Population Welfare provided timely relief and faci- litated the full execution of the fieldwork. Timing: The actual fieldwork was planned to be completed before the month of Ramadan which started in March 1991. Due to the unfortunate delays caused by heavy rains, almost all teams worked throughout the month of Ramadan without any break. Noncooperation: In a few areas in NWFP and Sindh, where the main cash crop is poppies or where dacoits reside, almost all households were apprehensive about talking to the interviewers, especially when questions were asked about household members and the ownership of durable consumer goods. PDHS team members were sometimes mistaken for members of the narcotics board or as television license examiners. Cultural Norms: In several PSUs in Balochistan and NWFP, respondents willingly completed the interview but refused to allow anthropometric measurements to be taken. Most women did not want any outsider to touch their children. Moreover, mothers did not want others to know the weight and height of their children to protect them from the evil eye. Sample Selection: The sample for the PDHS was selected at FBS headquarters. Some errors were detected in the sampling interval for households in the overall sample and in the husbands' sample. These problems caused some delays and confusion in the field. Most of these problems resulted in short-term difficulties but did not deter the overall progress of the project. 2.8 Coverage ofthe Survey Tables 2.4 and 2.5 show the results of the household and individual interviews for the women's sample and the husbands' sample. A total of 8,019 households were selected for the women's sample. About 90 percent of the selecl~l households were successfully contacted and interviewed. The shortfall was primarily due to dwellings that were vacant or households which were absent when they were visited by interviewers. Of the 7,404 households found to be occupied (including listed dwellings that could not be found), 97 percent were successfully interviewed. In other words, once a household was contacted, it was 15 Table 2.4 Results of the household and individual intelviewe for the women's sample Percent distribution of households and eligible women in the women's sample by results of the homehcld and individual interviews, and response rates, according to residence and province. Paldstan 1990-91 Residence Provino~ Result of intervlcw To ml Major Oth*r and response rate urban city urban Rural Punjab Sindh NWFP Balochistan Total SelectM households for women's sample Completed (C) 88.9 88.8 89.1 90.5 Household present but no 0.9 1.0 0.7 1.0 competent respondent at homo fliP) Refused (R) 1.0 1.4 0.4 0.4 Dv,~Uing not found (DNF) 1.0 0.6 1.6 0.9 Household absent (HA) 2.6 2.6 2.7 2.2 Dw~11ing v ac~nffadd~css 4.6 4.9 4.2 3.1 not a dwelling (DV) Dwelling destroyed (DD) 0.2 0.2 0.3 0.3 Other (O) 0.7 0.5 1.0 1.6 Total percent 100.0 100.0 100.0 100.0 Number 4050 2232 1818 3969 Household response rate for women's sample (HRR) ! 93.1 91.6 91.6 75.4 89.7 0.9 1.9 0.2 0.5 1.0 0.9 0.7 0.5 0.7 0.7 0.4 0.1 0.2 5.0 1.0 0.6 2.6 1.8 7.1 2.4 3.2 2.6 3.2 g.7 3.9 0.2 0.3 0.4 0.1 0.2 0.7 0.2 2.2 2.5 1.2 100.0 100.0 100.0 100.0 100.0 2790 2260 1756 1213 8019 96.8 96.7 97.0 97.5 97.7 97.2 99.1 9Z3 97.2 Eligible women Completed (EWC) 94.7 93.5 96.2 96.7 94.7 94.6 98.6 95.1 95.7 Not at home (EWNH) 2.3 2.9 1.5 1.9 2.1 4.0 0.7 0.8 2.1 Refused CEWR) 1.7 2.3 1.0 0.7 1.9 0.8 0.5 1.7 1.2 partly completed (EWPC) 0.5 0.6 0.3 0.3 0.9 . . . . 0.5 0.4 Other (EWe) 0.8 0.7 0.9 0.4 0.4 0.6 0.3 1.9 0.6 Total percent 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Number 3572 1947 1625 3338 2331 1900 1689 990 6910 Eligible women response rate (EWRR) 2 95.5 94.2 97.1 97.1 95.0 95.2 98.9 96.9 96.3 Overall response rate for women (ORRW# 92.5 91.0 94.2 94.7 92.9 92.5 98.0 89.5 93.5 -- Less than 0.05 percent fUsing the number of households falling into specific response categories, the household response rate (HRR) is calculated as: C C+ HP +R + DNF ZUsing the number of eligible women falling into specific response categories, the eligible women response rate (EWRR) is calculated as: EWC EWC + EWNH + EWR + EWPC ~The overall lesponse rate for women (ORRW) is calculated as: ORRW = HRR * EWRR 16 Table 2.5 Results of the household and individual interviews for the husbands' sample Percent distribution of households and eligible husbands in the husbands' sample by resultJ of the household and individual interviews, and response rates, aoc, ording to residence and province, Pakistan 1990-91 Residence Province Result of interview Total Major Other and response rate urban city urban Rural Punjab Sindh NWFP Balochistan Total Selected households for husbands' sample Completed (CH) Household present but no competent respondent at home (HPIO Refused (RH) Dwelling not found (DNFH) Household absent 0J~AH) DweUthg vacant/address not a dwelling (DVH) Dwelling destroyed (DDH) Other (OH) Total percent Number Household response rate for husbands' sample (HRRH) t ~ .0 89.8 90.3 90.7 93.0 91.4 93.8 77.0 90.3 0.7 0.7 0.8 1.2 1.0 2.0 -- 0.5 1.0 1.1 1.6 0.5 0,4 1.0 0.8 ~3 1.0 0.8 0.8 0.7 1.0 0.8 ~1 0.3 -- 4.7 0.8 2.3 2.3 2.3 1.9 0,4 2.1 1.7 6.4 2.1 4.1 4.3 4.0 3.3 3.3 3.0 2.0 8.4 3.7 . . . . . . 0.4 0.1 0.4 0.2 -- 0.2 0.9 0.7 1.2 13 1.1 - 2.0 2.0 1.1 1~.0 1~.0 1~.0 1~.0 1~.0 1~.0 1~.0 1~.0 1~.0 1351 7~ 6~ 1342 930 760 598 ~5 2693 97.1 96.8 97.5 97.4 97.9 96.8 99.6 92.6 97.2 Eligible husbands Completed (EHC) 72.9 70.8 75.6 82.0 81.7 67.5 79.4 83.1 77.1 Not at home (EHNH) 23.9 25.7 21.5 15.1 16.3 29.9 16.5 11.9 19.9 Postponed (EHP) 0.3 -- 0.7 . . . . . . . . 1.2 0.2 Refused 0SHR) 1.6 2.0 1.0 0.9 0.9 0.9 1.8 1.9 1.3 Partly completed 0SHPC) 0.3 0.4 0.2 0.5 0.5 0.4 0.3 0.4 0.4 Other 0SHO) L0 1.1 1.0 1.5 0.5 1.3 2.0 1.5 1.3 Total percent 100.0 100.0 100.0 100.0 100.0 I00.0 100.0 100.0 100.0 Number 955 537 418 802 564 539 394 260 1757 Eligible husbands response rate (EHRR) 2 73.7 71.6 76.3 83.3 82.2 68.4 81.1 84.4 78.0 Overall response rate for husbands (ORRI~ 71.6 69.3 74.4 81.1 80.5 66.2 80.8 78.2 75.8 -- Less than 0.05 percent 1Using the number of households falling into specific ~sponse categories, the household response ram for the husbands' sample (HRRH) is calculated as: CH CH+ HPH + RH + DNFH 2Using the number of eligible husbands falling into specific response categcflcs, the eligible husbands' response rate (EHRR) is calculated as: EHC EHC + EHNH + EHP + EHR + EHPC 3The ovendl response rate for husbands (ORRH) is calculated as: ORRH = HRRH * EHRR 17 almost certain to complete the household interview. The highest response rate for the household interview was recorded for NWFP (99 percent); the lowest was recorded for Baiochistan (92 percent). In more than 15 percent of the cases in Balochistan, either the dwellings were vacant or the households were absent due to the temporary migration of households because of severe cold weather in that region. In the interviewed households, 6,9 I0 women were identified as eligible for the individual interview. Interviews were successfully completed for 96 percent of the eligible women. The difference between the number of women targeted for interviewing and actual contacts was mainly due to the fact that the actual number of eligible women per household was lower than assumed in the sample design. The principal reason for nonresponse among eligible women was the failure to find them at home, despite repeated visits to the household. The refusal rate was low (only 1.2 percen0. A sample of 1,757 husbands of eligible women was identified as being eligible for the husbands' interview. However, only 77 percent of eligible husbands could be comacted and have interviews completed. The response rate was particularly low in Sindh where almost one-thinl of eligible husbands were not at home and in major cities where one-quarter of husbands were not at home. The major reason for the high level of nonresponse among husbands was their absence from the households and the fact that male interviewers could not contact them even after several visits. 18 CHAPTER 3 CHARACTERISTICS OF HOUSEHOLDS AND RESPONDENTS Tauseef Ahmed and Syed Mubashir All A profile of the demographic and socioeconomic characteristics of the population in the sample households is presented in this chapter to provide background information about the respondents interviewed in the PDHS. The characteristics of the population are compared with those reported from earlier surveys and censuses, wherever possible, to examine differentials and trends. 3.1 Household Composition Data on the household composition of the de jure population are presented for urban and rural areas in Table 3.1. The results show that households in Pakistan are predominantly headed by males) Only seven percent of all households are headed by females. This figure is slightly higher than the corresponding six percent obtained in the 1979-80 Population, Labour Force and Migration Survey. Pakistani households tend to be large with an average of six to seven persons living and eating together in a single household. The typical household in Pakistan has an average of 6.7 persons, but about one-quarter (26 percent) o f all households have four or fewer members. The breakdown by place of residence shows that there are more members in urban households (7.2 persons) than in rural households (6.5 persons). These results are consistent with the findings of the Pakistan Contraceptive Prevalence Survey (PCPS) and the Population, Labour Force and Migration Survey (PLM). The lower mean size of households in rural areas could be due partly to the migration of some household members to urban areas. Joint and extended family living arrangements are the norm in Pakistan. More than 60 percent of households have three or more related adults, while one-third have two related adults. Only one household in twenty had just one adult among the usual residents of the household. Another topic for which data are generally not available in Pakistan is the extent to which children live with families other than their own parents. In the PDHS, only one percent of households contained children under 15 years of age, who were not living with either of their natural parents. Many of these children are likely to be domestic servants, although some may be adopted or foster children. tRespondents to the household questionnaire were asked to identify the head of the household themselves. No standard definition was provided since the objective was to determine who was the head of the household according to the rcspondent's own definition of that concept. 19 Table 3.1 Household composition Percent distribution of households by sex of head of household, household size, and kinship of household members, and the percentage of households with children not living with either of their natural parents, according to orban-rmal residence, Pakistan 1990-91 Household Total Major Other char aeteristic urban city urban Rural Total Household head Male 92.1 93.0 90.9 93.2 92.9 Female 7.9 7.0 9.1 6.8 7.1 Total I00.0 I00.0 100.0 I00.0 I00.0 Number of usual members I 1.9 1.3 2.7 3.4 2.9 2 4.6 4.3 5.1 6.6 6.0 3 6.4 6.2 6,7 7.8 7.4 4 8.5 9.6 6.9 9.9 9.5 5 11.6 11.6 11.7 12.5 12.3 6 13.5 14.1 12.7 13.3 13.4 7 12.4 12.7 12.0 11.6 11.8 8 12.6 11.6 14.0 11.6 11.9 9 + 28.5 28.7 28.2 23.2 24.8 Total 100.0 100.0 100.0 100.0 100.0 Mean size 7.2 7.2 7.2 6.5 6.7 1979-80 PLM mean size 6.6 U U 6.1 6.3 1984-85 PCPS mean size 7.3 7.4 7.2 6.7 6.9 Kinship of household members One adult 3.0 2.2 4.2 5.4 4.7 2 related adults opposite sex 27.8 27.6 28.0 32.9 31.4 2 related adults same sex 1.4 1.3 1.6 1.7 1.6 3+ related addts 65.8 65.9 65.7 59.5 61,4 Other 1.9 3.0 0.6 0.5 0.9 Total 100.0 100.0 100.0 100.0 100.0 Percentage of households with children not living with either of their natural parents 1.0 0.6 1.6 1.1 1.1 Number of households 2120 1202 918 5073 7193 U = Unknown; no information 3.2 Age-sex Distr ibut ion of the Household Populat ion As in many developing countries, data collection efforts in Pakistan are subject to age misreporting and heaping on certain ages due to digit preference. Because of these limitations, special attention was paid in the PDHS to minimizing age reporting errors. Interviewers were given training in the techniques of probing to elicit age information that is as accurate as possible. Interviewers were also provided with reference calendars to help them in determining an approximate age for those respondents who were not able to report their exact age or date of birth. One calendar listed the dates of major national and local events for the last fifty years. Another calendar showed how to convert seasonal and Islamic months into dates in the 20 Gregorian calendar. Finally, interviewers were provided with an age conversion table to allow them to check the consistency of age and date of birth responses. In most rural areas, where knowledge about ages and dates of birth is not the norm, the PDHS interviewers were largely successful in estimating age information by using the calendars or calculating the ages of individuals relative to the age of any household member whose age could be determined. Nevertheless, errors in recording ages and dates of birth could not be totally eliminated. Table 3.2 shows the age distribution of the de facto male and female population enumerated in the PDHS and sex ratios by five-year age gmups. Because of continuing high levels of fertility, Pakistan's population is relatively young (see Figure 3.1). The median age of the de facto population is 17.6 years. Some age misreporting is evident from an examination of the age distributions. The sex ratios further help to highlight some of the errors in the data. The overall sex ratio for Pakistan is 1.08 males for each female. The sex ratios for the population as a whole am fairly stable up to age 40-44, but rather erratic thereafter. The relatively high sex ratio at age 45-49 (particularly in urban areas) suggests that in some cases interviewers may have "aged" women in that age group across the 50-year age boundary so that the women would be ineligible for the individual interview. Table 3.2 Household population by ase I residence and sex Percent d/strib~fion of the de facto household populatinn by five-year age group, according to urban-rural residence and sex, Pakistan 1990-91 Total urban Major city Other urban Rural Total Sex Sex Sex Sex Sex Age group Male Female ratio Male Female nttio Male Female ratio Male Female rado Male Female ratio 0-4 12.6 13.6 0.98 12.3 13.7 0.95 13.0 13.4 1.02 13.4 13,6 1.08 13.2 13.6 1.05 5-9 15.2 16,5 0.97 14.7 15.5 1.00 16.0 17.9 0.94 18.2 17,9 1.11 17,3 17.5 1.07 10-14 13.4 14.3 0.99 12.8 14.1 0.96 14.2 14.6 1.02 13.9 13.3 1.14 13.8 13.7 1.09 15-19 11.7 11.3 1.09 11,8 11.5 1.08 11,5 11.1 1.10 9.7 9.5 1.11 10.3 10.1 1.10 20-24 9.0 8.7 1.10 9.6 9.4 1.07 8.3 7.6 1.15 7.4 8.0 1.02 7.9 8.2 1.05 25-29 7.2 7.5 1.02 7.7 7.8 1.04 6,6 7.0 0.98 6.3 7.7 0.90 6.6 7.6 0.94 30-34 5.5 5.5 1.06 5.4 5.3 1.07 5.7 5.7 1.05 5.3 5.5 1.05 5.3 5.5 1.05 35-39 5.3 5.0 1.12 5.6 5.3 1.10 5.0 4.6 1.16 4.2 4.3 1.07 4.6 4.5 1.09 40-44 4.1 3.7 1.16 4.3 3.7 1.20 3,8 3.6 I.II 4.0 4.1 1.06 4.0 4.0 1.09 45-49 3.6 2.7 1.43 3.9 2,5 1.65 3.3 2.9 1.18 3.1 2.8 1.20 3.2 2.7 1.27 50-54 2.9 3.6 0.86 3.1 3,3 0.98 2.8 4.0 0.74 2.8 3.7 0.83 2.9 3.7 0.84 55-59 2.1 2.2 L04 2,1 2.2 1.01 2.2 2.1 1.09 2.1 3.0 0.78 2.1 2.7 0,85 60-64 2.6 2.2 1.24 2,7 2.3 1.21 2.6 2.1 1,28 3.2 2.5 1.38 3.0 2.4 1.34 65-69 1.6 I.I 1.51 1.5 1.2 1.38 1.6 1,0 1.70 1.9 1.7 1.21 1.8 1.5 1.28 70-74 1.5 1.0 1.47 1.3 I.I 1.28 1.6 1.0 1.75 2.5 1.2 2.36 2.2 I.I 2.09 75-79 0.5 0.2 2.37 0.5 0.2 3.~0 0,4 0.3 1.67 0.6 0,4 1.57 0.6 0.3 1.73 80 + I.I 0.9 1.20 0.9 0.9 I.I0 1.3 1.0 1.32 1,5 0.8 2.0~ 1.4 0.8 1.76 Mhalng] 0.I -- * . . . . * 0.I 0.I * 0.I 0.I * 0.I 0.I 1.38 Don't know Total 100.0 100.0 1.06 100.0 100.0 1.06 100.0 100.0 1.0~ 100.0 100.0 1.10 1(30.0 100.0 1.08 Nmnber 7480 7089 14569 4303 4070 8373 3177 3019 6196 16293 14876 31169 23773 21965 45737 - Len than 0.05 perccm * Baited on fewe~ than 25 tmweighted crees. Numhor not shown. 21 Age 80* 75-79 70-74 65-69 60-64 55-59 50-54 45-49 40-44 35-39 30-34 25-29 20-24 15-19 10-14 5-9 0 -4 Figure 3.1 Population Pyramid of Pakistan Mal~ 10 8 6 4 2 0 2 4 Percent 6 8 10 PDHS 19g0-91 Table 3.3 Population by age from selected sources Percent distribution of the population by age group, 1990-91 PDHS, 1984-85 PCPS and 1981 census 1990-91 1984-85 1981 Age PDHS PCPS census 0-4 13.4 16.4 15.3 5-9 17.4 16.3 16.0 10-14 13.7 12.8 13.2 15-19 10.2 10.1 9.5 20-24 8.1 8.0 7.6 25-29 7.1 6.9 6.7 30-34 5.4 5.3 5.6 35-39 4.6 4.9 5.1 40-44 4.0 4.1 4.7 45-49 3.0 3.2 3.7 50-54 3.2 3.1 3.6 55-59 2.4 2.3 2.0 60-64 2.7 2.6 2,7 65 -69 1.7 1.3 1.2 70-74 1.7 1.3 1.4 75 and over 1.6 1.4 1.7 Total 100.0 100.0 100.0 Source: Original data from Pakistan Contraceptive Prevalence Survey; Populatlon Census Organlsadon (1984) 22 The overall distribution of the population by age is quite similar to that of the 1981 census and previous demographic surveys (see Table 3.3). The percentage of the population below age five, however, is smaller in the PDHS than in the other sources. This shift in age composition can be attributed partly to a reduction in fertility levels in the recent past. However, some of this shift is due to the omission of young children and displacement in their ages, particularly in rural areas (see the discussion of omission and age displacement in Chapter 4). It is interesting to note that despite substantial differences among the three sources in the first three age categories, the percentage of the total population that is under age 15 years of age is exactly the same (44.5 percent) in the 1981 census and in the PDHS. 3.3 Mar i ta l Status The PDHS gathered information on the marital status of all household members age 15 and over. Examination of the marital status data for the de facto population (see Table 3.4) reveals a consistent picture, Table 3.4 Household population by age t sex and marital status Percent distribution of the de facto household population by marital status, acc.ordin 8 to age and sex, Pakistan 1990-91 MALE Never Divorced/ Age group married Married Widowed separated Missing Total Number 15-19 93.8 3.5 0.0 0.0 2.6 100.0 2448 20-24 75.3 23.1 0.4 0.0 1.2 100.0 1883 25-29 38.7 59.4 0.9 0.3 0.7 100.0 1561 30-34 15.1 82.2 1.6 0.5 0.6 100.0 1269 35-39 7.6 90.8 1.1 0.5 -- 100.0 1083 40-44 2.8 93.3 2.5 1.4 -- 100.0 951 45-49 2.0 93.4 3.7 1.0 -- 100.0 766 50-54 2.1 92.8 4.5 0.6 -- 100.0 678 55-59 1.8 93.3 4.6 -- 0.3 100.0 506 60-64 0.7 87.8 1 I. I 0.3 0. I 100.0 708 65+ 1.0 78.2 19.3 0.9 0.5 I00.0 1398 Total 35.3 59.5 3.8 0.4 0.9 100.0 13265 FEMALE Never Divorced/ Age group married Married Widowed separated Missing Total Number 15-19 78.1 18.4 0.3 0.2 3.0 I00.0 2219 20-24 39.4 59.7 0.1 0.6 0.2 I00.0 1798 25-29 11.8 86.0 0.9 1.0 0.4 100,0 1669 30-34 3.8 93.2 1.5 1.6 -- I00.0 1207 35-39 2.0 93.1 3.9 0.9 - 100.0 996 40-44 2.3 92.7 4.6 0.3 - 100.0 871 45-49 2.0 90.5 7.3 0.2 - 100,0 602 50-54 0.7 81.8 15.5 2.0 0.1 100.0 805 55-59 0.3 77.3 21.4 0.7 0.2 100.0 597 6{)-64 0.5 72.1 26.2 0.I I.I I00.0 528 65+ 0,3 48.1 48.9 0.3 2.4 100.0 839 Total 22.6 67.8 8.0 0.7 0.9 I00.0 12143 Note: Total includes 14 males and 10 females whose age is unknown. - Less than 0.0~ percent 23 relative to previous surveys, with regard to the universality of marriage for both males and females. Almost all females get married by their early thirties and males by their early forties. However, women tend to get married much earlier than men. For example, 60 percent of women age 20-24 am currently married but only 23 percent of men in that age group are married. Females exhibit an earlier transition than males to widowhood or to being divorced or separated. This finding is consistent with the cultural norm which permits the remarriage of widowed or divorced men but discourages the remarriage of women. The earlier transition of women to widowhood is also due to the age difference between husbands and wives and to higher male mortality in the older age groups. Nearly half of all women age 65 and over are widows, whereas only 19 percent of men in that age group are widowers. 3.4 Educational Attainment The level of educational attainment in a society is an important indicator of social development. Moreover, education is considered to be a major factor underlying social status. Educational attainment has also been shown to have a significant effect on fertility behaviour, contraceptive use, infant and child mortality, morbidity and issues related to family health and hygiene. Table 3.5 presents the distribution of the household population age five and over by level of education, according to sex, age, place of residence and province. Education has been one of the few neglected sectors which has not caught up with the growing population in Pakistan. Table 3.5 shows that 43 percent of males and 68 percent of females have never attended school. Overall, less than one-third of males (30 percent) and one-fifth of females (20 percent) have attended only primary school, while 16 percent of males and 7 percent of females were reported to have reached secondary school or higher. On average, males have completed 3.2 years of schooling, whereas females have completed only 1.6 years. The data indicate that males receive much more education than females. Such differentials are more prominent at higher levels of education. Men are 50 percent more likely than women to have attended only primary school, twice as likely to have attended middle school without going on to secondary school, and 2.3 times as likely to have attended secondary school or higher education. The differentials in level of education according to age group provide an indication of the develop- ment in the educational attainment of the population over time. A steadily increasing pememage of both males and females have attended school and the increases are particularly dramatic at the secondary level of education and above. The sex differential in educational attainment continues to be significant for all ages, with little indication that the gap is decreasing, except in the youngest age group. The findings further indicate that one-half of males (51 percent) and four-fifths of females (79 percent) in rural areas have received no education. Among both males and females, the relative urban-roral differentials are most pronounced at the highest educational level. The median and mean number of years of schooling are slightly higher in major cities than in other urban areas. The urban-rural difference is undoubtedly due to a lack of facilities or their inaccessibility in rural areas, while male-female differentials could be attributed to cultural norms and the social constraints faced by women. Punjab and Sindh have relatively low percentages of females who have received no education, while in NWFP and Baiochistan more than 80 percent of females have never been to school. The level of educational attainment is much higher for males than for females in all provinces, but the disparity is particularly striking in NWFP. The PDHS also asked about the current status of school attendance for all persons under age 25. Table 3.6 presents the percentage of the de facto male and female population (age 6-24) who are enrolled in school by age, sex and place of residence. Overall, more than half (54 percent) of all school age children (age 6-15) were currently enrolled in school; the figures were 64 percent formales and 44 percent for females. The enrulment rate is much higher in major cities (75 percent) than in rural areas (46 percent) (see Figure 3.2). 24 Table 3.5 Educational level of bouscheld population pcxcer~ distribution of the de facto male and female household population age five and over by highest level of cducauJon ammded, accordin 8 to ~lec~l background charactc~cs, pakistan 1990-91 MALE Nmnber Backsrcund No of Median Mean characteristic education Primary Middle Secondary+ Missing Toad peaom years years Ap group 5-9 44.4 54.8 0,4 -- 0.4 100.0 4102 0.7 0.6 10-14 23.8 51.4 20.0 4.5 0.3 180,0 3274 3.6 3.2 15-19 28.4 17.8 20,8 32`6 0.3 100.0 2448 6.3 5.3 20-24 33.9 19,1 13.2 33.5 0.3 100.0 1883 5.7 5.4 25-29 39.7 20.7 12.4 27.1 0.1 100.0 1561 5.2 4.8 30.34 45.5 16.7 11.3 26.0 0.4 1~).0 1269 4.4 4.4 35-39 44.7 17,1 10,2 27.7 0.3 100.0 1083 4.5 4.5 40-44 50.6 20,2 8,1 20.9 0.2 100,0 951 0.0 3.8 4549 54.3 14.3 8.7 22.3 0.4 100.0 766 0.0 3.7 50.54 61.6 15.9 6.1 16.0 0.4 180.0 678 0.0 2.9 55-59 64.6 17.2 6.1 11.4 0.8 180.0 505 0.0 2.4 60-64 73.8 11.3 5.6 9.3 0.1 100.0 708 0,0 1.8 65+ 80.1 9.3 5.0 5.2 0.4 100.0 1398 0.0 1.3 Resldeace Toad mban 26.9 29.8 13.7 29.4 0.3 100.0 6535 5.0 5.0 Major city 27.0 27.0 13.4 32.3 0.2 180.0 3772 5.2 5.2 Other urbtn 26.7 33.5 14.0 2.5.6 0.3 100.0 2763 4.7 4.6 Rural 50.7 29.8 9.3 9.8 0.4 100.0 14106 0.0 2.4 Province Punjab 40.8 30.2 12.0 16.9 0,1 100,0 12330 1.8 3.4 Sindh 44.0 31.5 7.8 16.0 0.7 100.0 4962 1.0 3.2 NWFP 46.7 27.6 II.0 14.4 0.2 100.0 2.597 1.0 3.0 Balochistan 63.4 20.8 6.6 7.1 2,0 100.0 752 0,0 1.7 Total 43.1 29.8 10,7 16,0 0,3 100.0 20641 1.3 3.2 FEMALE Number Background No of Median Mean characteristic education Prhnary Middle Secondary+ Missing Total persons years years Age Ilroup 5-9 58.7 IO-14 48.5 15-19 54.9 20-24 63,9 25-29 72.0 30-34 75.3 35-39 79.0 40-44 83.4 45-49 86,3 50.54 93.0 55-59 92.8 60-64 94.5 65+ 95,3 Residence Tots] uxbtn 42`8 Major city 37.6 Omer ta~m 49.9 Rural 79.4 province Punjab 63.7 ShuSh 66.2 NV/PP 81.6 Balochadan 88.5 Total 67,6 ~.7 0.1 -- 0.4 I~.0 3340 0.0 0.4 35.5 12.4 3.2 0.5 1~.0 2998 1.0 2.1 15.5 10,5 18,9 0,2 I~.0 ~19 0.0 3.2 13.6 6,4 16,1 0,1 1~.0 1798 0.0 2.8 10.0 5.0 12.9 -- 1~.0 1669 0.0 2.2 10.4 4.0 9.5 0,7 1~,0 1207 0.0 1.8 9.1 4.8 6.9 0.2 1~.0 996 0,0 1.5 7.0 3.3 6.0 0.3 1~.0 ~1 0.0 1.2 6.5 2.8 3,4 1,0 I~.0 602 0.0 0.9 3.0 1.9 1.7 0,4 1~,0 805 0.0 0,5 3,9 1.5 1.0 0,7 I~.0 597 0,0 0.4 1,8 1,3 1.6 0,9 I~,0 528 0.0 0.3 2.2 ~3 0.9 1,4 1~,0 ~9 0.0 0.2 28.4 10.5 18,0 0.3 100.0 6126 1.5 3.4 28.9 11,6 21.6 0.3 100.0 3511 2.8 4.0 27.6 9.0 13.2 0.2 100.0 2615 0.9 2.7 15.8 2.6 1.6 0.5 100.0 12855 0.0 0.7 22.2 6A 7.7 0,2 100.0 11389 0.0 1.8 20.2 4.6 0.2 0.8 100.0 4345 0.0 1.7 12.2 3.0 2.9 0.3 100.0 2570 0.0 0.8 7.1 1.4 1.2 1.8 100.0 675 0.0 0.4 19.9 5.2 6.9 0.4 100.0 18981 0.0 1.6 Note: Excludm 14 males and 10 females whine age is enknovm. -- Le~ than 0.05 pezc~t 25 Table 3.6 School enrolment Percentage of the de facto household population 6-24 ye~s of age cttrrendy enrolled in school by age, sex and urban-rural residence, Pakistan 1990-91 Male Female Total Total Major Other Total Major Other Total Major Other Age group urban city urban Rural Total urbaa city urban Rural Total urban city urban Rural Total 6-15 76.3 75.8 76.9 58.6 63.8 68.1 73.3 61.9 31.9 43.6 72.2 74.6 69.2 46.1 54.1 6-10 78.1 77.9 78.4 56.8 62.6 71.7 75.5 67.1 33.2 44.8 74.9 76.7 72.7 45.7 54.1 11-15 74.1 73.4 75.0 61.4 65.5 63.8 70.6 55.8 29.8 41.8 68.9 72.0 65.1 46.8 54.2 16-20 40.8 41.5 39.8 32.4 35.4 31.2 35.1 24.8 5.4 14.3 36.1 38.3 32.8 19.0 25.0 21-24 15.4 15.9 14.6 8.6 11.1 7.2 8.1 5.7 1.9 3.9 11.6 12.2 10.5 5.4 7.7 100 80 60 40 20 0 Percent Figure 3.2 School Enrolment by Age and Place of Residence 6-15 16-20 21-24 Age (Years) [ m Major City ~ Other Urban ~-~ Rural i [ I PDHS 1990-91 Since the concentration of the population is relatively high in urban areas, there is a greater incentive for both public and private agencies to invest in setting up schools. The increased school enrolment in urban areas is likely to be a function of the greater availability and easy accessibility of educational institutions in general as well as higher educational aspirations for children in urban areas. 26 At age 16-20, one in four children was still attending school. By age 21-24, fewer than one in ten was still in school. Table 3.6 shows the greater enrolment of male than female children at all ages. The sex differentials in enrolment am more pronounced in rural areas, and to some extent in small cities and towns, but nearly disappear in major cities. The sex differentials in enrolment increase with age, which at least partially reflects the greater dropout rate for female children. 3.5 Housing Characteristics Selected housing characteristics are shown in Table 3.7 for households with at least one eligible woman who completed the interview. Electricity is nearly universal in urban areas (96 pereen0 while less than half of rural households (47 percent) were reported to have electricity. Overall, 61 percent of all households have electricity. The questionnaire included information on the source of drinking water for the household. A safe, accessible source of drinking water is important for the health and welfare of household members. Table 3.7 shows that only 18 percent of all households have water piped inside their houses and 9 percent have water piped onto their property. In total, more than 90 percent of all households in major cities have access to running piped water. Piped water is available to only 63 percent of households in other urban areas and 15 percent of households in rural areas. In rural areas, nearly 70 percent of households obtain their drinking water from wells and tubewells, while 13 percent rely on rivers, canals and kamzes for their drinking water. The use of different types of sanitation facilities varies greatly by place of residence. About 90 percent ofaU households in major cities have flush toilets in their houses compared to 55 percent in other urban areas. 2 In contrast, the toilet facilities in mral areas are quite rudimentary. Seventy-two percent of rural households have no toilet facility at all and an additional 22 percent have only a pit latrine or a bucket. One of the more important socioeconomic indicators for survey households is the quality of their housing, as measured by the construction material of the walls and roofs. More than 92 percent of households in major cities have baked brick and cement walls, compared to only two-thirds of households in other urban areas. In contrast, 60 percent of rural households live in houses with mud or unbaked brick walls and about 16 percent of their houses have wooden or bamboo walls. Similarly, a large majority of bouseholds in major cities (73 percent) live in houses with roofs made of concrete or T-irons or wood with bricks. Fifty-eight percent of households in other urban areas live in houses which have the same type of roofing materials. The other extreme is reported in rural households, where the roofs of houses are made primarily of wood or bamboo (72 percent). At the national level, a little more than one-third of households live in houses with cement and baked brick walls and roofs made with concrete or bricks with T-irons. Finally, the number of persons per room used for sleeping was calculated as a measure of crowding. At the national level, only 20 percent of households have 1-2 persons per room and 37 percent of households have 3-4 persons sleeping in one morn. At the other extreme, 20 percent of households have seven or more persons sleeping in one room. On the average, five persons sleep in one room. The degree of crowding was nearly as great in urban areas as in rural areas. 2 Flush toilets are defined as toilets in which water is carried down waste pipes, whether the water is piped into the to'flet or poured in by buckets. 27 Table 3,7 Housleli characterlsfic= Percem ~fisttilanlon ~ houscholds with eliaible women by housing characteristics, according to ud~m- rand residence, Paldstan 1990-91 Housing Total Major Other characteristic urbmx city urban Rural Total FJec~ldty Yes 95.7 98.1 92.5 46.6 61.4 No 4.3 1.9 7.5 53.4 38.6 Total I00.0 I00,0 I00.0 I00,0 I00.0 Source of drinking water Piped into resklence 48.1 58.5 34.1 5.5 18.3 Piped omo property 20.3 20,1 20.6 4.4 9.2 Public tap 11.2 13.4 8.1 5.4 7.1 Well with pmnp. tubewell 15,5 4.7 30.0 55.1 43.2 Well without lured pump 2.2 0.7 4.2 13.5 10.1 River. canal, knrez 0.9 0.8 1.1 12.5 9.0 Tanker, vendor 0,3 0,3 0,2 0.5 0.4 Rainwater . . . . . . 0.9 0.6 Other 1.3 1.2 1.3 1,7 1.6 Missing 0.3 0.2 0.4 0.4 0,4 Total 100,0 I00.0 100.0 100.0 100,0 Sanitation facility Flush 74.5 89.3 54.6 5.8 26.5 Bucket 14.4 6.2 25,4 10.0 11.3 Plt lmrine 4.4 2.8 6.5 11.5 9.3 Other 0,7 0.5 0.9 0.1 0.3 No facilities 6,0 1,1 12.6 71.9 52,1 MAssing 0.I 0.I 0.I 0.7 0.5 Total 100,0 100.0 100,0 100,0 100.0 Material of walls Baked bdchs, cement 81,1 92,4 65.9 19.2 37.8 Unbaked bclcks, mud 16.3 6.9 28.9 60.0 46.9 Wood/baraboo 2.3 0.6 4.7 15.7 11.6 Other 0.3 0,1 0.5 5.1 3.6 Total 100.0 100.0 I00.0 I00.0 I00.0 Material of reef RCC/RBC 45.0 57.8 27.9 4,1 16,4 T-Iron/wood/brick 21.5 15.2 29.9 19,5 20,1 Asbestos/into shceu 10.6 16.7 2.4 3.6 5.7 Wood/bamboo 22.8 10.2 39.6 72.2 57.3 Other 0,1 0.1 0.1 0.6 0.4 Total 100.0 100.0 100.0 100,0 100.0 Persons per ~deeplng roem 1-2 21.6 22.4 20.5 19.3 20.0 3-4 39.6 39.2 40.1 36.1 37.2 5-6 20.6 21,3 19.6 23,9 22,9 7 + 18.1 17.1 19.6 20.4 19,7 Missing/Don't know 0,1 0,1 0.2 0,2 0.2 Total 100.0 100,0 100.0 100,0 100,0 Mean 4.6 4.5 4.6 4.8 4.7 Number ef homelmlds 1633 936 698 3796 5429 o. Less than 0,05 l~t 28 3.6 Presence of Household Durable Goods In order to obtain additional information on the socioeconomic status of households, household respondents were asked if specific household goods were present in their homes. Table 3.8 shows that 35 percent of all households in Pakistan have a radio and a little more than one-quarter (27 percent) have a television. As expected, both these items show large differenlials between urban and rural areas, but the differential for televisions is particularly prominent. Table 3.8 Household durable goods Percentage of households possessing vm'ious durable consumer goods, by urban-rural residence, Pakistan 1990-91 Total Major Other Item urban city urban Rural Total Radio 51.9 58.4 43.2 28.3 35.4 Television 64.2 74.7 50.1 10.9 27.0 Refrigerator 37.6 46.3 26.0 4.0 14.1 Room cooler 13.2 13.0 13.5 1.2 4.8 Washing machine 43.4 50.2 34.2 2.9 15.1 Water pump 23.0 21.5 25.0 3.2 9.2 Bicycle 39.9 34.4 47.2 31.4 33.9 Motorcycle 17.6 21.4 12.5 3.3 7.6 Car, van or ta'aeter 6.4 7.1 5.5 3.8 4.6 Number of households 1633 936 698 3796 5429 Refrigerators, room coolers, and washing machines are all concentrated in urban areas. Only a few households in rural areas reported owning any of these items. Bicycles are the most commonly owned means of transport in all areas. At the national level, 34 percent of all households own at least one bicycle. About 8 percent of all households have a motorcycle, but motorcycles are more common in major cities (21 percent). Less than five percent of households own a car, a van or a tractor. 3.7 Background Characteristics of Female Respondents Women were eligible for the individual interview if they were ever married, age 15-49, and stayed in the household the night before the household interview was conducted. Eligible women were asked their age, marital status, educational level, place of residence, work status and physical mobility, in addition to many other questions on demographic and health status. Table 3.9 presents information on the background characteristics of all 6,611 eligible women who were interviewed. More than half of these women (57 percent) were in the 20-34 age group, with the largest number in age 25-29. The age distribution of currently married women in their childbearing years from four sources is compared in Figure 3.3. The PDHS age distribution is closest to the age distribution for the PCPS. The relatively low proportion of currently married women age 15-24 in the PDHS is consistent with the evidence that the average age of marriage in Pakistan has been rising over time. 29 A large majority of ever-married women (96 percent) were currently married and only a negligible proportion were either widowed, divorced or separated. About 80 percent of women had never attended school and only 7 percent were educated up to the secondary or higher levels. Alraost 70 per- cent of women were residents of rural areas, 17 percent resided in major cities and the rest were located in other urban areas. A majority of respondents (60 percent) were from Pun- jab, 23 percent were from Sindh, 13 percent from NWFP and 4 percent from B alochistan. The norm in Pakistan is for women to stay home and take care of the house and the children. It is not common for women to join the labor market and their mobility is often restricted. In the PDHS, only 17 per- cent of women were reported to be currently working at the time of the survey and an additional 4 percent had worked only before marriage. Three-quarters of all women stated that they had never worked. To gauge the extent of their physical mobility, women were asked whether they would need to be accompanied by someone if they needed to go to a hospital or clinic for medical treat- ment. While one-quarter of women reported that they could go to a hospital alone, 71 percent reported that they would need to be accompanied by someone. The restriction on their physical mobility can be explained partly by cultural norms. An alternative hypothesis is that women usually do not seek medical treatment for minor illnesses and they leave the house only when their illness has become serious. Under these circum- stances, they would need to be accompanied to go to a hospital. Table 3.9 Background characteristics of female respondents Percent distrlbutinn of ever-runrrled women by background characteristics, Paldstan 1990-91 Weighted Unweighted Background Weighted number of number of characterize percent women women Age 15-19 6.5 428 407 20-24 16.0 1059 1064 25 -29 22.6 1494 1469 30-34 18.0 1187 1200 35-39 14.8 951 1031 40-44 12.8 844 820 45-49 9,3 617 620 Marital status Married 96.3 6364 6393 Widowed 2.4 159 148 Divorced 0.3 22 19 Separated 1.0 65 51 Residence Total urban 30.5 2019 3384 Major city 17.4 1151 1820 Other urban 13,1 868 1564 Rural 69.5 4592 3227 Province Punjab 59.7 3948 2207 Sindh 23.1 1529 1798 NWFP 13.3 878 1665 Balochistan 3.9 255 941 Education level attended No education 79.2 5237 5055 Prima~ 9,1 601 600 Middle 4.4 288 320 Secondary 6.2 410 522 Higher 1. l 75 114 Work status Currently working 16.8 1111 1057 Worked only before marriage 4.4 290 292 Worked only after marriage 0.7 44 52 Worked before and after marriage 1.1 72 74 Never worked 1 76.7 5073 5111 Missing 0.3 21 25 Mobtllty Could go to hospital alone 25.1 1660 1699 Would need to be accompanied 70.8 4682 4441 Depends or missing 4.1 269 471 Total 100.0 6611 6611 b'Never worked" means that the woman il not currently working and whe did not work either before marriage or just after marriage. 30 25 Figure 3.3 Distribution of Currently Married Women by Age, Pakistan, 1975-1991 Percent 20 15 10 O i i - ] i i 15-19 20-24 25-29 30-34 35-39 40-44 Age 45-49 i PDHS 1990-91 -- PCPS 1984-85 | I ~ Pop. Ceneus 1981 + PFS 1975 Source. Population Welfare Division (1986) In Table 3.10, variations in the level of education by age group, place and province of residence, and work status are examined. In all age groups, no less than three-quarters of women reported that they had never atCnded school. In general, younger women were more likely to have attended school than older women. A comparison of educational attainment as measured by the 1975 PFS and the PDHS confirms that levels of educational achievement for women have been increasing over time. Even among women residing in major cities, 48 percent had no education and only one-quarter had attended a secondary school or a higher level of education. At the other extreme, 90 percent of women from rural areas had no education and only 1 percent had attended secondary school. The provincial educational pattern foUows the general pattern of development. Punjab and Sindh, which are more developed, have a lower percentage of women who had no education and 8-9 percent had attended secondary school or gone beyond secondary school. In Balochistan, 96 percent of women had no education and only 1 percent had reached secondary school. Women who were currently working were the least educated group. Eighty-four percent of working women had never been to school and only seven percent had some secondary or higher education. The highest average level of education was exhibited by those who worked only before marriage or both before and after marriage. Those who worked just after marriage also had a relatively high level of education: 13 percent reported that they had attended secondary school or a higher level of education. 31 Table 3.10 Level of education Percent distribution of ever-mm'ried wornea by the highest level of education aaonded, according to selected background characteristics, Pakisum 1990-91 Background No charactczisfic education Primal 7 Middle Secondary Higher Total Number Age 15-19 81.0 10.9 3.6 4.2 0.2 100,0 428 20-24 75.3 12.6 5,l 5,9 1.0 100,0 1059 25-29 75.7 8,7 5.6 8.8 1.1 100.0 1494 30-34 77.7 9.2 4.0 7.5 1.6 100.0 1187 35-39 79.8 8.8 4.2 5.8 1.4 I00.0 981 40-44 84.0 7.2 3A 4.3 1,l 100.0 g44 45-49 88.5 5.4 2.7 2.5 0.8 100.0 617 Residence Total u~ban 55.0 14.g 9.1 17.5 3.6 100.0 2019 Major city 47.7 15.7 10.6 20.7 5.3 100.0 1151 Other urban 64.6 13.6 7.1 13.3 1.5 100.0 868 Rural 89.9 6.6 2.3 1.2 -- 100.0 4592 Province Punjab 76.5 10.3 5.2 6.7 1.3 100.0 3948 Sindh 76.8 10.2 3.8 7.9 1.3 100.0 1529 NWFP 90.6 4.0 2.4 2.7 0.3 100.O 878 Balochistan 96.3 1.9 0.6 1.0 0.2 100.0 255 Work status x Currently working 84.3 5.8 3.0 5.1 1.8 100.0 1111 Worked only before marriage 64.2 11.6 6.3 13.5 4.4 100.0 290 Worked only after marriage 74.0 9.1 3.8 12.4 0.6 100.0 44 Worked before & after marriage 63.9 15.1 4.8 7.1 9.1 100.0 72 Never worked 2 79.2 9.6 4.6 6.0 0.7 I(D.0 5073 Total 79.2 9.1 4.4 6.2 1.1 I00.0 6611 -- Less than 0.05 percant 1Excludes 21 women with missing information on work status. 2"Never worked" means that the woman is not currently working and she did not work either before marriage or just after marriage. 3.8 Exposure to Mass Media As an indicator of exposure to mass media, each woman interviewed was asked whether she usually reads a newspaper, watches television, or listens to radio at least once a week. Table 3.11 shows that 14 percent of women read a newspaper weekly, 30 percent watch TV and 27 percent listen to radio. As expected, there is a close association between the level of education and exposure to the three types of media. The low level of exposure to radio and TV among uneducated women may be explained by their lack of access to these facilities. There is also a large differential in media exposure between urban and 32 rural women, Women in major cities are more likely to watch TV (78 percent) than to read newspapers (43 percent) or listen to the radio (47 percent). In contrast, rural women listen to radio (21 percent) more than they watch "IV (13 percent) or read newspapers (5 percent). Women from small cities or towns follow a pattern similar to women from major cities, however exposure to all forms of media is the highest in major cities. Women in Sindh are more exposed to all types of media than women in other provinces. Women in Punjab have a pattern of media exposure similar to the national pattern. More than four times as many women in NWFP are exposed to radio and TV as are exposed to newspapers, while Balochi women are most often exposed to radio. In general, Balochi women have the least access to these media, perhaps because of a lack of media facilities in Balochistan and because of the large geographical area over which the population is dispersed. Table 3.11 Exposure to mass media Percentage of ever-merried women who usually read a newspaper, wash television, or listen to radio at least once a week by selected background characteristics, Pakistan 1990-91 Read Watch Listen Number Background newspaper TV to radio of characteristic weekly weekly weekly women Age 15-19 11.0 24.8 31.8 428 20-24 14,9 28.8 30,3 1059 25-29 17.3 30.2 29.3 1494 30-34 14.7 30.5 26.5 1187 35-39 14.8 31.7 26.6 981 40-44 10.1 30.6 22.0 844 45-49 8.7 28.9 21.9 617 Residence Total urban 34.1 67.5 41.1 2019 Major city 42.6 77.9 46.5 1151 Other urban 22.7 53.6 34.0 868 Rural 5.1 13.3 20.9 4592 Province Punjab 14.4 29.1 26.3 3948 Sindh 19.6 40.2 33.1 1529 NWFP 5.4 21.5 23.8 878 B alochistan 2.8 7.3 15.0 255 Education level attended No education 1.8 18.8 21.5 5237 Primary 43.1 57.8 43.1 601 Middle 65.9 70.4 50.4 288 Secondary 75,8 88.0 53.1 410 Higher 93.8 99.5 59.4 75 Total 13.9 29.8 27.1 6611 33 REFERENCES Population Census Organisation [Pakistan]. 1984. 1981 Census Report of Pakistan. Islamabad: Population Census Organisation (Census Report No. 69). Population Welfare Division [Pakistan]. 1986. Pakistan Contraceptive Prevalence Survey, 1984-85. lslamabad: Ministry of Planning and Development. 34 CHAPTER 4 FERTILITY Fred Arnold and Mehboob Sultan One of the major objectives of the PDHS is to estimate fertility levels, trends and differentials. Information on fertility will help to determine the impact of changes in the use of family planning and other changes in the proximate determinants of fertility. The fertility estimates presented in this chapter are based on the reported birth histories of ever-married women 15-49 years old who were interviewed in the PDHS. Respondents were first asked to report the aggregate number of sons and daughters they had ever given birth to in their lifetime. To encourage complete reporting, women were asked separately about children still living at home, those living elsewhere and children who had died. The birth history also obtained information on the sex, date of birth and survival status of each child. This information was used to calculate measures of current fertility and fertility trends over time, as well as cumulative measures of the number of children ever born. In addition, estimates of birth intervals and the mother's age at the initiation of childbearing were calculated from data on the timing of births. To obtain complete and accurate information on reproduction, interviewers were trained to probe carefully to facilitate the respondent's recall and to check any documents that may include birth dates for children. Moreover, for any intervals of more than three years between births, interviewers were required to record the reason for the long interval to help identify any live births that may have been missed during that time period. In spite of the precautions taken, the PDHS is subject to the same types of errors that are typical of all retrospective demographic surveys. These include the underreporting of births (particularly for children who died immediately after birth or at a very early age) and the mistiming of births. These types of problems are particularly prevalent in countries such as Pakistan where the level of female literacy is low. In previous demographic surveys in Pakistan, births have been misplaced away from the survey date because of a pattern of exaggerating children's ages that increases with age (Rethefford, et al. 1987). A further complication in the PDHS is the displacement of births out of the most recent five-year period (from 1986 to the time of the survey). This has been a significant problem in many DHS surveys in other countries (Arnold 1990). The apparent reason for this type of displacement is that interviewers were trying to avoid a lengthy set of questions on health that were asked only about children bom since January 1986. Because of the needs of policy makers for more detailed health data, the health section has been lengthened in recent DHS surveys and the displacement problem persists. Moreover, since height, weight and ann circumference measurements were also taken only on living children born in 1986 or later, there is an even stronger incentive for moving the dates of children's births out of that time period. In the PDHS, displacement was a serious problem, with nearly twice as many births being reported in 1985 as in 1986 (see Appendix C, Table C.4). For this reason, fertility and mortality rates in this report are presented for six-year periods, so that the transference of most displaced births will occur within a single time period rather than across time period boundaries. The omission of recent births has been a feature of all retrospective demographic surveys in Pakistan. This problem is often attributed to inaccurate reporting by respondents. In the PDHS, the omission of recent births may be compounded by the underenumeration of births by interviewers who are trying to circumvent the health questions and to avoid weighing and measuring young children. The decline in the average annual number of births from 1556 in 1982-85 to 1145 in 1986-90 (Appendix C, Table C.4) is undoubtedly due in part to the omission of children born in the five years before the survey. 35 It is difficult, however, to correct the fertility estimates for the incomplete reporting of births since some of the estimated fertility decline is a real phenomenon and an unknown portion is attributable to data er rors . 4.1 Fertility Levels and Trends Until recently, fertility rates in Pakistan have remained high with little evidence of a sustained fertility decline (Shah and Cleland 1988; Rukanuddin and Farooqui 1988; Shah, Pullum and Irfan 1986; Retherford and Alam 1985). In recent years, however, fertility has begun to decline in response to a rapidly increasing age at marriage and a rise in the prevalence of contraceptive use. Various summary measures of fertility have been calculated from the PDHS to provide a complete picture of recent fertility, including the crude birth rote (CBR), the general fertility rate (GFR), age-specific fertility rates (ASFR) and the total fertility rate (TFR). These estimates are described in the following sections. Crude Birth Rate The crude birth rate (per thousand population) is the least sophisticated measure of fertility, but it is the most commonly used and easily understood. Several attempts have been made to estimate the CBR in Pakistan, but there is still no agreement on its precise magnitude. In the PDHS, the CBR is calculated by summing the product of the age-specific fertility rates and the proportion of women in each age group out of the total de facto (male and female) population at all ages. Since the ASFRs relate to births during the past six years, the CBR calculated from the PDHS pertains to the same period and is centered on the years 1987-88. Table 4.1 shows the crude birth rates for selected years derived from various surveys. The PDHS estimates a CBR of 35 per thousand population. The CBRs estimated from previous surveys are 39 for the 1968-69 National Impact Survey (NIS), 41 for the 1975 Pakistan Fertility Survey (PFS), and 37 for the 1984-85 Pakistan Contraceptive Prevalence Survey (PCPS). The Pakistan Demographic Survey (PDS) estimated a CBR of 43 for each year during the period 1984-87 and 40 for 1988. The PDS rates are, however, based on indirect measures involving the matching of vital events and are considered to be on the high side. The CBRs by residence indicate that fertility is slightly higher in rural areas (36) than urban areas (34). The provincial differentials are also worth noting. The lowest CBR is observed for Sindh (33) and the highest for Balochistan (38). Karachi had a slightly higher CBR (36) than major cities as a whole (34). This may be due to an age-sex distribution which favours higher fertility or the somewhat lower level of contraceptive prevalence in Karachi than in other major cities. Comparing the CBRs from the 1984-85 PCPS and the 1990-91 PDHS, it is observed that there was a small decline of four percent in the CBR between the two surveys. The CBR declined in each type of place of residence and each province, except for major cities and for Sindh where it exhibited a slight increase. It should be noted that the CBRs obtained from various sources are the by-products of methodological procedures, response errors, enumerator biases, problems of coverage and sampling errors, which may be of different nature and magnitude in different surveys. However, it appears that the fertility transition at least started in Pakistan after the mid-1970s. 36 Table 4.1 Crude birth rates according to selected surveys Crude birth rate per thousand for selected surveys 1968-1991, Pakistan Urban-rural residence 1990-91 1984-85 1975 1968-69 and province PDHS PCPS PFS NIS Residence Total urban 33.7 U U U Major city 33.5 32.6 U U Other urban 34.0 38.5 U U Rural 35.6 37.1 U U Province Punjab 35.5 37.6 U U Sindh 32.8 32.0 U U Karachi 36.2 U U U NWFP 35.3 36.4 U U Balochistan 38.3 45.4 U U Total 35.0 36.6 40.5 39 Note: The period covered by the CBR estimates is six years prior to the interview for the PDHS and one year prior to the interview fur the other three surveys. The estimated crude birth rate from the PDHS for one year prior to the survey is 34.2. U = Unknown; no information Source: Population Planning Council of Pakistan (1976), Population Welfare Divison (1986) General Fertility Rate The general fertility rate (GFR) is calculated by dividing the number of births occurring during a specific period of time by the number of women of reproductive age (15-49 years of age) and multiplying the result by 1,000. The PDHS estimated the GFR to be 177 for the country as a whole--that is, 177 births to every 1,000 women (see Table 4.2). The observed GFR is higher in rural areas (184) than urban areas (163). Within urban areas, the GFR is lower in major cities (157) than in other urban areas (170). The highest GFR is observed in Balochistan (201), followed by Punjab and NWFP (178 each) and Sindh (171). The GFR for Pakistan was almost the same in the PDHS and the PCPS with an increase in urban rates and a slight decrease in rural rates between the two surveys. 37 Table 4.2 Currant fertility Age-specific and cumulative fertility rates and the crude birth rate f~ the six yeats preceding the survey, by urban-tufa/residence and province and for Karaehi. pakistan, 1990-91 Residence Province Total Major Other Age urban city urban Rural Punjab Sindh NWFP Ba/ochislan Karacbi Total 15-19 59 55 64 97 79 88 86 149 70 84 20-24 224 225 222 235 226 235 227 267 241 230 25-29 268 259 281 268 275 242 287 251 272 268 30-34 225 211 243 231 237 211 233 190 213 229 35-39 126 116 141 157 159 l lS 149 116 119 147 4044 [49] [47] [531 [851 [70] [79] [771 [82] a [63] a [73] 45-49 [29] [21] a [38] a [44] [30] a [51] a [41] a b b [40] TFR 15-49 4.90 4.67 5.21 5.58 5.39 5.12 5.50 5.84 5.03 5.36 TFR 15-44 4.86 4.56 5.02 5.36 5.24 4.86 5.30 5.28 4.89 5.16 GFR 162.5 157.0 170.4 184.1 178.0 170.7 177.8 201.1 173.5 177.0 CBR 33.7 33.5 34.0 35.6 35.5 32.8 35.3 38.3 36.2 35.0 Note: Rates are calculated for all women 15-49, using infornmtlon on women's age and marital states from the household questionna/le and on the number of births from the woman's questionnaire. Figures in brackets are partially truncated rates. aBased on fewer than 500 person-months of exposure bBased on fewer thma 250 person-months of exposure, rates not shown TFR: Total fe~lity rate expressed per woman GFR: Genera/fertility rate (births divided by number of women 15-44), expressed per 1,000 women CBR: Crude birth rate, expressed per 1,000 population Age-Specific and Total Fertility Rates Although the GFR is a more refined measure than the CBR, beth are crude summary measures. The changes observed in these rates may not provide a complete picture; better estimates of fertility can be obtained by examining the age-specific fertility rates (ASFRs) and the total fertility rate (TFR). A historical series of fertility estimates from four national surveys is shown in Table 4.3. The total fertility rate flTR) is a summary measure that indicates the number of children a woman would bear during her reproductive years if she were to experience the age-specific fertility rates prevailing at the tune of the survey. Mathematically, the 'I'FR is five times the sum of the age-specific fertility rates for each five-year age group. The Pakistan Contraceptive Prevalence Survey (PCPS) includes only currently married women in its sample, whereas the other three surveys interviewed ever-married women. In order to calculate the fertility rates for all women, it is assumed that no births occur outside of marriage. 38 Table 4.3 Current fertility according to selected surveys Age-specific fertility rates for selected surveys 1975-1991, Pakistan 1990-91 1984-85 1979-80 1975 Age group PDHS PCPS PLM PFS 15-19 84 64 99 131 20-24 230 223 283 275 25-29 268 263 313 315 30-34 229 234 263 259 35-39 147 209 188 188 40-44 73 127 101 77 45~19 40 71 48 11 Total fertility rate, 15-49 5.4 6.0 6.5 6.3 Source: Alam, Irfim mad Farooqui [1984] and Population Welfare Division (1986) According to the Pakistan Fertility Survey (PFS) and the Population, Labour Force and Migration Survey (PLM), the total fertility rate in the 1970s was between 6.3 and 6.5 children per woman. The PCPS recorded a drop to 6.0 children per woman and the PDHS registered a further decline to 5.4 children per woman (a decline of 10 percent since the 1984-85 PCPS and 15 percent since the 1975 PFS). 1 According to the PDHS, if current age-specific fertility rates were to remain unchanged in the future, the average woman in Pakistan would have 1.6 children by the time she reaches age 25, 2.9 children by age 30, more than four children by her thirty-filLh birthday, and 5.4 children by the end of her childbearing years. Trends in age-specific fertility rates are somewhat erratic, although fertility is generally lower in the two most recent surveys (see Table 4.3 and Figure 4.1). A comparison of the fertility estimates from the PFS and the PDHS shows that fertility declined most rapidly (by more than one-third) in the 15-19 age group, reflecting a pattem which is consistent with the increasing age at marriage. Substantial fertility declines are also evident at ages 20-39. lit should be noted, however, that the average "II~K estimated by the Pakistan Demographic Survey for 1984-1988 (6.9 children per woman) would suggest that fertility has not yet begun to decline in Pakistan (Federal Bureau of Statistics 1990). 39 Figure 4.1 Age-Specific Fertility Rates Pakistan, 1970-1991 Births per 1,000 Women 350 300 250 200 150 100 50 0 i i i i 15-19 20-24 25-29 30-34 35-39 40-44 Age 45-49 PDH$ t990-91 (6 yra) PLM 1979-80 (1975-79) - - PCP$ 1984-85 (1 yr) + PF8 1975 (1970-75) i Source : A lam, I r lan and Farooqgl [1984l; Population Welfare Division (1986) Differentials in fertility by type of place of residence are shown in Table 4.4 and Figure 4.2. Overall, urban areas have lower fertility rates than rural areas and within urban areas major cities have lower fertility. Fertility rates in urban and rural areas were very similar during the prime childbearing years (ages 20-34), but differences in urban and rural fertility levels arc striking in the youngest and oldest age groups (see Table 4.2). Overall, at current fertility rates, the average woman living in a large city can be expected to have ncarly one child less than her rural counterpart (4.7 children compared to 5.6 children). Provincial differences in fertility are quite modest. The TFR for women age 15-49 ranges from 5.1 in Sindh to 5.8 in Balochistan. For women age 15-44, the range of fertility estimates is even more restricted. At the provincial level, it is preferable to compare the estimates of fertility at ages 15-44 rather than 15-49 since the age-specific fertility rates at age 45.49 are based on only a small number of years of exposure to the risk of pregnancy. An additional reason for focusing on the 15-44 age group is that the l l ,R which includes women age 45.49 uses data which are progressively truncated as one moves back in time. A separate estimate of fertility is shown for Karachi, which has a total fertility rate of 5.0 for women age 15-49 and 4.9 for women age 15-44. While Karachi's fertility is lower than the national average, it is somewhat higher than the fertility reported for other major cities in Pakistan. As noted earlier, this finding is consistent with the fact that the reported contraceptive prevalence rate for Karachi is lower than that reported for other major cities. 40 Current fertility rates are related not only to the geographical area in which a woman resides but also to her educational attainment. Women without any formal education have a rvR that is nearly one child higher than women who have attended primary school and two children higher than women who have gone beyond the middle school level (see Table 4.4). Table 4.4 also shows the mean number of children ever born to women age 40-49---that is, women who are approaching the end of their child- bearing years. A comparison of this cumulative measure of childbearing with the "l'lq~. gives a rough indication of the trend in fertility over the last several decades. For all women, the total fertility rate is exactly one child less than the mean number of children ever born. This difference provides further evidence that fertility has started to decline in Pakistan, although if there are errors in recording recent births in the birth history, the decline may not be as rapid as the comparison suggests. The differences between the two meas- ures am greatest for women in Sindh and Punjab as well as for those living in urban areas (particularly in Kamchi and other major cities). Therefore, women in these areas appear to be leading the way in the early stages of the fertility decline. Educa- tional attainment is strongly related to both cumu- lative and current fertility levels, suggesting that educated women have been experiencing lower fer- tility for a long period of time. The most direct way of observing fertility trends is to examine changes in age-specific fertili- Table 4.4 Fertility b,y background characteristics Total fertility rate for the six years preceding the survey, and mean number of children ever born (CEB) to women 40-49 years of age, by selex:ted background characteristics Pakistan 1990-91 Total Me~m no. Background fertility of CEB eheraeteristic rate (women 40-49) Res~ence To~ ~ban 4.9 6.3 M~ore~ 4.7 6.3 Other~ban 5.2 6.4 Rural 5.6 6,4 Pro~nce Punj~ 5.4 6.3 Sin~ 5.1 6.6 Karac~ 5.0 7.1 NWFP 5.5 6.1 Ba~c~stan 5.8 5.7 Educatlon level attended No education 5.7 6.5 Primary 4.9 6.1 Middle 4.5 5.3 Secondary+ 3.6 4.3 To~ 5.4 6.4 Note: Figures are calculated for all women 15-49, using information on women's age and marital status from the household questionnaire and on the number of births from the woman's questionnaire. ty rates over time based on the PDHS birth history data (see Table 4.5). The trend in fertility over a period of more than two decades can be seen for women age 15-34. Only partial information is available for older women because of truncation in the data for earlier time periods. The lowest estimated fertility rate in every age group is observed for the most recent six-year period. A comparison of the two most recent periods reveals that estimated fertility has fallen most rapidly (by over forty percent) in the youngest and oldest age groups. Fertility also reportedly declined by more than one-quarter in every other age group. Declines of the magnitude shown for the middle and older age groups seem unlikely given the continuing low level of contraceptive prevalence in Pakistan. The fertility decline in these age groups is probably exaggerated by errors in the coverage and timing of births in the PDHS. This conclusion seems particularly warranted in light of the experience of previous demographic surveys in Pakistan. In evaluating the quality of three large-scale national demographic surveys and the 1981 census, Rethefford et al. (1987) noted that the estimated "tl"K fell below five children per woman during the two years preceding each of the surveys. However, there was no credible evidence that fertility had actually fallen during any of these periods. The authors concluded that fertility was severely underestimated in the five- year period preceding each survey. 41 Figure 4.2 Total Fertility Rate (TFR) and Mean Number of Children Ever Born (CEB) RESIDENCE Urban Rural PROVINCE Punjab Sindh NWFP Balochistan EDUCATION No edue. Primary Middle Seoondary* 1 2 3 4 5 6 Number of Births [ ~ TFR (Preoeding 6 yrs) ~ CEB (Women 40-49) PDHS 1990"91 Table 4.5 Fertility trends Age-specific fertility rates for six-year periods preceding the survey, by mother's age at the t/me of birth, Pakistan 1990-91 Number of years preceding survey Mother's age 0-5 6-11 12-17 18-23 15-19 84 145 139 116 20-24 230 317 294 278 25-29 268 367 331 [3341 30-34 229 319 [3091 ,. a 35-39 147 [212] [227]" U 4044 [73] [137] U U 45-49 [4O] U U U Note: Age-specific fertility rat~s are per 1,000 womea~, Figures in brackets are partially I~qmeated rates. U = Unknown; no hfform.ation Based on fewer than 250 person-months of exposure, rates not shown ased on fewer than 500 person-months of exposure 42 Table 4.5 also shows that fertility registered modest gains in the two earliest intervals between six- year periods. Fertility may have actually risen over that time period but possible data errors again need to be considered in interpreting the results. The apparent rise in fertility in the earlier periods may be due to a commonly observed tendency among older women to underreport early births or to displace the birth dates of those children forward in time (Potter 1977). The pattern of fertility change over time discussed above is also evident for women with marital durations of less than 10 years, but estimated fertility has fallen steadily over time for women married more than 15 years (see Table 4.6). In the most recent six-year period, the estimated fertility rate declines consistently as the duration of marriage increases. Women who have been married for more than 15 years reported sharply lower fertility than women with shorter marriage durations. Table 4.6 Fertility by marital duration Fertility rates for ever-married women by duration since first marriage in years fur six-yeas periods preceding the survey, Pakistan 1990-91 Number of years preceding survey Mastiagedurafion at birth 0-5 6-11 12-17 18-23 04 339 386 343 309 5-9 298 412 381 379 10-14 248 346 325 341 15-19 186 263 286 a 20-24 98 187 a a 25-29 52 (95) a NA Note: Rates in parentheses are based on fewer than 500 person-months of IgXpe$1~'l~, NA = Not applicable abased on fewer than 250 person-months of exposure, rates not shown 4.2 Children Ever Born The cumulative number of children ever born is shown in Table 4.7 for all women and for currently married women. The figures for all women are calculated by assuming that all births occur within marriage. Women in their childbearing years in Pakistan have had an average of three children and currently married women have had an average of just over four children. The steady increase in the average number of children ever born by age is a normal function of the family building process. Women who are currently at the end of their childbearing years have had, on average, more than six births. Mortality has had a significant impact on family size, however, since the average woman age 40-49 has had one child who died by the time of the survey. Taking the difference between the mean number of children ever born (6.6) and the mean number of children surviving (5.6), it is seen that, overall, there has been a loss of 15 percent of births among currently married women age 40-49. Early childbearing is relatively rare in Pakistan. Only 12 percent of women in the 15-19 age group have ever had a child and even in the 20-24 age group only a minority of women have ever given birth. Eventually, however, nearly all women bear children. Among currently married women age 35-49, 43 only three percent have never had a child. This low level of childlessness indicates that primary sterility is low in Pakistan. This is consistent with the findings of a low level of primary sterility in the 25 DHS surveys carried out from 1986 to 1989 (Arnold and Blanc 1990). T@le 4.7 Children ever born and livin s Percent distribution of all women and cummtly marred wo~nen by nmnbcr d children ever born (CEB) and mean number of children ever born and living, accorc~g to age group, Paldllan 1990-91 Number of d~Idnm ever born Age 0 1 2 3 4 5 6 7 8 9 Number Mean Mean no. of no, of of living 10+ Tond women CEB children ALL WOMEN 15-19 87.8 9.2 1.9 1.0 0.I . . . . . . I0¢.0 1720 0.2 0.I 20-24 54.3 17.0 15.3 8.2 3.6 0.9 0.6 0.I . . . . I(~.0 1747 1.0 0,8 25-29 23.0 10.4 15.0 17.4 15.3 11.0 4,6 2.3 0.8 0.2 0.I 100.0 1745 2.6 2.3 30-34 9.2 5.0 8.8 14.8 16.7 13.6 12,7 9.4 5.1 2.9 1.9 I00.0 1241 4.3 3.7 35-39 5.4 3.4 4.3 8.6 10.2 16.8 16,9 12.9 8.8 6.2 6.4 I00.0 1005 5.5 4.8 40-44 5.5 1.5 3.0 7.3 9.0 12.0 13.2 15.4 11.5 7.1 14.6 100.0 865 6.3 5.4 45-49 5,5 2.7 3.3 7.7 5.8 10.5 13,4 11.4 13.8 I0.0 15.8 I(i0.0 630 6.4 5.5 To~al 34.7 8.5 8.5 9.4 8.4 8.0 6,9 5.5 3.9 2,5 3.5 I00.0 8953 3.0 2.6 CURRENTLY MARR/ED WOMEN 15-19 51.4 36.8 7.5 4.0 0.2 . . . . . . . . 100.0 418 0.6 0.6 20-24 24.3 27.9 25,2 13.7 6.1 1.6 1.1 0.1 -- 100.0 1041 1.6 1.4 25-29 9.5 IL8 17.2 20.5 18,2 13.2 5.5 2.7 0.9 0.2 0.2 100.0 1452 3.1 2.7 30-34 4.7 4.6 8.7 15.4 17.8 14.5 13.7 I0.0 5.4 3.1 2.1 I00.0 1147 4.6 4,0 35-39 2.7 3.3 4.6 9.0 8.9 I%5 17.7 13.5 9.4 6.6 6.8 I00.0 932 5.7 5.0 40-44 3.1 1,1 3.1 7.3 8.9 11.7 13.5 16.4 12,3 7.1 15.5 100.0 803 6.5 5.6 45-49 3.8 2.8 3.5 7.4 6.1 10.1 13.7 11.4 14.5 10.4 16.4 100.0 572 6.6 5.6 Total 11.5 11.4 11.5 12.8 11.4 10.8 9.4 7.5 5,4 3.4 4.8 100,0 6364 4,1 3.5 -- L~ss than 0.05 percent Although currently married women age 45-49, on average, have had 6.6 children, the range of family sizes is quite wide. Ten percent have had fewer than three children and an additional 13 percent have had three or four children. About one-quarter (24 percent) have had five or six children and another quarter (26 pcrecnt) have had seven or eight children. At the high end of the spectrum, more than one in four women have had nine or more live births and one in six women have had 10 or more births. More than half of currently married women age 45-49 have had at least seven live births. Cumulative fertility for currently married women has shown some signs of a decline over time in every age group except age 15-19 (see Table 4.8). Although the overall mean number of children ever born was identical in the 1975 Pakistan Fertility Survey and the 1984-85 Pakistan Contraceptive Prevalence Survey, the mean number of children ever born declined slightly during that period in the 44 majority of age groups. Between the 1984-85 PCPS and the 1990-91 PDHS, the overall mean number of children ever born declined from 4.3 to 4.1. Because of a decline in mortality during that period, however, the mean number of surviving children remained constant at 3.5 children per woman. In fact, the mean number of surviving children aetoally increased from 3.2 to 3.5 between the time of the 1975 PFS and the 1984-85 PCPS, again due to a decline in mortality between the two surveys rather than to an increase in the cumulative number of children ever born. Table 4.8 Children ever born ~ceording to selected surveys Mean number of children ever born and children still living for currently married women, by age group, selected surveys 1975-1991, Pakistan Mean number of children ever bern Mean number of living children 1990-91 1984-85 1975 1990-91 1984-85 1975 Age PDHS PCPS PFS PDblS PCPS PFS 15-19 0.6 0.6 0.6 a 0.6 0.6 0.5 s 20-~ 1.6 1.8 1.9 1.4 1.5 1.5 25-29 3.1 3.4 3.4 2.7 2.8 2.8 30-34 4.6 5.0 5.2 4.0 4.2 4.0 35-39 5.7 6.1 6.4 5.0 5.1 4.9 40-44 6.5 7.0 7.5 5.6 5.5 5.2 45-49 6.6 7.5 7.4 5.6 5.7 5.1 15-49 4.1 4.3 4.3 a 3.5 3.5 3.2 a alncludes currently married women age 10-14 Source: Population Welfare Division (1986) and Population Planning Coencil of Pak'tstan (1976) Differentials in cumulative fertility by selected socioeconomic characteristics of respondents and their husbands are shown in Table 4.9. The largest differentials are observed for the woman's educational attainment. Women with some secondary school education or higher have 1.4 fewer children, on average, than women with no education. For the oldest age group (age 35 or higher) this differential widens to more than two children per woman. Differentials in fertility are less pronounced for the husband's education, particularly for men whose wives are in the youngest age groups. Women whose husbands are in professional, technical, clerical or service jobs have a relatively small number of children ever born. It is surprising, however, that the woman's own work experience is only weakly related to her cumulative fertility. 45 Table 4.9 Mean number of children ever bern Mean number of children ever born by background ehm'acteristies of currently married women mad their husbands according to age, Pakistan 1990.91 Age of woman Background characteristic 15-24 25-34 35+ Total Woman's education No education 1.3 3.9 6.4 4.3 Primary 1.5 3.7 6.0 3.7 Middle 1.1 3.7 5.3 3.5 Secooda~'+ 1 .O 2.8 4.2 2.9 Woman's work status l Currently working 1.5 4.0 6.3 4.4 Worked previously 1.5 3.5 6.1 3.8 Never worked 1.3 3.7 6.2 4.0 Husband's education No education 1.3 4.0 6.4 4.4 Primary 1.4 3.7 6.5 4.1 Middle 1.3 3.6 6.7 3.7 Secondary+ 1.3 3.4 5.3 3.6 Husband's occupation Professional, technical 1.5 3.I 5.8 3.7 Administrative, managerial 1.5 3.4 5.1 4.1 Clerical 1,3 3.3 5.8 3.7 Sales 1.3 4.1 6.1 4.3 Service 1.2 3.4 5.6 3.3 Agriculture, fishing 1,3 3.8 6.3 4.3 Production, transportation, labor 1.4 3.8 6.4 4.1 Not classifiable 1.0 3.7 6.1 3.9 Total 1.3 3.7 6.2 4.1 :"Worked previously" means that the women is not eurently working and she worked before marriage end/or just after marriage. '*Never worked" means that the woman is not cur~erttly working and she did not work either before marriage or just after marriage. 4.3 Birth Intervals Previous research has demonstrated that children bom too close to the time of a previous bi~ah are at increased risk of dying. The risk is particularly high when the interval between births is less than 24 months. Previous birth intervals for children born in the five years preceding the survey are shown in Table 4.10. The median interval since the previous birth is 29 months. One of every three bit'tlm occurred less than 24 months after the previous birth and half of those had very short birth intervals of less than 18 months. Another one-third of births (36 percent) had previous birth intervals of two years and the remaining one-third (31 percent) had intervals of three years or more. 46 Tat~e 4.10 Birth intervals Percent distribution of births in the five years preceding the survey by number of months since previous birth, aeeoxding to selected background ebaracteristies, paldstan 1990-91 Median Number of months since previous birth months since Number Background previous of ebaraet~ristie 7-17 18-23 24-35 36-47 48+ Tolal birth b'u'ths Age 15-19 32.4 21.0 39.6 6.7 0.3 100.0 23.7 68 20-29 19.8 18.7 37.0 12.6 11.9 100.0 26.8 2452 30-39 14.3 15.9 35.7 14.4 19.8 100.0 30.8 2242 40+ 10.3 10.9 27.8 15.7 35.4 100.0 37.9 547 Birth order 2-3 20.2 16.7 36.6 12.5 14.0 100.0 27.4 2056 4-6 13.6 16.4 37.0 13.8 19.2 100.0 30.0 2085 7+ 16.0 17.3 31.1 14.9 20.7 100.0 30.4 1169 Sex of prior birth Male 17.7 16.2 33.3 14.8 18.0 100.0 29.1 2762 Female 15.6 17.3 37.9 12,3 16.9 100.0 29.0 2608 Survival of prior birth Living 14.7 16.3 36.6 13.9 18.7 100.0 30.0 4670 Dead 31.2 20.2 28.5 11. l 9.0 100.O 23.7 640 Residence Total urban 20.6 19.5 33.0 12.6 14.2 100.0 26.5 1649 Major city 21.9 20.6 31.9 i2.1 13.6 100.0 25.7 938 Other urban 19.0 18.0 34,5 13.4 15.1 100.0 27.7 712 Rural 14.9 15.5 36.7 14.0 19.0 100.0 30.3 3660 Province Punjab 16.3 16.3 38.4 13.2 15.8 100.0 28.9 3238 Sindh 19.9 18.1 26.8 11.4 23.9 100.0 28.3 1158 NWFP 11.5 17.3 37.2 17.2 16.9 100.0 31.3 709 Balochislan 22,5 13,7 34,3 18,8 10.7 100,0 27,3 205 Education level attended No education 15.0 16.4 36.3 14.0 18.4 100.0 29.8 4192 Primary 22.2 20.2 32.3 12.9 12.5 100.0 26.2 524 Middle 26.3 14.2 39.3 7.7 12.5 100.0 25.9 233 Secondary+ 22.6 17.1 29.5 13.9 17.0 100.0 27.5 361 Total 16.7 16.7 35.6 13.6 17.5 100.0 29.1 5310 Note: First-order births axe excluded. The interval for multiple births is the number of months since the preceding pregnancy that ended in a live birth. The median birth interval is relatively short for younger women, for urban residents, for women living in Balochistan and for women who had received some formal education. In Balochistan, birth intervals are relatively short for both modem (urban, educated) women and traditional women. This 47 f'mding is probably due to a complex set of circumstances including negligible contraceptive use in Baloehistan and a later age at marriage and shorter breastfeeding among modem women. Birth intervals are also shorter than average for second and third order births, Second and third order births were also most likely to be in the high risk group---that is, births occurring within 24 months of the preceding birth. As expected, children whose prior sibling had died before the time of the survey had the shortest previous bilth intervals. A majority of children whose prior sibling had died were born less than 24 months after the birth of the previous child. 4.4 Age at First Birth The age at which a woman bears her first child has important demographic and health consequences. On the demographic side, early initiation into childbearing is generally a major determinant of large family size and rapid population growth, particularly in countries in which family planning is not widespread. On the health side, beating children at an early age entails significant risks to the health of both the mother and the child. Early childbearing also tends to restrict educational and economic opportunities for women. Table 4.11 presents the distribution of Pakistani women by their age at first birth. The majority of women in Pakistan did not have their first birth unfd after their twentieth birthday. Childbearing before age 15 has always been uncommon and it is becoming increasingly rare over time. More than one-fifth of women age 25-49, however, had their first birth before age 18 whereas about 40 percent had their first birth during their teenage years. In recent years, there has been a rapid decline in the extent to which women begin childbearing during their teenage years. For example, whereas 42 percent of women age 25-29 had their first birth before age 20, only 30 percent of women age 20-24 had their first child that early. Table 4.11 Age at first birth Percent dis~bution of ever-manSed women by age at first birth, nccording to current age, Pakistan 1990-91 Current age Median Women Age at first birth Number age at with no of first birth <15 15-17 18-19 20-21 22-24 25+ Total women birth 15-19 87.8 1.5 6.7 4.1 . . . . 100.0 1720 a 20-24 54.3 3.3 13.9 13.3 10.6 4.8 -- 100.0 1747 a 25-29 23.0 5.3 18.7 18.1 15.7 14.0 5.3 100.0 1745 21.0 30-34 9.2 4.4 20.0 17.3 16.1 19.7 13.3 100.0 I?AI 20.9 35-39 5.4 3.4 16.4 18.4 16.5 20.6 19.3 100.0 1005 21.4 40-44 5.5 6.3 15.6 15.6 15.5 19.8 21.8 100.0 865 21.7 45-19 5.5 4.8 12.4 15.7 12.9 22.2 26.5 100.0 630 22.6 -- Less than 0.05 percent aLess than 50 percent of the women have had a birth by the beginning of the age group. Differentials in the age at first birth are shown in Table 4.12. The median age at first birth for all women age 25-49 is 21.3 years. Overall, there is little variation in the median age at first birth by place of residence or by education, except for women who have attended the highest education level. For the youngest age group, the median age at first birth is lowest in rural areas, in Baiochistan and among 48 Table 4.12 Age at first bitlla by backh, ruund characteristics Median age at first birth among woman aged 25-49 yeats, by current age and selected background chsaacteristics Cm~ant age Background Ages charaetea'isfic 25-29 30-34 35-39 40-44 45-49 25-49 Residence Total urban 21.7 20.9 20.8 21.5 21,4 21,3 Major city 21.8 21.0 20.6 20.7 21.3 21.1 Other urban 21.6 20.8 21.2 22.1 21.5 21.4 Rural 20.6 20.9 22.0 21.7 23.2 21.3 Province Punjab 21.6 21.0 21.6 21.8 22.7 21.6 Sindh 19.8 20.6 21.2 21.0 21.7 20.6 NWFP 21.2 20.8 21.0 21.3 23.6 21.4 Balochistan 18.3 20.4 22.0 22.7 25.3 20.3 Education level attended No education 20.3 20.5 21.3 21.6 22.8 21.0 Primary 20.7 21.5 21.5 20.0 21.0 21.1 Middle 22.1 20.2 19.7 20.6 23.2 21.1 Secondary+ 25.0 24.7 23.0 23.3 22.5 24,0 Total 21.0 20.9 21.4 21.7 22.6 21.3 women with little or no education. These patterns, however, are not regular across all age groups. 2 The most consistent pattern is the late initiation of childbearing among women who have gone beyond middle school. 4.5 Teenage Fertility Some information on teenage fertility was already presented in the section on age at first birth. More detailed findings on teenage fertility are discussed in this section. Table 4.13 presents information on the childbearing experiences of women age 15-19. Column one shows the percentage of teenagers who are already mothers; column two shows the percentage who are pregnant with their first child. The sum of these two columns indicates the percentage of young women who have already begun childbearing. Overall, one in eight teenage women was a mother and another four percent were pregnant with their first child at the time of the survey. The proportion who have started childbearing increases with age. For example, at ages 15 and 16, only six percent of women have begun childbearing. After age 16, the proportion increases steadily to a level of 31 percent by age 19. While these figures demonstrate that there is a substantial amount of teenage childbearing in Pakistan, it is noteworthy that more than two-thirds of women who are 19 years old have not begun ch'tldbearing. Early childbearing is particularly characteristic of rural women and women who have not attended school. Regional differences in early childbearing are not as pronounced, but women in Balochistan are somewhat more likely to begin childbearing early than are women in other provinces. ~'he results for women age 45-49 should be interpreted cautiously since in demographic surveys older women often omit their first birth (particularly if the child died) or report the timing of their first birth erroneously. 49 Table 4.13 Teenage fertility Percentage of teenagers 15-19 who are mothers or pregnant with their first child, by selected background characteristics, Pakistan 1990-91 Percentage who are: Percentage Pregnant who have Number Background with ftrst begun of characteristic Mothers child childbearing teenagers Age 15 3.0 2.5 5.5 173 16 3.7 2.5 6.1 381 17 7.6 5.4 13.0 260 18 15.1 3,8 18.9 630 19 27.4 3.3 30.6 276 Residence Total urban 7.8 2.1 9.9 583 Major city 5.7 2.1 7.9 318 Other urban 10.2 2,1 12.3 264 Rural 14.5 4.3 18.7 1137 Province Punjab 12.6 4.3 16.9 1015 Sindh 12.0 2.4 14.4 345 NWFP 11.7 2.2 13.9 281 Balochistan 15.3 5.2 20.5 52 Education level attended No education 17.4 5.0 22.4 1007 Primary 8.3 2.I 10.4 248 Middle 4.9 -- 4.9 115 Secondary+ 3.1 2. I 5.3 256 Total 12.2 3.5 15.7 1720 -- Less than 0.05 percent Teenage women who have begun childbearing are not fikely to have had more than one birth (see Table 4.14). Only one respondent age 15-17 had two or more births and fewer than 10 percent of women age 19 had given birth to more than one child. Overall, g8 percent of women 15-19 have never given birth and only three percent have delivered more than one child. These findings suggest that the recent increase in the average age at first marriage in Pakistan has had a dampening effect on early childbearing and caused the large majority of women to delay childbearing at least until they have completed their teenage years. 50 Table 4.14 Children ever born to teenagers Percent dis~bution of teenagers 15-19 by number of children ever born (CEB), sccording to single year of age, Pakistan 1990-91 Number of children ever bern Age 0 1 2+ Total Mean Number number of of CEB teenagers 15 97.0 3.0 -- 100.0 173 16 96.3 3.7 -- 100.0 381 17 92.4 7.3 0.3 100.0 0.1 260 18 84.9 11.3 3.8 100.0 0.2 630 19 72.6 17.9 9.5 100.0 0.4 276 Total 87.8 9.2 3.0 100.0 0.2 1720 -- Less then 0.05 percent or mean less then 0.05 children 51 REFERENCES Alam, Iqbal, Mohammad lffan and Naseem Iqbal Farooqui. [1984]. Fertility Levels, Trends and Differentials in Pakistan: Evidence from the Population, Labour Force and Migration Survey 1979-80. Studies in Population, Labour Force and Migration, Project Report No. 1. Islamabad: Pakistan Institute of Development Economics. Amold, Fred. 1990. Assessment of the Quality of Birth History Data in the Demographic and Health Surveys. In An Assessment of DHS-I Data Quality, 81-111. DHS Methodological Reports, No. 1. Columbia, Maryland: Institute for Resource Development/Macro Systems, Inc. Amold, Fred and Ann K. Blanc. 1990. Fertility Levels and Trends. DHS Comparative Studies, No. 2. Columbia, Maryland: Institute for Resource Development/Macro Systems, Inc. Federal Bureau of Statistics [Pakistan]. 1990. Pakistan Demographic Survey - 1988. Karachi: Statistics Division. Population Census Organisation [Pakistan]. 1984. 1981 Census Report of Pakistan. Islamabad: Population Census Organisation (Census Report No. 69). Population Census Organisation [Pakistan]. 1985. HandBook of Population Census Data. Islamabad: Statistics Division. Population Planning Council of Pakistan. 1976. Pakistan Fertility Survey: First Report, 1976. Voorburg, Netherlands: Intemational Statistical Institute. Population Welfare Division [Pakistan]. 1986. Pakistan Contraceptive Prevalence Survey, 1984-85. Islamabad: Ministry of Planning and Development. Potter, Joseph E. 1977. Problems in Using Birth History Analysis to Estimate Trends in Fertility. Population Studies 31(2):335-64. Retherford, Robert D. and lqbal Alam. 1985. Comparison of Fertility Trends Estimated Alternatively from Birth Histories and Own Children. Papers of the East-West Population Institute No. 94. Honolulu: East- West Center. Retherford, Robert D., G. Mujtaba Mirza, Mohammad Irfan, and Iqbal Alam. 1987. Fertility Trends in Pakistan -- The Decline That Wasn't. Asian and Pacific Population Forum 1(2):1-10. Rukanuddin, Abdul Razzaque and M. Naseem Iqbal Famoqui. 1988. The State of Population in Pakistan, 1987. Islamabad: National Institute of Population Studies. Shah, Iqbal H. and John G. Cleland. 1988. High Fertility in Bangladesh, Nepal, and Pakistan: Motives and Means. Paper presented at the IUSSP Seminar on "Fertility Transition in Asia: Diversity and Change," Chulalongkom University, Bangkok, March 1988. Shah, lqbal H., Thomas W. Pullum, and Muhammad Irfan. 1986. Fertility in Pakistan during the 1970s. Journal of Biosocial Science 18(2):215-29. 52 CHAPTER 5 KNOWLEDGE AND USE OF FAMILY PLANNING Nasra M. Shah and Syed Mubashir Ali In a country with a persistently high population growth rate such as Pakistan, the level of knowledge about family planning and the use of family planning methods remain very important demographic issues. Information on contraceptive use by various methods is of particular importance to policy makers, pro- gramme managers and researchers for formulating future programme strategies. An assessment of the extent of knowledge and use of contraception, therefore, constituted one of the primary objectives of the PDHS. This chapter describes women's knowledge of modem and traditional contraceptive methods as well as of their sources, the level of contraceptive use and the timing of contraceptive initiation, accessibility of family planning services and the cost of contraceptive methods. Differentials in knowledge and use according to demographic and socioeconomic characteristics of the respondents are also discussed. 5.1 Knowledge of Family Planning Methods and Sources Levels and Trends The question used to elicit knowledge about family planning was phrased: "Now I would like to talk about family planning--the various ways or methods that a couple can use to delay or avoid a pregnancy. Which ways or methods have you heard about?" The respondent was first asked to report all the methods she knew without any prompting. Once she completed her spontaneous reporting, the interviewer read out the names and a short deseription of the remaining methods on the list and asked if she knew each one of them. In this way, her "complete" knowledge of contraception was obtained. The contraceptive methods included in the survey are shown in Table 5.1. Almost four-fifths of ever-married and currently married women reported knowledge of at least one method. Almost all the women who reported such knowledge knew of a modem method. One-quarter of all women knew of a traditional method, mostly periodic abstinence or withdrawal. Female sterilisation, the pill, and injection were the best known methods. A distinction can be made, however, between prompted and unprompted knowledge. If just unprompted knowledge is considered, only about half of all women reported that they knew any method. Women who reported knowing of a method were asked if they knew where they could go to obtain the method. Of all currently married women, 25 to 30 percent knew where they could obtain the pill, an IUD, or an injection. The largest percentage (37 percent) knew where they could get sterilisation services. In the case of periodic abstinence, only 9 percent of women knew where to get advice on how to use this method. 53 Table 5.1 Knowledge and source of contraceptive methods Percentage of ever-married women and currently married women who know specific contraceptive methods mad who know a source (for information or seawices), by specific method, Pakistan 1990-91 Ever-married women Currently married women Ever- Currently manied rmnried womcn womex l Contraceptive Know Unprompted Prompted Know Unprompted Prompted Know a Know a method method knowledge knowledge method knowledge knowledge sourc~ source Any method 77.9 49.1 28.8 77.9 49.3 28.6 46.2 46.3 Any modern method 77.3 46.4 30.8 77.2 46.7 30.5 44.8 44.9 Pill 62.2 30.2 32.0 62.2 30.2 32.0 29.6 30.1 IUD 51.6 19.2 32.3 51.5 19.3 32.1 24.9 25.1 Injection 62.1 27.7 34.3 62.2 27.9 34.3 29.8 30.1 Vaginal method 12.7 3.1 9.5 12.7 3.1 9.6 7.0 7.0 Condom 35.0 14.7 20.3 35.3 14.8 20.5 19.1 19.3 Female sU~filisation 69.6 24.0 45.6 69.7 24.2 45.5 37.0 37.0 Male sterilisation 20.2 3.0 17.2 20.2 3.0 17.3 11.4 11.5 Any traditional method 25.6 10.4 15.2 25.7 10.4 15.3 NA NA Periodic abstinence 17.8 5.7 12.0 17.8 5.7 12.0 9.3 9.4 Withdrawal 14.2 3.2 11.0 14.3 3.2 11,0 NA NA Other 3.5 3.5 NA 3.5 3.5 NA NA NA Number of women 6611 6611 6611 6364 6364 6364 6611 6364 NA = Not applicable Comparison of the level of contraceptive knowledge in the PDHS with earlier surveys reveals some unusual patterns. The comparisons are not straightforward, since some surveys included prompted responses about specific methods, while others included only the respondent's spontaneous (unprompted) knowledge. Looking only at unprompted knowledge, the percentage of currently married women who reported knowing about any method fluctuated from 76 percent in the 1975 Pakistan Fertility Survey (PFS) to 26 percent in the 1979-80 Population, Labour Force and Migration Survey (PLM) to 49 percent in the 1990-91 PDHS. Looking at prompted and unprompted knowledge combined, the percentages still vary greatly, from 97 percent of married women interviewed in 1968-69 in the National Impact Survey--when the family planning programme was only four years old---to 62 percent in the 1984-85 Contraceptive Prevalence Survey and 78 percent in the PDHS (see Table 5.2). Thus, it appears that the overall level of contraceptive knowledge, which showed a decline in the last two decades, has made headway in recent years, assuming that the measurement in the PDHS was of the same quality as in the earlier surveys. The reason for the fluctuations in the level of contraceptive knowledge is uncle~ however, it is possible that the social climate that influences the respondent's reporting of knowledge of contraception might have become more restrictive beginning in the late 1970s, as was poinled out by analysts of the 1979-80 Pakistan Population, Labour Force and Migration Survey (Soomro et al. [1984]). 54 Table 5.2 Trends in contraceptive knowledge Percentage of currently married women who know specific contraceptive methods, 1990-91 PDHS, 1984-85 PCPS, and 1968-69 N-IS Contraceptive 1990-91 1984-85 1968-69 method PDHS PCPS NIS Any method 77.9 61.5 97.0 Pill 62.2 54.1 37.7 IUD 51.5 43.4 72.1 Injection 62.2 46.7 U Vagina/method 12.7 16.2 38.8 Condom 35.3 28.9 42.3 Female sterilisation 69.7 50.6 47.9 Male sterilisation 20.2 18.8 36.7 Periodic abstinence 17.8 5.8 13.7 Withdrawal 14.3 9.0 16.5 Note: Figures are for total knowledge,, i.e., knowledge based on unprompted (spontaneous) and prompted responses. U = Unknown; no information Source: Population Welfare Division (1986); Soomro et al. [1984]. Differentials in Knowledge Table 5.3 shows the knowledge of modem contraceptive methods and the source for methods among currently married women by selected background characteristics. In terms of respondent's age, women 30-39 years, who are likely to have the greatest need for contraception, reported the highest levels of knowledge (81 percent had heard of a modem method), Only 66 percent of women age 15-19 knew of a modem method of contraception. There are large differences in reported knowledge between urban and rural areas (see Figure 5.1). Ninety-four percent of currently married women residing in major cities knew of at least one modem method of contraception and three-fourths knew where to obtain a method. Among rural women, 71 percent knew of a modem method and 34 percent knew where to obtain a method. A comparison with earlier surveys indicates that the urban-rural differential in knowledge of family planning has increased over time (data not shown). While women in the provinces of Punjab, Sindh and NWFP had knowledge of modem methods ranging from 74 to 83 percent, only 37 percent of the Balochi women reported knowing of a modem method. Similarly, only about half as many B alochi women knew a source for a modem method as women in the other provinces. 55 Table 5.3 Knowledge of modem contraceptive methods and source for methods Percentage of currently married women who know at least one modem contraceptive method and who Imow a source (for information or services), by background characteristics, Pakistan 1990-91 Know Know Know source for Number Background any a modem modem of characteristic method method 1 method women Age 15-19 66.3 65.8 32.3 418 20-24 75.0 74.4 39.2 1041 25-29 77.4 76.7 43.6 1452 30-34 81.8 81.2 48.4 1147 35-39 81.5 81.3 53.2 931 40-44 78.7 77.9 47.2 803 4549 77.8 76.6 44.1 572 Residence Total urban 91.3 90.6 69.9 1930 Major city 94.5 93.8 75.7 1098 Other urban 87.2 86.4 62,3 832 Rural 72.0 71.4 34.0 4434 Province Punjab 80.6 79.9 46.2 3768 Sindh 74.4 73.9 45.3 1486 NWFP 83.6 83.3 44.8 856 Balochistan 38.5 36.5 23.8 254 Education level attended No education 73.8 73.0 37.7 5044 Primary 91.7 91.7 64.3 573 Middle 93.6 93.6 72.4 279 Secondary+ 95.6 95.0 81.9 468 Work status 2 Currently working 72.7 72.5 40.0 1033 Worked before marriage only 85.8 85.7 57.7 282 Worked after marriage only 81.9 81.9 64.1 43 Worked before and 75.7 74.5 39.2 69 after man'iage Neve~ worked 78.5 77.7 45.0 4916 Total 77.9 77.2 44.9 6364 IInolud~ pill, IUD, injec~on, vaginal methods (diaphragm/foam/jelly), condom, female sun'ilisa~un md male su~ilisatiun 2Excludes 21 women with miuing informsfion on work status. "Never worked" means that the woman is not ~dy working and she did not work either before marriage or just after marriage. 56 100 80 60 40 20 Figure 5.1 Knowledge of Modern Contraceptive Methods and Sources among Currently Married Women by Residence Percent o _/ Major City Other Urban Rural m Know a Modern Method ~ Know a Source i I PDHS 1990-91 The major difference in contraceptive knowledge by education is between women with no education and those who have at least some education. Only 73 percent of women with no education knew of a modem method, compared with 92 percent of those who had gone to primary school and 95 percent of those who had received at least some education at the secondary level. Similarly, only 38 percent of the women with no education knew where to obtain a method, compared with 82 percent of the ones with secondary or higher education. Finally, the relative level of contraceptive knowledge of working and nonworking women does not show any clear pattern. One might expect worldng women to have greater exposure and knowledge; however, the results in Table 5.3 indicate that this is not always the ease. Women who worked only before or alter marriage are more likely to know a modem contraceptive method (86 percent and 82 percent, respectively) than those who have never worked (78 percent), but those who have never worked are slightly more likely to know a modem method than those who are currently working (73 percent). Earlier research in Pakistan has shown that working women typically belong to the lower socioeconomic stratum (Shah 1986) and are likely to be less knowledgeable about contraception. This is consistent with the finding in Table 3.10 that working women are the least educated group overall. 57 5.2 Contraceptive Use Ever Use of Family Planning Methods In the PDHS all respondents who knew at least one method were asked whether they had ever used the known methods. This was further probed by asking whether they "ever used anything or tried in any way to delay or avoid getting pregnant." Table 5.4 shows that one-fifth of ever-married and of currently married women had used a contraceptive method at some time in the past. Sixteen percent of currently married women had used a modem method, while 9 percent had used a traditional method, Periodic abstinence was the most commonly used traditional method. Table 5.4 Ever use of contraception Percentage of ever-marrled and of currently married womea who have ever used any conlxaceptive method, by specific method and age. Pakistan 1990-91 Age of woman Contraceptive method 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Total EVER-MARRIED WOMEN Any method 3,3 12.7 18.6 23.5 31.1 25.5 19.1 20.3 Any modern method 2.4 8.0 15.1 19.1 25.1 20.0 15.4 16.0 Pill 1.1 2.2 3.7 5.7 6.0 7.0 4.2 4.5 IUD 0.6 1.4 4.0 4.9 4.1 3.9 2.2 3.4 Injection 0.4 1.5 2.7 4.0 5.9 3.2 3.0 3.2 Vaginal method --- 0.4 0.3 0.4 1.1 0.4 0.4 0.5 Condom 1.4 3.9 8.8 9.2 10.2 6.6 4.6 7.1 Female sterilisation 0.5 0.8 2.7 7.6 7.6 6.6 3.5 Male sterilisation -- 0.1 0.1 0.2 -- 0.4 0.1 Any traditional method 1.2 6.4 8.5 9.9 13.3 9,9 6.6 8.7 Periodic abstinence 1.1 2.4 5.2 5.2 8.2 6.3 3.7 4.9 Withdrawal 0.5 3.0 3.6 5.3 6.2 2.7 1.9 3.7 Other 0.2 2.5 1.4 1.8 1.6 2.7 2.1 1.8 Nuraber of women 428 1059 1494 1187 981 844 617 6611 CURRENTLY MARRIED WOMEN Any method 3.1 12.9 19.1 24.0 32.1 25.9 19.6 20.7 Any modern method 2.3 8.2 15.5 19.4 25.8 20.1 15.8 16.2 Pill 0.9 2.3 3.7 5.9 6.3 7.0 4.5 4.5 IUD 0.4 1.4 4.1 4.9 4.1 3.6 1.8 3.3 Injection 0.4 1.6 2.7 4.2 6.2 3.2 3.3 3.3 Vaginal method -- 0.4 0.3 0.4 1.2 0.4 0.4 0.5 Condom 1.3 4.0 9.1 9.4 I0.I 6.2 4.7 7.2 Female sterilisation -- 0.5 0,9 2.7 8.0 8.0 6.8 3.5 Male sterilisation . . . . 0.1 0.1 0.2 -- 0.4 0.I Any traditional method 1.0 6.5 8.7 10.0 13,9 10.3 6.7 8.9 Periodic abstinence 0.9 2.5 5.3 5.2 8.4 6.5 3.8 5.0 Withdrawal 0.5 3.1 3.7 5.4 6.5 2.8 1.9 3.8 Other -- 2.5 1.4 1.8 1.7 2.8 2.1 1.9 Number of wom~n 418 1041 1452 1147 931 803 572 6364 - L~s than 0.05 I~rceat 58 The most commonly used modem methods were the condom (7 percent), followed by the pill (5 percent) and female sterilisation (4 percent). Less than one percent of the women reported ever having used a vaginal method such as the diaphragm, spermicides, or suppositories and a negligible proportion (0.1 percent) reported the use of male sterilisation. In terms of age, one-fifth or more of ever-married and of currently married women age 30-44 had used a modem method. Contraceptive use rates were highest in the age group 35-39, where knowledge was reported to be the highest (Table 5.3). A low use rate was observed among the youngest women: only 2 percent of those age 15-19 had ever used a modem method. The use of traditional methods was highest (13- 14 percent) in the age group 35-39, followed by 10 percent in the adjacent age groups 30-34 and 40-44. Levels of ever use from the PDHS are considerably higher (21 percen0 than those estimated from previous surveys. Data from the 1968-69 National Impact Survey (NIS), the 1975 Pakistan Fertility Survey (PFS), the 1979-80 Population, Labour Force and Migration Survey (PLM) and the 1984-85 Pakistan Contraceptive Prevalence Survey (PCPS) show levels of ever use among currently married women to be 12 percent, 11 percent, 5 percent, and 12 percent, respectively (Soomro et al. [1984]; Population Welfare Division 1986). While differences in the methods asked about andintheimplementationofthe surveysmay account for some of the differences, it seems likely that there has been a substantial recent increase in the ever use of contraception in the country. Current Use of Family Planning Methods Regarding current use at the time of survey, ~ 12 percent of currently married women reported that they were using some method to delay or prevent pregnancy (see Table 5.5). Three-fourths of the current users were using a modem method and one-fourth a traditional method. The most widely used method was female sterilisation (4 percent), followed by the condom (3 percent) and the IUD (1 percent). Less than one percent were using either the pill or injection (a recently introduced method). 1In the PDHS, no reference period was def'med for current use. The woman was asked whether she or her husband were cune~tly using a method. 59 Table 5.5 Currant use of contraception Percent distribution of curranfly married women by current use of contraceptive methods, according to age, Pakistan 1990-91 Age of woman Con l~,m~l~l~Jv@ method 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Total Any method 2.6 6.3 9.6 13.4 20.4 15.8 I 1.8 11.8 Any modern method 1.9 3.8 7.4 9.6 15.8 12,8 10.3 9,0 Pill 0.2 0.8 0.8 0.7 0.9 0.8 -- 0,7 IUD 0.4 0.7 1.8 1.9 1.4 1.1 0.4 1.3 Injection 0.4 0.4 0.4 0.6 1,6 1.1 1.1 0.8 Vaginal method . . . . . . . . 0.1 . . . . . . Condom 0.8 1.5 3.6 3.6 3.8 1.8 1.9 2.7 Female sterilisation -- 0.5 0.9 2.7 7.9 8.0 6.8 3.5 Male sterilisation . . . . . . 0.1 0.2 . . . . . . Any traditional method 0.7 2.5 2.3 3.8 4.5 3.0 1.5 2,8 Periodic abstinence 0.5 0.7 1.0 1.6 2.6 1.7 0.4 1.3 Withdrawal 0.1 1.1 1.0 1.9 1.8 1.3 0.3 1.2 Other -- 0.6 0.3 0.3 0.2 -- 0.8 0.3 Not currently using 97.4 93.7 90.4 86.6 79.6 84.2 88.2 88.2 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Number 418 1041 1452 1147 931 803 572 6364 -- Less than 0.05 percent The level of contraceptive use varies with the age of women, increasing from less than three percent for married women age 15-19 to a high of 20 percent for women 35-39, and decreasing thereafter. It is assumed that younger women are less likely to use contraception because they have recently started their families, while older women, whose current use is more than the national average, are more likely to use due to the pressure of frequent pregnancies and large family size. Younger women typically use a temporary method such as the condom, periodic abstinence, or withdrawal, while sterilisation is more common among older women (8 percent of women age 35-44 have been sterilised). It is notable that even among women age 35 and over, the condom is the second most widely used method. Table 5.6 shows the contraceptive use rates (ever use and current use) among non-pregnant women. A total of 22 percent o

View the publication