Liberia - Demographic and Health Survey - 1988

Publication date: 1988

Liberia Demographic and Health Survey 1986 Bureau of Statistics Ministry of Planning and Economic Affairs ®DHS Demographic and Health Surveys Institute for Resource Development/Westinghouse REPUBLIC OF LIBERIA LIBERIA DEMOGRAPHIC AND HEALTH SURVEY 1986 by Dorothy Chieh-Johnson Anne R. Cross Ann A. Way Jeremiah M. Sullivan Bureau of Statistics Ministry of Planning and Economic Affairs Monrovia, Liberia and Institute .for Resource Development/Westinghouse Columbia, Maryland, U.S.A. February 1988 This report presents the findings of the Liberia Demographic and Health Survey, implemented by the Ministry of Planning and Economic Affairs in 1986. The survey is part of the worldwide Demographic and Health Surveys (DHS) Programme which is designed to collect data on fertility, family planning, and maternal and child health. Additional information on this survey can be ob- tained from the Ministry of Planning and Economic Affairs, P.O. Box 9016, Monrovia, Liberia. The Liberia Demographic and Health Survey was carried out with the assistance of the In- stitute for Resource Development (IRD), a subsidiary of Westinghouse Electric Corporation, with offices in Columbia, Maryland. Funding for the survey was provided under a contract with the U.S. Agency for International Development (Contract No. DPE--3023-C~30-4083-00). Addition- al information about the DHS program can be obtained by writing to: DHS, IRD/Westinghouse, P.O. Box 866, Columbia, MD, 21044, U.S.A. (Telex 87775). CONTENTS Page CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi MAP OF LIBERIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1. BACKGROUND 1.1 History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2 Geography and Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.4 Educational System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.5 Religion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.6 Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.7 Population and Family Planning Policies and Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.8 Health Priorities and Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 1.9 Objectives of the Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.10 Organization of the Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 1.11 Background Characteristics of the Surveyed Women . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2. MARRIAGE AND EXPOSURE TO THE RISK OF PREGNANCY 2.1 Current Marital Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.2 Polygyny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3 Age at First Union . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.4 Exposure to the Risk of Pregnancy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 2.5 Breastfeeding, Postpartum Amenorrhea and Abstinence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3. FERTILITY 3.1 Fertility Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.2 Levels and Differentials in Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.3 Fertility Trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.4 Children Ever Born . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.5 Age at First Birth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 iii Page 4. CONTRACEPTIVE KNOWLEDGE AND USE 4.1 Contraceptive Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.2 EverUse of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.3 Current Use of Contraception . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 4.4 Source for Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.5 Attitude about Pregnancy and Reasons for Non-use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 4.6 Intention To Use In the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.7 Approval of the Use of Family Planning Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 5. FERTILITY PREFERENCES 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 5.2 Future Fertility Preferences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 5.3 Ideal Number of Children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 5.4 Unwanted Fertility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 6. MORTALITY AND HEALTH 6.1 Mortality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6.2 Maternity Care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 6.3 Child Health Indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 APPENDIX A: SURVEY DESIGN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sample Design and Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characteristics of the Sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . APPENDIX B: ESTIMATES OF SAMPLING ERROR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . APPENDIX C: SURVEY QUESTIONNAIRES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 83 85 87 95 117 iv TABLES Page 1.1 PERCENT DISTRIBUTION OF WOMEN AGED 15--49 BY BACKGROUND CHARACTERISTICS, LIBERIA, 1974, 1984, AND 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.2 PERCENT DISTRIBUTION OF WOMEN BY EDUCATION, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 2.1 PERCENT DISTRIBUTION OF WOMEN BY CURRENT MARITAL STATUS, ACCORDING TO CURRENT AGE, LIBERIA 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.2 PERCENTAGE OF WOMEN WHO HAVE NEVER MARRIED BY AGE GROUP, LIBERIA, 1974, 1984, AND 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.3 PERCENTAGE OF CURRENTLY MARRIED WOMEN WHO ARE IN A POLYGYNOUS UNION, BY AGE, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.4 PERCENT DISTRIBUTION OF WOMEN BY AGE AT FIRST UNION AND MEDIAN AGE AT FIRST UNION, ACCORDING TO CURRENT AGE, LIBERIA 1986 . . . . . . . . . . . . . . . . . . . 19 2.5 MEDIAN AGE AT FIRST UNION AMONG WOMEN AGED 20-49 BY CURRENT AGE, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA 1986 . . . . . . . . . . . . . . . . 21 2.6 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN BY EXPOSURE STATUS, ACCORDING TO AGE, LIBERIA 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.7 PERCENTAGE OF WOMEN WHO ARE CURRENTLY EXPOSED TO THE RISK OF PREGNANCY, BY MARITAL STATUS, ACCORDING TO AGE, LIBERIA 1986 . . . . . . . . . . . 23 2.8 PROPORTIONS OF WOMEN STILL BREASTFEEDING, POSTPARTUM AMENORRHEIC, AND ABSTAINING, BY MONTHS SINCE BIRTH, AND MEDIAN AND MEAN DURATIONS OF BREASTI~tzEDING, AMENORRHEA AND ABSTINENCE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 2.9 MEAN DURATION OF POSTPARTUM BREASTI~EEDING, AMENORRI-IEA, AND SEXUAL ABSTINENCE (CURRENT STATUS ESTIMATE BASED ON BIRTHS WITHIN 36 MONTHS OF THE INTERVIEW DATE), ACCORDING TO BACKGROUND CHARACTERISTICS OF MOTHER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 3.1 TOTAL FERTILITY RATES FOR 198346, 1980-82, AND THE FIVE-YEAR PERIOD IMMEDIATELY PRIOR TO THE SURVEY AND MEAN NUMBER OF CHILDREN EVER BORN TO WOMEN 40-49, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 3.2 AGE-SPECIFIC FERTILITY RATES (PER THOUSAND WOMEN) FOR FIVE-YEAR PERIODS PRIOR TO THE SURVEY, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.3 PERCENT DISTRIBUTION OF ALL WOMEN AND CURRENTLY MARRIED WOMEN BY NUMBER OF CHILDREN EVER BORN, ACCORDING TO AGE, LIBERIA, 1986 . . . . . . . . . 33 V Page 3,4 MEAN NUMBER OF CHILDREN EVER BORN TO EVER-MARRIED WOMEN AGED 15--49, ACCORDING TO AGE AT FIRST MARRIAGE AND YEARS SINCE FIRST MARRIAGE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 3.5 PERCENT DISTRIBUTION OF ALL WOMEN BY AGE AT FIRST BIRTH, ACCORDING TO CURRENT AGE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 3.6 MEDIAN AGE AT FIRST BIRTH AMONG WOMEN AGED 20-49, ACCORDING TO CURRENT AGE AND BACKGROUND CHARACTERISTICS OF WOMEN, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4.1 PERCENTAGE KNOWING ANY CONTRACEPTIVE METHOD AND KNOWING A SOURCE (FOR INFORMATION OR SERVICES) FOR A METHOD AMONG ALL WOMEN AND CURRENTLY MARRIED WOMEN BY METHOD, LIBERIA, 1986 . . . . . . . . . . . . . . . . . 38 4.2 PERCENTAGE KNOWING ANY MODERN CONTRACEPTIVE METHOD AND PERCENTAGE KNOWING A SOURCE FOR MODERN METHODS AMONG CURRENTLY MARRIED WOMEN ACCORDING TO BACKGROUND CHARAC'I~RISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 4.3 PERCENTAGE WHO HAVE EVER USED CONTRACEPTIVE METHODS AMONG ALL WOMEN AND CURRENTLY MARRIED WOMEN BY METHOD, LIBERIA, 1986 . . . . . . . 40 4.4 PERCENT DISTRIBUTION OF ALL WOMEN AND WOMEN WHO HAVE EVER USED PERIODIC ABSTINENCE BY KNOWLEDGE OF THE FERTILE PERIOD DURING THE OVULATORY CYCLE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 4.5 PERCENT DISTRIBUTION OF ALL WOMEN AND CURRENTLY MARRIED WOMEN, ACCORDING TO CONTRACEPTIVE METHODS CURRENTLY USED, LIBERIA, 1986 . . . 41 4.6 PERCENTAGE CURRENTLY USING ANY CONTRACEPTIVE METHOD AND CURRENTLY USING ANY MODERN CONTRACEPTIVE METHOD AMONG CURRENTLY MARRIED WOMEN BY BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 4.7 PERCENT DISTRIBUTION OF CURRENT CONTRACEPTIVE USERS B Y THE MOST RECENT SOURCE OF SUPPLY (INFORMATION) ACCORDING TO THE METHOD CURRENTLY USED, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.8 PERCENT DISTRIBUTION OF NON-USERS EXPOSED TO RISK OF PREGNANCY BY ATTITUDE TOWARD BECOMING PREGNANT IN THE NEXT FEW WEEKS, ACCORDING TO THE NUMBER OF LIVING CHILDREN, LIBERIA, 1986 . . . . . . . . . . . . . . 47 4.9 PERCENT DISTRIBUTION OF NON-USERS EXPOSED TO RISK OF PREGNANCY WHO WOULD BE UPSET IF THEY BECAME PREGNANT, BY MAIN REASON FOR NON-USE, ACCORDING TO AGE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.10 PERCENT DISTRIBUTION OF CURRENTLY MARRIED NON-USERS BY INTENTION TO USE IN THE FUTURE, ACCORDING TO NUMBER OF LIVING CHILDREN, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 4.11 PERCENT DISTRIBUTION OF CURRENTLY MARRIED NON-USERS INTENDING TO USE IN THE FUTURE BY PREFERRED METHOD, ACCORDING TO CURRENT AGE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 vi 4.12 4.13 4.14 5.1 5.2 5.3 5.4 5.5 5.6 5.7 6.1 6.2 6.3 6.4 Page PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN KNOWING A CONTRACEPTIVE METHOD BY HUSBAND'S AND WIFE'S .t/ITfI'UDE TOWARD THE USE OF FAMILY PLANNING, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN KNOWING A METHOD BY NUMBER OF TIMES DISCUSSED FAMILY PLANNING WITH HUSBAND, ACCORDING TO CURRENT AGE, L1BERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 PERCENTAGE OF CURRENTLY MARRIED WOMEN KNOWING A METHOD WHO APPROVE OF FAMILY PLANNING AND WHO SAY HUSBAND APPROVES OF FAMILY PLANNING, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . 52 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN AGED 15-49 BY FERTILITY PREFERENCES, ACCORDING TO NUMBER OF LIVING CHILDREN, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN AGED 15-49 BY FERTILITY PREFERENCES, ACCORDING TO AGE GROUP, LIBERIA, 1986 . . . . . . . . . . . 55 PERCENTAGE OF CURRENTLY MARRIED WOMEN AGED 15--49 WHO WANT NO MORE CHILDREN, ACCORDING TO NUMBER OF LIVING CHILDREN AND BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 AMONG CURRENTLY MARRIED WOMEN AGED 15-49, THE PERCENTAGE WHO ARE IN NEED OF FAMILY PLANNING AND THE PERCENTAGE WHO ARE IN NEED AND INTEND TO USE FAMILY PLANNING IN THE FUTURE, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 PERCENT DISTRIBUTION OF WOMEN AGED 15-49 BY IDEAL NUMBER OF CHILDREN, ACCORDING TO ACTUAL NUMBER OF LIVING CHILDREN, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 MEAN IDEAL NUMBER OF CHILDREN FOR WOMEN AGED 15-49 ACCORDING TO AGE AND BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . 61 PERCENT DISTRIBUTION OF WOMEN AGED 15-49 WHO HAD A BIRTH IN THE 12 MONTHS PRIOR TO THE SURVEY BY WHETHER THEY WANTED THE CHILD THEN, LATER, OR NOT AT ALL, ACCORDING TO BIRTH ORDER, LIBERIA, 1986 . . . . . . . . . . . . 62 INFANT AND CHLIDHOOD MORTALITY ESTIMATES BY TIME PERIOD, SEX AND URBAN-RURAL RESIDENCE, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64 SOCIOECONOMIC DIFFERENTIALS IN INFANT AND CHILD MORTALITY, 1976-86, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 DEMOGRAPHIC DIFFERENTIALS IN INFANT AND CHILD MORTALITY, 1976-86, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 MEAN NUMBER OF CHILDREN EVER BORN, SURVIVING AND DEAD AND PROPORTION OF CHILDREN DEAD BY AGE OF WOMEN, LIBERIA, 1986 . . . . . . . . . . . . 70 vii Page 6.5 PERCENT DISTRIBUTION OF WOMEN AGED 15-49 WHO HAVE GIVEN BIRTH IN THE FIVE YEARS PRECEDING THE SURVEY BY TYPE OF PRENATAL CARE PRIOR TO THE MOST RECENT BIRTH, AND PERCENTAGE OF SUCH WOMEN WHO RECEIVED A TETANUS TOXOID INJECTION PRIOR TO MOST RECENT BIRTH, ACCORDING TO BACKGROUND CHARACTERISTICS OF MOTHER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . 72 6.6 PERCENT DISTRIBUTION OF WOMEN AGED 15-49 WHO HAVE GIVEN BIRTH IN THE FIVE YEARS PRECEDING THE SURVEY BY TYPE OF ASSISTANCE AT DELIVERY OF THE MOST RECENT BIRTH, ACCORDING TO BACKGROUND CHARACTERISTICS OF MOTHER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 6.7 AMONG ALL CHILDREN UNDER 5 YEARS OF AGE, THE PERCENTAGE WITH HEALTH CARDS AND THE PERCENTAGE RECORDED ON HEALTH CARD OR REPORTED BY MOTHER AS HAVING BEEN IMMUNIZED AND, AMONG CHILDREN WITH HEALTH CARDS, THE PERCENTAGE FOR WHOM BCG, DPT, POLIO, AND MEASLES ARE RECORDED AND THE PERCENTAGE CONSIDERED TO BE FULLY IMMUNIZED, ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTHER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 6.8 AMONG ALL CHILDREN UNDER 5 YEARS OF AGE, THE PERCENTAGE REPORTED BY THE MOTHER AS HAVING HAD DIARRHEA IN THE FOUR WEEKS PRECEDING THE SURVEY AND, AMONG CHILDREN WITH DIARRHEA, THE PERCENTAGE RECEIVING VARIOUS TREATMENTS, ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTFIER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . 77 6.9 AMONG ALL CHILDREN UNDER 5 YEARS, THE PERCENTAGE REPORTED BY THE MOTHER AS HAVING FEVER 1N THE FOUR WEEKS PRECEDING THE SURVEY AND, AMONG CHILDREN WITH FEVER, THE PERCENTAGE RECEIVING VARIOUS TREATMENTS, ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTHER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 6.10 AMONG ALL CHILDREN UNDER 5 YEARS, THE PERCENTAGE REPORTED BY THE MOTHER AS SUI~I~t~RING FROM COUGHING OR DIFFICULT BREATHING IN THE FOUR WEEKS PRECEDING THE SURVEY AND, AMONG CHILDREN WITH A COUGH OR DIFFICULT BREATHING, THE PERCENTAGE RECEIVING VARIOUS TREATMENTS ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTHER, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 6.11 AMONG ALL CHILDREN UNDER 5 YEARS OF AGE, THE PERCENTAGE REPORTED BY THE MOTHER AS EVER HAVING MEASLES, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 APPENDIX A A.1 DATA ON SAMPLE IMPLEMENTATION AND RESPONSE RATES, LIBERIA, 1986 . . . . . . . 86 APPENDIX B B.1 LIST OF SELECTED VARIABLES FOR WHICH SAMPLING ERRORS ARE CALCULATED, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 B.2 MEANS, STANDARDS ERRORS, AND CONFIDENCE INTERVALS FOR SELECTED VARIABLES BY DOMAIN, LIBERIA, 1986 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 viii FIGURES Page 4.1 FAMILY PLANNING KNOWLEDGE AND USE, CURRENTLY MARRIED WOMEN 15-49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 4.2 CURRENT USE OF FAMILY PLANNING BY METHOD, CURRENTLY MARRIED WOMEN AGE 15-49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.3 CURRENT USE OF FAMILY PLANNING BY EDUCATION AND RESIDENCE, CURRENTLY MARRIED WOMEN 15-49 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 4.4 SOURCE OF FAMILY PLANNING SUPPLY, CURRENT USERS . . . . . . . . . . . . . . . . . . . . . . . . 46 5.1 FERTILITY PREFERENCES, CURRENTLY MARRIED WOMEN AGE 15--49 . . . . . . . . . . . . . 54 5.2 FERTILITY PREFERENCES BY NUMBER OF LIVING CHILDREN . . . . . . . . . . . . . . . . . . . . 55 6.1 TRENDS IN INFANT AND CHILD MORTALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 6.2 DIFFERENTIALS IN INFANT MORTALITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 ix PREFACE The Liberia Demographic and Health Survey (LDHS) was conducted as part of the worldwide Demographic and Health Surveys (DHS) program, in which surveys are being carried out in countries in Africa, Asia, Latin America, and the Middle East. Liberia was the second country to conduct a DHS and the first country in Africa to do so. The LDHS was a national sample survey, designed to collect information on fertility, family planning, mortality and health. Fieldwork for the Liberia Demographic and Health Survey was conducted from February to July 1986 under terms of an agreement between the Government of Liberia, through the Ministry of Planning and Economic Affairs (MPEA), and the Institute for Resource Development (IRD), a subsidiary of Westinghouse Electric Corporation. Financial assistance was provided by the U.S. Agency for International Development (USAID). Technical and administrative services for the survey were jointly provided by the Ministry of Planning and Economic Affairs and IRD. Implementation of the DHS survey in Liberia came at an opportune time as the Liberian Na- tional Population Commission (NPC) was in the process of formulating a national population policy for Govemment's approval. The survey will also be useful in providing baseline informa- tion to the Southeast Region Primary Health Care Project (SER/PHC) for planning its program to motivate and educate the people in Sinoe and Grand Gedeh Counties in preventive health measures. The LDHS is set to attain additional objectives, some of which are: • to collect data on the family planning knowledge and behavior of women; to ascertain the reasons for high or low fertility among women and the use or non-use of family planning methods; • to collect data on some health-related matters such as immunizations, breastfeeding and prenatal check-ups; • to obtain experience in conducting surveys to monitor changes in birth rates, health and the use of family planning; • to provide Liberian data to the DHS database for international comparative research. The planning of the LDHS commenced in 1985 at which time a statistical subcommittee was set up in the Bureau of Statistics, MPEA. The subcommittee consisted of representatives of the Ministry of Health and Social Welfare, the Family Planning Association of Liberia, and the Bureau of Statistics, MPEA. The role of the subcommittee was to further develop and adapt the model questionnaire of the DHS to reflect Liberia's situation. Members of the subcommittee from the Bureau of Statistics were instrumental in providing the 1984 census frame and other car- tographic documents useful at the listing and enumeration stages. The success of this large undertaking could not have been realized without the relentless effort and dedication of several institutions and individuals, especially the employees of the Ministry of Planning and Economic Affairs and IRD/Westinghouse, respectively. In particular, I wish to ex- tend my gratitude to the following individuals and institutions who contributed to the success of the LDHS Project. xi Administrative: Hon. Emanuel O. Gardiner, Hon. Paul R. Jeffy, Hon. J. Rudolph Johnson (former Ministers of MPEA); Hon. Amelia Ward, Deputy Minister for Economic Planning and Statistics; Hon. T. Edward Liberty, Assistant Minister for Statistics, Bureau of Statistics; Hon. Abraham Y. Turay, Assistant Minister for Administration, and tlis staff. Technical Committee: Hon. T. Edward Liberty, Chairman/Assistant Minister for Statistics; Mr. Philip Gadegbeku, LDHS Project Director/Co-Chairman; Mr. A. Massalee, Director of Population Division/Member; Mrs. Dorothy Chieh-Johnson, LDHS Project Coordinator/Member; Mr. S.N. Goswami, UN Advisor to MPEA/Member; Hon. J. Prall, Assistant Minister and Registrar General, MOH&SW/Member; Dr. McArthur Wolo and Mr. R. Ainsworth of PHC, MOH&SW/Members. Technical: Mrs. Dorothy Chieh-Johnson, LDHS Project Coordinator, MPEA; Ms. Anne R. Cross, Dr. Ann Way and Dr. Jeremiah SuLlivan, DHS Coordinators, IRD/Westinghouse; Dr. Alfredo Aliaga, Sampler, IRD/Westinghouse and Mr. Lawrence Akoi, Assistant Sampler, MPEA. Field Staff: Ms. Sandra Howard, Field Coordinator, University of Liberia; Ms. Viola Wesley, Ms. Faith Lawrence, and Messrs. Alfred Jaryan, Theo Barlay, MuUy Sandi and Whoniyan Bryant, Field Supervisors, and their Field Editors and Enumerators. Data Processing Staff: Ms. Jeanne Cushing and Ms. Elizabeth Britton, IRD/Westinghouse; Mr. R. Togba, Ms. Rozana Soko and Ms. Hawa Sherif, Data Processing Supervisors, Bureau of Statistics, and their assistants. Institutions: Ministry of Health and Social Welfare; Family Planning Association of Liberia If:PAL); Ministry of Internal Affairs (MIA); and all concessions. Finally, I extend my sincere thanks to all the Superintendents of the various counties and to all those who contributed one way or another to the success of the Liberia Demographic and Health Survey. Elijah E. Taylor Minister, Ministry of Planning and Economic Affairs, R.L. xii MAP OF LIBERIA / LOFA COUNTY GUi~e 0 3 COUNTY ~ NIMBA "~,p, <.¢ % TERRITORY GRAND BASSA COUNTY ;OUNTY ITERR'~ L~ % GRAND GEDEH _ COUNTY SASSTOWN MARY ~LAND Scole: I: 3tOOOtO00 xiv SUMMARY The Liberia Demographic and Health Survey (LDHS) was a national-level survey conducted from February to July 1986, covering a sample of 5,239 women aged 15 to 49. The purpose of the survey was to provide planners and policymakers with data regarding fertility, family planning, and matemal and child health. A secondary objective was to collect baseline information for the Southeast Region Primary Health Care Project. Survey data indicate that fertility is high in Liberia, with women having an average of about six and one-half births by the time they reach the end of their childbearing years. This level of fer- tility is found both from data on the mean number of children ever born to older women, as well as from the recent total fertility rate. Furthermore, it appears that fertility has been more or less constant in the recent past, and may be increasing slightly. Women in urban areas and in Montser- rado County have somewhat fewer children than women in other areas. The most significant dif- ference in fertility rates is by educational level, where women with secondary or higher education average fewer than 5 births per woman, as compared with 7 births for women with primary educa- tion. Childbearing in Liberia begins at an early age; over half of Liberian women have their first birth before they reach age 20. One reason for the high fertility levels in Liberia is that marriage (whether legal, traditional, or consensual) is early and almost universal. Half of all Liberian women marry before they reach age 18 and less than two percent remain unmarried their entire life. There is evidence of a trend toward later marriage---the median age at union has risen from about 16 for older women, to over 18 for the 20-24 year olds. Urban women marry a year and a half later on average than their rural counterparts, while women with secondary education marry almost four years later than women with no education. Polygyny is common in Liberia, with almost 40 percent of currently married women reporting that their husbands have other wives. Traditional practices of breastfeeding and postpartum abstinence still provide substantial protection from pregnancy after the birth of a child. Babies are breastfed for an average of 17 months, which no doubt helps to extend the average duration of postpartum amenorrhea to 11 months. Sexual abstinence after a birth is also long in Liberia, with an average duration of 13 months. LDHS data show a trend toward the weakening of traditional breastfeeding and abstinence practices among younger, urban and more educated women. Another factor leading to high fertility is the low level of contraceptive use in Liberia. Al- though 72 percent of Liberian women know at least one contraceptive method, only 22 percent have ever used a method, and only 8 percent are currently using. Both knowledge and use of fami- ly planning is low among currently married women, with only 6 percent of married women cur- rently using a method. Low rates of use may be due in part to lack of awareness of sources for obtaining family planning services, since less than half of all women know of a place to get a method. Most users rely on modem methods. The pill is the predominant method; more than half of all current users rely on it. Female sterilization, periodic abstinence, and the IUD account for the majority of the rest of the users. Contraceptive use is higher among women with more children, urban women, women in Montserrado County, Christian women, and women from the Grebo, Kru/Sapo, and Lorma tribes. The most outstanding differential in family planning use, however, is by education level; the use rate among women with some secondary schooling is nearly ten times the rate of women who never went to school. With regard to sources of supply, 40 percent of cur- rent users obtain their methods through the Family Planning Association of Liberia, while 29 per- cent rely on government services and 23 percent on private sources. The most common reason for non-use cited by women who are exposed to the risk of becom- ing pregnant and do not want to get pregnant immediately is fear of side effects. Other reasons in- clude disapproval of family planning by either the parmer or the respondent herself, cost of methods, and difficulty in obtaining them. Approval of family planning is not widespread in Liberia--slighdy less than half of currently married women knowing about family planning ap- prove of its use by couples and only 30 percent feel that their husbands approve. Despite the low level of contraceptive use, LDHS data indicate that the potential need for fami- ly planning is great. Seventeen percent of married women say that they want no more children and 33 percent want to wait at least two years before having their next child. This means that half of all married women are potentially in need of family planning either to limit or to space births. Furthermore, 30 percent of women who had a birth in the 12 months prior to the survey indicated that their last birth was either unwanted or mistimed. With regard to health issues, LDHS data indicate that out of every 1,000 births, 144 die before reaching their first birthday and 220 die before reaching age five. While these rates indicate high levels of mortality, the rates for earlier time periods are even higher, and there is a clear trend of declining childhood mortality over the past decade. As expected, childhood mortality is higher among boys than girls and among children of rural than urban mothers. The most significant dif- ferentials are those associated with the length of the preceding birth interval. The infant mortality rate for intervals of less than two years (203) is almost three times the rate for 4 years or more (72). About two out of every three Liberian children under five has received some immunization; however, only about one out of five children age one year and over with health cards is fully im- munized, i.e., has received BCG and measles vaccinations and completed three doses of DPT and polio. LDHS data indicate high levels of prevalence of certain childhood diseases. Of children under five, 40 percent had diarrhea in the four weeks before the survey, 50 percent had fever, and 37 per- cent had respiratory difficulties. Antibiotics and traditional medicine are the most common treat- ments for diarrhea, antimalarial medicine for fever, and cough symp for respiratory problems. Almost 20 percent of children under five have had measles. 2 1. BACKGROUND 1.1 History Liberia, one of the oldest republics on the continent of Africa, is unique because of its non- colonial background. Available information suggests that the spatial settlement of various ethnic groups as found today began sometime in the 14th century. It appears fairly certain that most of the tribes arrived between the twelfth and sixteenth centuries from the north, northeast, and the east; many came from the savanna areas near the Sahara Desert, fleeing from the Islamic Jihad (Moslem Holy War). It is further suggested that the migration of tribes into Liberia occurred in stages, following, among other factors, the decline of the Mali and Songhay Empires in 1375 and 1591, respectively. These migrations were probably encouraged by the favorable agricultural and economic conditions and the availability of land. The earliest information about the geographic area now known as Liberia came from European explorers and seafarers who frequented the west coast of Africa during the 15th century. Pedro de Sintra, a Portuguese, reached Liberia first in 1461 and again in 1462, when he started a Portuguese trade monopoly which lasted until 1515. Some of the principal items traded were ivory, gold and malegueta pepper. At the end of the 16th and the beginning of the 17th centuries, the Dutch came to Liberia in search of items traded by the Portuguese. Both the Dutch and the Portuguese gave names to most of the rivers, capes, and mountains found on the west coast, in- cluding Cape Mount (Cabo do Monte), Cestos River, Cape Palmas and CavaUa River. The Portuguese also taught the Vai and the Kru ethnic groups their languages. The English and French were also active on the west coast of Africa. The arrival of emancipated slaves from the United States of America marked the beginning of a new era in Liberia. In 1816, the American Colonization Society was founded by a group of American philanthropists, whose primary aim was to resettle the freed slaves in the land of their forefathers. In December, 1821, Mssrs. Ayres and Stockton of the Society bought the future site of Monrovia and a strip of coastland known as Providence Island from the chiefs of the De and Mambo tribes. These were used as settlements for freed men of color who began arriving in 1822. The colony did not have an easy start, as the chiefs and original inhabitants were reluctant to give up their lands, leading to several tribal attacks on the new colony. In addition, diseases and lack of technical and financial assistance contributed to the difficult struggle of the colony. In 1825, Liberia acquired its name and its first constitution. As years went by, many small set- tlements were founded and joined to form the Commonwealth of Liberia. However, the unwilling- ness of the British Government to accept the Commonwealth as a sovereign govemment during a dispute over levying import duties on British trading ships, led to Liberia's declaration of inde- pendence in 1847. Since its independence, Liberia has gained political, social and economic cohesion, facing the challenge of time. The descendants of the freed slaves, known as Americo-Liberians, and constituting less than five percent of the populace, govemed the country on a colonial pattem of indirect rule, thus trans- ferring the socioeconomic and political system of the United States to this land. After more than a century of settlers' oligarchy, Liberia saw a change in its political structure on April 12, 1980, when a coup d'etat ushered in the first indigenous leader. Since the inception of the indigenous government, otherwise known as the Second Republic, the constitution and some socio-political characteristics have remained patterned after the United States of America, with which Liberia has maintained a traditional informal as well as formal affiliation over the years. The indigenous people of Liberia comprise about 96 percent of the country's population and are distinguishable into sixteen major tribes. 1.2 Geography and Climate Located on the west coast of Africa between a longitude of 7" and 12 ° West and a latitude of 4" and 9* North, the Republic of Liberia covers an area of approximately 99,068 square kilometers. It is bordered by Sierra Leone on the west, Guinea in the north, and Ivory Coast in the east. The Atlantic Ocean provides a long coastline of 550 kilometers in the south (see map). The climate is humid tropical, with a long rainy season April to October and a dry season November to March. The average annual rainfall is 400 cm.; the coastal region receives over 500 cm., with rainfall diminishing inland. Because of their altitude, areas around Mt. Nimba and Voinjama in the extreme north receive more rain than the central part of Liberia. The average annual temperature is 28°C and whilst there is very little variation throughout the year, daily and seasonal temperature ranges do increase significantly inland from the coast. The humidity is generally extremely high, especially in the coastal area where it ranges from 90 to 100 percent. Between December and early February, the Harmattan, a dry wind, sweeps across Liberia, bringing dust and causing high temperatures during the day and low temperatures at night. Due to the climatic conditions, Liberia has a tropical rain forest vegetation. Exceptions are the extreme north and northwest with its savanna woodlands and the coastal belt where a number of different types of vegetation are found. Liberia's forest vegetation is characterized by a predominance of leguminous trees and a small volume of timber. 1.3 Economy Liberia's economy is considerably influenced by the importation of raw materials, equipment, and a wide variety of consumer goods. Production for the export sector is carried out on a large scale through foreign investment. The most important activity is the mining and shipment of iron ore; but, due to declining world demand, its share of the export market has decreased considerab- ly. Other exports include rubber, timber, diamonds, and increasingly, agricultural commodities. The Government is the largest single employer in the country. The manufacturing industry is small and is mainly geared to supplying goods for the domestic market. Construction activity is also limited and is mainly determined by the investment in the concession sector of the economy. About 70 percent of the Liberian population is engaged in traditional agriculture, growing rice, cocoa, coffee, and other cash crops. Realizing fairly low yields and moderate incomes, this sector has little influence on the economy as a whole. A period of high economic g;'owth was experienced from 1964 to 1974, when the average an- nual Gross Domestic Product (GDP) growth rate was 5.7 percent. This was mainly attributed to expanded export of iron ore and rubber and the exploitation of the rain forest resources. However, between 1975 and 1980, the growth rate of the GDP decreased to 1.7 percent as a result of the 4 sluggish demand for Liberia's exports and consequent low prices paid for its products on the world market. 1.4 Educational System For quite some time, Liberia has subscribed to the principle of universal education. As early as 1839, a public school law and its 1912 revision promulgated compulsory education in Liberia. Furthermore, the National Socioeconomic Development Plan (1976-1980) made "universal basic education" an explicit development objective. Although the educational system has expanded rapidly over the last three decades in response to this national commitment, the fulfillment of universal education even at the primary level is yet to be attained. The educational system in Liberia is of two types, formal and informal. The informal consists of the "bush schools" for boys and girls, while the formal comprises three levels: elementary (grades 1-6), secondary (grades 7-12), and higher education (degree programs). The Ministry of Education is the government's arm responsible for administering primary and secondary schools, including those that provide vocational and technical instruction. It also organizes primary teacher training and supervises both public and private schools. The University of Liberia and Cuttington University College are the only two universities in the country. Generally, school attendance rates are higher in urban areas than in rural areas. In rural, areas, particularly among small farmers who constitute the majority of the population, literacy is not a re- quirement for daily life. As a result, "Western" type education is not usually adopted. Many small farmers strongly believe that "Western" education will alienate their children from traditional beliefs and values and be a disruptive influence within the family; hence, they are not enthusiastic about sending their children to school. This belief probably accounts for the prevailing low level of literacy, particularly among women. Based on data from the 1984 Census, only 34 percent of the men and 17 percent of the women aged 10 years and over were able to read and write English. This, however, is a slight im- provement over 1974 Census figures which showed that only 30 percent of men and 12 percent of women were literate. Moreover, there was an improvement in the attendance rates of the school-aged population from 1974 to 1984. In 1974, about 26 percent of the school-aged population were enrolled in school--35 percent of males and 17 percent of females---whereas by 1984, the rates had almost doubled, to about 46 percent of the school-aged population attending school. The differential by sex also narrowed, with 57 percent of males and 34 percent of females attending school. 1.5 Religion Liberia is predominantly a Christian nation. Based on data from the 1984 census, about 68 per- cent of the population are Christian, 14 percent are Muslim, while the remaining 18 percent belong to the category "Other or No Religion." The distribution by ethnic affiliation shows that the Kpelle, Bassa, Grebo, Kru, and Gio tribes are predominantly Christian, while the Mandingo, Vai and Gola ethnic groups are predominantly Muslim. 1.6 Population Size and Structure Liberia's population has more than doubled during the past three decades, from an estimated 824,000 persons in 1950 to 2.1 million persons according to the 1984 census. Between 1962 and 1974, the population grew by 47.9 percent, whereas between 1974 and 1984 the change was 39.8 percent (MPEA, 1986: Table 1). The rate of intercensal population growth has remained the same with an average annual growth rate of 3.3 percent per annum from 1962 to 1974 compared to 3.4 percent from 1974 to 1984. The population of Liberia is characterized by a young age distribution which is the result of high fertility and declining mortality in recent years. In 1974, the population under 15 years of age was 615,000, or 41 percent of the total population (MPEA, 1977: Table 3). By 1984, this number had increased to 907,000, or 43 percent of the total population (MPEA, 1987). The momentum generated by high fertility and declining mortality also has an impact on the number of women of reproductive age; this group has increased from 377,000 women in 1974 to 497,000 in 1984. Such a large increase in the childbearing population implies that even if the num- ber of births per woman dropped rapidly, population growth would remain high. Mortality Mortality levels in Liberia have been declining. In 1974, the crude death rate was estimated at 17 per thousand population per year, with the rate for males slightly higher than that for females. The infant mortality rate for both sexes in the same year was estimated at 141 per thousand live births (University of Liberia and MPEA, 1981). Recent estimates from the 1984 Census indicate that the infant mortality rate has declined to 127 per thousand births (Republic of Liberia, 1987). Similarly, the estimated expectation of life at birth for both sexes has increased from 49 years in 1974 to 53 years in 1979-1984 (University of Liberia and MPEA, 1981: 89; Republic of Liberia, 1987: 3). Fertility Data from the 1970-71 Liberia Population Growth Survey (LPGS), the 1978 National Demographic Survey (NDS) and the 1974 and 1984 censuses have all shown that the level of fer- tility in Liberia is high. In 1974, the crude birth rate was estimated to be 49 per 1000 population per year and the total fertility rate (TFR) to be 6.7 children (University of Liberia and MPEA, 1981: 81, 84). Migration and Urbanization Although lifetime immigration into Liberia is small, it appears to be increasing. In 1984, about 95,000 or 4.6 percent of the total enumerated population were foreign-born. This was about 36,000 persons more than in 1974 when the foreign-bom comprised only four percent of the total population. In 1984, over 90 percent of the lifetime immigrants were from other African countries; three-fifths were from neighboring countries of Sierra Leone, Guinea, and Ivory Coast 6 (MPEA, 1987). Analysis of data from the LPGS shows that the major reasons for migrating were related to job opportunities. The number of persons living in urban areas increased from 29 percent in 1974 to ap- proximately 39 percent in 1984, almost a 10-point increase in I0 years. The rapid growth in Liberia's urban population is not only due to natural increase, but to a large extent, the high rate of rural-to-urban migration. Monrovia, the national capital, is the most urbanized area in Liberia. Its population grew from 46.6 percent of the total urban population in 1974 to 51.6 percent in 1984. 1.7 Population and Family Planning Policies and Programs Family planning activities were initiated in Liberia in 1956, with the establishment of the Family Planning Association of Liberia (FPAL). In 1972, based on a Presidential Proclamation, family planning was officially incorporated into Liberia's health programs. Since then, FPAL's role has been to assist the Ministry of Health and Social Welfare in implementing comprehensive, effective, and efficient family planning services to promote childspacing as a basic human right for the welfare of individuals and couples. Currently, family planning services are provided through clinics jointly administered by FPAL and the Ministry of Health and Social Welfare in nine of the thirteen counties in Liberia. Available data indicate that most acceptors use oral contraceptives, IUDs, and condoms, while a few use in- jectables. Natural family planning is gradually gaining support in Liberia. Although still at the develop- ment stage, a natural family planning program organized by the National Catholic Secretariat covers three counties in Liberia. Natural family planning does not involve the use of artificial methods but only natural rhythm method. The program does not have organized clinics, but provides services through health authorities in the various counties. 1.8 Health Priorities and Programs The Ministry of Health and Social Welfare is responsible for meeting the health and social wel- fare needs of the Liberian citizenry, by providing a viable health care delivery system which will permeate every urban and mral community of Liberia. In its continuing efforts to expand health services to a majority of the Liberian population, the Ministry has gradually shifted from the cost- ly curative, intensive programs of the 1970s to cost-effective, preventive-oriented primary health care programs. The health policy of the Govemment is to provide health care for all its people through a Na- tional Health Delivery System. This system is designed to provide, in a complementary manner, preventive and curative health services throughout the country. Particular emphasis is placed on matemal and child health services, environmental sanitation, immunization, and health education. The goal of the system is to extend health coverage for the population from 35 percent of the population to 90 percent by the year 2000, at an annual rate of 3 percent (MPEA, no date: 118, 119). Recently, the govemment initiated the Southeast Region Primary Health Care (SER/PHC) Project, a USAID-funded program focused on Sinoe and Grand Gedeh Counties. Among other things, the SER/PHC project aims at decreasing infant mortality, increasing immunization coverage of young children, educating mothers about oral rehydration therapy for diarrhea, in- creasing the contraceptive prevalence rate, and increasing the number of deliveries by trained health workers. Its immunization efforts are aimed at combatting the six major childhood dis- eases: measles, tetanus, polio, tuberculosis, whooping cough and diptheria. Health conditions in Liberia have been improving over the last twenty years. Life expectancy, for instance, has risen, while the supply of physicians and of hospital beds has improved in rela- tion to the size of population. Nevertheless, only about 35 percent of Liberia's population have ac- cess to any form of modem medical services. In 1980, the total number of health facilities included 58 hospital and health centers and 310 health posts and clinics (MPEA, no date: 115). 1.9 Objectives of the Survey The major objective of the LDHS was to provide data on fertility, family planning, and mater- nal and child health to planners and policymakers in Liberia for use in designing and evaluating programs. Although a fair amount of demographic data was available from censuses and surveys, almost no information existed concerning family planning, health, or the determinants of fertility, and the data that did exist were drawn from small-scale, subnational studies. Thus, there was a need for data to make informed policy choices for family planning and health projects. A more specific objective was to provide baseline data for the Southeast Region Primary Health Care Project. In order to effectively plan strategies and to eventually evaluate the progress of the project in meeting its goals, there was need for data to indicate the health situation in the two target counties prior to the implementation of the project. Many of the desired topics, such as immunizations, family planning use, and perinatal care, were already incorporated into the model DHS questionnaire; nevertheless, the LDHS was able to better aceomodate the needs of this project by adding several questions and by oversampling women living in Sinoe and Grand Gedeh Counties. Another important goal of the LDHS was to enhance tile skills of those participating in the project for conducting high-qual!ty surveys in the future. Finally, the contribution of Liberian data to an expanding international dataset was also an objective of the LDHS. 1.10 Organization of the Survey The LDHS was a national-level survey, with oversampling in Sinoe and Grand Gedeh Coun- ties so that separate estimates could be produced for the Southeast Region Primary Health Care Project. The 156 enumeration areas covered in the LDHS were selected with probability propor- tional to size, using the 1984 Population Census as a sampling frame. Field teams listed all households in each selected area, after which individual households were selected for interview. Because of oversampling, the sample is not serf-weighting at the national level, and the figures given in this report are based on weighted data. A more complete description of the sample design is given in Appendix A. The LDHS utilized two questionnaires: one to list members of the selected households (Household Questionnaire) and the other to record information from all women aged 15-49 who were present in the selected households the night before the interview (Individual Questionnaire). Both questionnaires were produced in Liberian English and were pretested in September 1985. The Individual Questionnaire was an early version of the DHS model questionnaire. It covered three main topics: (1) fertility, including a birth history and questions concerning desires for future childbearing, (2) family planning knowledge and use, and (3) family health, including prevalence of childhood diseases, immunizations for children under age five, and breasffeeding and weaning practices. Both questionnaires are reproduced in Appendix C. The field staff for the LDHS consisted of 24 female interviewers, 6 field editors, 6 super- visors, and one fieldwork coordinator. Except for the supervisors, who were experienced MPEA staff members, all field staff were specially recruited for the LDHS. Two training courses were held simultaneously in Monrovia, and Zwedru, Grand Gedeh, for the entire month of February, 1986. Training included practice interviewing, both in the classroom, as well as in the field. Since most of the interviews were conducted in the local dialects, the training course also covered prac- tice in asking questions in the vernacular. After training, six teams were formed, one for each of the major dialects covered in the sample. Data collection began in late February and was largely completed by July, except for some call-backs. Data from the questionnaires were entered onto microcomputers at the Bureau of Statistics of- fice in Monrovia. The data were then subjected to extensive checks for consistency and accuracy. Errors detected during this operation were resolved either by referring to the original question- naire, or, in some cases, by logical inference from other information given in the record. Finally, dates were imputed for the small number of cases where complete dates of important events were not given. Information on the completeness of date reporting is of interest in assessing data quality. With regard to dates of birth of individual women, 42 percent of respondents reported both a month and year of birth, 21 percent gave a year of birth in addition to current age, and 37 percent gave only their ages. With regard to children's dates of birth in the birth history, 85 percent of births had both month and year reported, 12 percent had year and age reported, 1 percent had only age reported, and 2 percent had no date information. 1.11 Background Characteristics of the Surveyed Women In the Liberia Demographic and Health Survey, 5,026 households were interviewed, repre- senting a response rate of 90 percent. Within the interviewed households, 5,340 eligible women were identified, of which 5,239 (98 percent) were interviewed (see Appendix A for details on response rates). Eligibility for the individual interview was based on de facto criteria, i.e., women who were between 1549 years of age and had stayed in the selected household the previous night. This section of the report briefly examines the demographic and other social characteristics of these women. Knowledge of these characteristics is not only useful on its own, but also provides a rough measure of the quality of the data and allows the reader to interpret other survey findings more easily. The distribution of LDHS respondents by background characteristics is given in Table 1.1, along with comparable data from the 1974 and 1984 censuses. The data show that the age distribu- tion of LDHS respondents is comparable to that of the country as a whole, with two minor excep- 9 TABLE 1.1 PERCENT DISTRIBUTION OF WOMEN AGED 15--49 BY BACKGROUND CHARACTERISTICS, LIBERIA, 1974, 1984, AND 1986 1974 1984 1986 Characteristic Census Census LDHS Aqe 15-19 22.1 23.8 21.7 20-24 17.8 20.8 19.7 25-29 17.7 17.5 20.6 30-34 15.3 12.6 12.6 35-39 12.2 I I .2 I I .9 40-44 8.2 7.7 6.2 45-49 6.7 6.4 7.3 Urban-Rural Urban 28.5 37.8 a 43.2 Rural 71.5 62.2 a 56.8 Education b None 89.5 79.4 62.6 Primary 6.7 8.9 18.4 Secondary or more 3.8 l l .7 19.0 Reliqion Christian NA 67.7 54.2 Muslim NA 14.2 14.4 Trad'I/Other NA 12.5 None NA 18.1 18.9 Trlbe c Bassa 14.7 14.4 12.7 Gio 9.0 8.1 7.7 Gola 4.4 4.0 4.6 Grebe 7.6 8.5 7.3 Kpelle 20.0 19.8 16.3 Krahn 4.9 3.8 4.2 Kru/Sapo d 7.7 7.1 10.6 Lorma 5.7 5.6 6.0 Mandingo 4.0 4.9 6.0 Mane 7.6 7.2 7,9 Other/None 14.4 16.6 16.8 All Women lO0.O lO0.O lO0,O NA = not ava i lab le a Refers to a l l women b For the two censuses, refers to women aged 15 and above c For the two censuses, refers to women aged 15-44 d For the two censuses, includes Kru only I0 tions. The proportions of LDHS respondents in both the 25-29 and 45~9 age groups are higher than in the 1974 and 1984 censuses and are higher than the proportions in the immediately preced- ing age groups. In a country with high fertility such as Liberia, one would expect the proportions to decrease gradually at each successive age group. These two anomalies may reflect a bias in age reporting in the LDHS. In any case, the discrepancies are minor and, on the whole, the age dis- tribution of LDHS respondents appears to be accurate. The data indicate that two out of every five women in Liberia are from urban areas. Com- parison with data from the 1974 Census show the large increase in urbanization that has taken place in the past 12 years. The fact that census data are based on a de jure definition of residence as compared to the de facto definition of residence used in the LDHS probably has an extremely minor effect. All women interviewed in the LDHS were asked if they had ever attended school. Those who responded positively were further asked the highest level of school attended according to the country's formal educational system. Respondents were grouped into three categories: those with no education, those who completed between 1-6 years of primary school (primary), and those who completed some secondary school or higher (secondary or more). Among the women surveyed, 63 percent have no education, 18 percent have had some primary education, and 19 percent have secondary or higher education. Unfortunately, census data for women of comparable age are not available, however, data for women aged 15 and above for 1974 and 1984 do show evidence of a trend toward higher education of women. All women interviewed in the survey were classified into four religious categories: Christian, Muslim, traditional/other, and none. The classification Christian includes all Christian denomina- tions. The LDHS findings show that over one-haft of respondents are Christian, about 14 percent are Muslim, 13 percent belong to other or traditional religions, while almost 20 percent report no religious affifiation. With the exception of the Muslim category, the 1986 figures do not agree very closely with those from the 1984 Census. It appears that many women who were reported as Christians in the 1984 Census were reported in either the "Traditional/other" or the "none" category in the LDHS. No question on religious affiliation was included in the 1974 Census. Respondents in the survey were grouped into eleven major ethnic groups, namely: Bassa, Gin, Gola, Grebo, Kpelle, Krahn, Kru/Sapo, Lorma, Mandingo, Mano and "other/none." Although there are sixteen major tribes in Liberia, some have been grouped into the "other/none" category due to their small size. The distribution of women by tribal affiliation is given in the bottom of Table 1.1. The data from all three sources show similar patterns, with the Kpelle and Bassa tribes being the largest. The LDHS shows a somewhat smaller proportion of Kpelle women than the two censuses . In later sections of this report, differentials in many variables will be presented by background characteristics of respondents. It is important to bear in mind that there are also interrelationships between the background characteristics themselves. Although not all these interrelationships can be explored in this report, Table 1.2 presents data conceming the relationship between educational attainment of women and other characteristics. 11 TABLE 1.2 PERCENT DISTRIBLrrlON OF WOMEN BY EDUCATION, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Characteristic Education Number No Second- Educ- ary or Total ation Primary More Percent Wtd. Unwtd. Aqe 15-19 20-24 25-29 30-34 35-39 40-44 45-49 UrbBn-Rural Urban Rural 36.7 41.I 22.2 I00.0 1,137 1,169 48.5 20.8 30.7 lO0.O 1,030 982 67.8 12.1 20.1 lO0.O 1,081 1,012 74.3 10.2 15.5 lO0.O 658 660 84.3 5.7 lO.O lO0.O 626 640 86.2 7.3 6.5 lO0.O 327 370 88.0 6.5 5.5 lO0.O 380 406 46.2 20.5 33.3 I00.0 2,262 1,944 75.2 16.8 8.1 I00.0 2,977 3,295 R@qiQn Sinoe 65.2 22.2 12.6 lO0.O 150 834 Grand Gedeh 69.1 23.5 7.4 lO0.O 293 920 Montserrado 44.1 20.0 35.9 lO0.O 1,459 1,060 Rest of country 70.I 17.1 12.8 lO0.O 3,337 2,425 Reliqion Christian Muslim Trad'I/Other None 50.I 21.I 28.8 I00.0 2,838 3,133 80.7 I0.7 8.6 lO0.O 754 602 69.4 19.7 I0.9 I00.0 658 602 80.4 15.8 3.9 I00.0 989 902 66.0 19.0 15.0 I00.0 664 490 64.9 20.6 14.6 I00.0 401 294 60.5 19.8 19.8 I00.0 244 177 39.4 23.0 37.6 I00.0 380 570 74.9 15.2 9.9 I00.0 854 661 54.7 26.7 18.6 I00.0 219 521 46.3 26.5 27.2 I00.0 555 1,039 59.2 19.2 21.6 I00.0 312 230 85.2 6.1 8.7 I00.0 317 262 64.7 23.3 12.0 I00.0 413 301 62.3 12.3 25.4 lO0.O 880 694 62.6 18.4 19.0 lO0.O 5,239 5,239 3,282 964 993 5,239 Tribe Bassa Gio Gol a Grebo Kpelle Krahn Kru/Sapo Lorma Mandingo Mano Other/None All Women Wtd. Number 12 Education is inversely related to age, that is, older women are generally less well-educated than younger women. For example, whereas only 37 percent of women 15-19 have had no formal education, over 85 percent of women aged 35 and over are uneducated. The proportion of respondents with no formal education is considerably higher in rural areas (75 percent) than in urban areas (46 percent). Libefian legislation provides compulsory schooling for all children up to 16 years of age; however, in the rural traditional setting, particularly among small farmers who constitute a majority of the rural populace, due to cultural belief and traditions, some families did not send their daughters to formal schools. This practice is gradually declining, as evidenced by the fact that younger women are better educated. Another factor influencing the urban-rural differential is that access to schools is more difficult in rural areas. Although there is almost no difference between urban and rural areas in the proportion of women with primary education, there is a dramatic difference between urban and rural women who have achieved secondary level or higher education. The urban proportion (33 percent) is four times that of the rural (8 percent). This phenomenon is not surprising since there am very few secondary schools or colleges in the rural parts of Liberia. Most often, rural girls who have com- pleted primary school would have to migrate to urban areas, particularly to Monrovia, or remain in the rural areas and get married. Table 1.2 shows that Montserrado County has the smallest proportion of uneducated women (44 percent) as compared to 65 percent in Sinoe, 69 percent in Grand Gedeh, and 70 percent in the rest of the country. There is very little difference among the four subregions with regard to the proportion of women with primary education. However, a greater disparity is evident when it comes to secondary level or higher education. Over one-third of women in Montserrado County fall in this category, as compared to Sinoe (13 percent), rest of the country (13 percent), and Grand Gedeh with the least (7 percent). As mentioned earlier, this is consistent with the fact that there are many more high schools and post-secondary schools in Montserrado County than in any other county. Table 1.2 indicates that one of every two Christian women is uneducated, compared to four of every five Muslim women and women with no religious affiliation. Furthermore, Christian women are more likely to have attended secondary or higher level institutions (29 percent) than women who are Muslim (9 percent), traditional/other (11 percent) or who have no religion (4 per- cent). Three-fifths or more of women in the tribal groups with the exception of the Grebo (39 per- cent), the Kru/Sapo (46 percent), and the Krahn (55 percent) have had no formal education. More than one-fifth of women in these tribes as well as women in Lorma and the "other~one" category have been to secondary school. The Mandingo and the Kpelle have the lowest proportion--nine and 10 percent, respectively--who have attended secondary school or higher, while the remaining tribes report between 12 and 20 percent. 13 2. MARRIAGE AND EXPOSURE TO THE RISK OF PREGNANCY 2.1 Current Marital Status In Liberia, as in most societies in the world, childbearing takes place mainly within socially prescribed and relatively stable marital unions. Thus, study of the patterns of marriage is essential to the understanding of fertility patterns in any society. In the Liberia Demographic and Health Survey, marriage was loosely defined to include any legal or customary union of a man and a woman as husband and wife, as well as other stable cohabitation, such as a man and a woman living together and having sexual relations without any legal or customary binding. Table 2.1 shows that, based on the above definition of marriage, one out of every five respon- dents in the LDHS bad never married, 67 percent were currently married, while the rest (11 per- cent) were either widowed, divorced, separated, or no longer living together. Table 2.1 also shows the variations in marital status by current age of the respondents. As ex- pected, the proportion of women who have never married decreases substantially with increasing age, from 64 percent of women !~5-19 to only about one percent of those over the age of 35. The proportion of women reported as living together is considerably higher in each age group when compared to those legally married except for the latter three age groups. The extent of widowhood is relatively small, particularly in the younger age groups, 15-19 through 30-34 years. Similarly, the proportion of women divorced is small and increases with age. TABLE 2.1 PERCENT DISTRIBUTION OF WOMEN BY CURRENT MARITAL STATUS, ACCORDING TO CURRENT AGE, LIBERIA, 1986 Marital Status No Longer Never Living Living Total Wtd. Age Married Married Together Widowed Divorced Together Percent Number 15-19 64.0 9.7 22.0 0.4 1.4 2.5 I00.0 1.137 20-24 24.7 20.2 45.3 0.I 2.8 7.I I00.0 1.030 25-29 7.9 34.2 45.1 0.6 3.2 9.0 lO0.O 1.081 30-34 6.2 38.3 43.5 1.7 3.5 6.8 I00.0 658 35-39 1.2 42.4 43.1 2.4 4.5 6.4 lO0.O 626 40-44 1.7 40.3 39.7 6.6 4.4 7.3 I00.0 327 45-49 0.5 51.0 30.8 6,2 5,7 5.8 I00,0 380 All Ages 21.4 29.2 38.3 1.6 3.1 6.3 I00.0 5,239 15 LDHS data on nuptiality can be compared with data from other sources in order to assess trends in marriage patterns and to evaluate the quality of data. Table 2.2 shows the proportions of women who have never married by age group, from the 1974 and 1984 censuses as well as from the LDHS. TABLE 2.2. PERCENTAGE OF WOMEN WHO HAVE NEVER MARRIED BY AGE GROUP, LIBERIA, 1974, 1984, AND 1986 1974 a 1984 b 1986 Age Census Census LDHS 15-19 57.7 64.3 64.0 20-24 21.4 29.1 24.7 25-29 9.1 13.9 7.9 30-34 5.0 7.3 6.2 35-39 3.6 4.5 1.2 40-44 3.0 3.6 117 45--49 2.8 3.1 0.5 All Ages 19.8 25.7 21.4 a From MPEA, 1977, Table 9 b From MPEA, 1987 The data show a substantial increase between 1974 and 1984 in the proportion of women at each age group who have never married, which implies that the age at marriage has been rising. This trend appears to reverse between 1984 and 1986, with a decrease in the proportions never married; however, at least some of the differences between 1984 and 1986 reflect definitional dif- ferences. Although the intent in both the censuses and the LDHS was to define marriage so as to include consensual unions, the census forms did not give the wording of questions and, since the codes given on the form for marital status were "never married," "married," "widowed," and "divorced/separated," is likely that many enumerators merely asked respondents if they were mar- ried, without explaining that consensual unions were to be included as marriages. In the LDHS, respondents were first asked if they had ever been married or lived with a man, and, if yes, whether they were currently married, living with a man, widowed, divorced, or no longer living together. Because the LDHS questions explicitly covered consensual unions, it is likely that they resulted in somewhat higher estimates of marriage than the censuses. 2.2 Polygyny In the LDHS, women who were either legally or traditionally married or living together with men were asked whether their husbands/partners had other wives. The analysis of their responses 16 TABLE 2.3 PERCENTAGE OF CURRENTLY MARRIED WOMEN WHO ARE IN POLYGYNOUS UNIONS, BY AGE, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Characteristic Aq~ 15-~9 20-24 25-29 30-34 35-39 40-44 45-49 All Ages Urban-Rural Urban Rural Reqion Sinoe Grand Gedeh Montserrade Rest of country Education No education Primary Secondary or more Reliqion Christian Muslim Trad'I/Other None Tribe Bassa 0io Gola Grebo Kpelle Krahn KrulSapo Lorma Mandingo Mano Other/None All Women 29.9 26.1 25.7 36.9 32.9 26.2 39.6 30.2 34.0 38.3 45.9 43.7 46.0 44.1 43,9 42.6 20.0 25.0 36.4 44.2 41.5 40.8 40.0 35.4 34.7 50.4 53.4 62.4 63.6 54.2 64.1 55.4 25.0 22.1 19.3 33.3 27.1 21.4 42,0 25.8 35.7 37.2 43.2 41.I 43.6 4].9 39.9 40,7 36,1 40,0 42.4 44.7 43.8 41.8 45.3 42.4 27.5 31.2 38.5 33.7 32.9 34.7 20.2 31.9 23.0 18.3 16.8 19.6 21.5 * 18,5 26.3 27.7 33.0 37.5 33.3 38.5 45.6 33.7 51.0 49.6 47.6 50.5 59.1 52.1 54.3 51.0 28.4 23.4 29,2 33.2 41.3 34.1 35.8 31.5 28.8 39.4 44.0 45.3 47.3 39.1 35.0 40.8 33.3 25.0 43.8 43.1 43.2 43.0 44.4 39.0 33.3 30.4 37.3 34.3 39.4 45,8 36.0 * 34.8 * 23.1 * 23.9 33.5 27,3 36.5 41.1 48.2 62.1 61.8 41.3 17.1 26.2 30,7 36.6 30.7 48.6 35.2 31.2 23.0 39.9 53,7 70.4 46,8 42.5 69.8 47.4 22.1 20.4 20.7 40.3 26.5 27.B 46.2 27.8 * 49.7 51.6 48.3 50.0 * 44.7 60.9 56.1 55.4 45.8 64.6 * * 57.4 34.8 38.8 22.0 28.5 50.0 50.0 30.4 34.2 40.5 38.5 41.2 42.8 48.6 34.2 42.2 41.6 32.4 33.7 37.7 40.8 41.2 39.9 42.6 38.0 *Fewer than 20 unweighted cases 17 reveals that 38 percent of these women are in polygynous unions (see Table 2.3). Unfortunately, there are no other sources of demographic information to which these data can be compared to detect any trend over time. However, the table shows that the prevalence of polygynous unions in Liberia increases slightly with the age of the women, which may indicate that the practice of polygny is gradually eroding. On the other hand, the data might merely reflect the fact that as women get older, their husbands are more likely to take second wives. Polygyny is more common in rural Liberia (43 percent of currently married women) than in the urban areas (30 percent). In rural Liberia, where people are predominantly farmers, the basic theories as to why polygyny is prevalent are that men want many wives, not only to enhance their political and economic status, but also to satisfy their desire for children. The more wives a man has, the stronger he is considered, the larger his farm (each wife cultivates a plot of land) and the greater the number of his children. It is somewhat surprising that the propertion of polygynous unions in urban areas is as high as reported. First of all, it might be assumed that the economic incentive for polygyny is less in urban areas where life is more competitive and the cost of living much higher than in rural areas. Moreover, urban residents are primarily engaged in economic activities apart from farming, thus eliminating the need for wives to cultivate plots of land. Nonetheless, the desire for children is one of the ultimate goals of most Liberian unions; even in monogamous unions, many men have children outside marriage by their girlfriends. In this type of situation, married women or women living together with a man, get to know who their husband's girlfriends are and may have reported them as other "wives." Hence, it is possible that not all of the polygynous unions reported in the LDHS are stable, cohabiting unions, but may include somewhat more casual liaisons. There are large differences in the extent of polygyny between regions. Over one-half of the married women in Grand Gedeh County are in polygynous unions, compared to 35 percent in Sinoe County, 26 percent in Montserrado County, and 41 percent in the rest of the country. To some extent, these regional differences reflect differences in the age, urban-rural, and ethnic com- position of the women in the counties. Table 2.3 further reveals that polygyny is inversely related to educational attainment such that uneducated women are more likely to be polygynous than women who have attained secondary or higher level of education. Since education and urban- rural residence are highly correlated, it is difficult to state the separate effects of these two vari- ables on polygyny. A higher proportion (51 percent) of Muslim women than Christian women (34 percent) are in polygynous unions. Although polygyny is contrary to the Christian religion, no attempt was made in the survey to measure the degree of devotion to any of the religious beliefs. Thirty-two percent of women in other or traditional religions and 41 percent of women who do not have any religious affiliation are in polygynous marital unions. Differences in the extent of polygyny by tribal groups range from a high of 57 percent among the Mandingo to a low of 24 percent among the Gola tribe. 2.3 Age at First Union All women who had ever been in either a legal or consensual union were asked the month and year when they started living with their first partners; if they could not remember the year, they were asked how old they were. Caution should be taken in interpreting the responses, since, as in 18 most developing countries, people have difficulty placing events in time. In the LDHS, 32 percent of ever-married women reported both a month and year of first marriage, 37 percent gave the year only, and 29 percent gave their age. Less than two percent of the respondents had the dates of their first marriage imputed, mostly relying on the dates of their first births. Another note about age at marriage concerns the custom of sending girls to live in the households of their future husbands at a young age. A small proportion of women reported very young ages at marriage due to this custom. Although the marriage was presumably not consum- mated until the girls matured, and therfore, they were not at risk of becoming pregnant, the data have been left as reported. Table 2.4 shows the distribution of women by age at first marriage. Fifty percent of Libefian women enter into marriage before reaching age 18. An additional 12 percent first marry between the ages of 18-19 years, while 17 percent marry at age 20 or older. TABLE 2.4 PERCENT DISTRIBUT/ON OF WOMEN BY AGE AT FIRST UNION AND MEDIAN AGE AT FIRST UNION, ACCORDING TO CURRENT AGE, LIBERIA, 1986 Aqe at First Union Current Never Total Wtd . Median Age Married <15 15-17 18-19 20-21 22-24 25+ Percent Number Age" 15-19 64.0 I I .6 21.0 3.4 100 1,137 - 20-24 24.7 16.6 31.7 15.8 7.5 3.7 100 1,030 18.2 25-29 7.9 16.3 35.0 17.3 I I . I 8.6 3.9 100 1,081 17.9 30-34 6.2 19.2 38.2 13.4 10.1 8.8 4.1 100 658 17.2 35-39 1.2 15.6 39.5 14.3 I0.4 9.1 9.9 100 626 17.2 40-44 1.7 29.1 40.5 I0.9 4.3 7.6 5.8 I00 327 16.0 45-49 0.5 19.4 41.4 8.7 12.0 7.2 I0.8 100 380 16.6 All Ages 21.4 16.6 33.1 12.1 7.4 5.7 3.7 I00 5,239 - Omitted due to censoring Defined as the age by which one-half of women have ever-married It can also be seen in Table 2.4 that younger women have a higher median age at first union than older women, which implies that age at marriage in Liberia has been increasing over time. This hypothesis is consistent with the report that the median age at marriage increased by about 1 year among females in the 1962-1974 intercensal period in Liberia (U.S. Bureau of the Census, 1982), as well as with the data presented in Table 2.2 on the increase hi the proportions never mar- fled by age between the 1974 and 1984 censuses. 19 Regarding differentials in age at first marriage, Table 2.5 presents the median ages at first mar- riage for women aged 25-49 according to background characteristics of the women. The data show a higher median age at first marriage among urban women (18.5 years) than rural women (16.8 years). There is also an indication of an upward trend in age at first marriage among younger women in both urban and rural areas, although it is more pronounced among urban women. An examination of differentials by county reveals that Grand Gedeh County has the lowest median age at marriage (16.4 years), with Sinoe County (17.6 years) and the rest of the country (17.0 years) reporting slightly higher ages at first marriage. Montserrado County displays the highest median age at marriage (18.8 years) and this pattern extends throughout all age groups. Education is highly correlated with age at first marriage (Table 2.5). The higher the level of education, the higher the median age at first marriage. At each age group, women with no educa- tion marry at much younger ages than women who have attended secondary school. Among religious groups, Christians report the highest median age at first marriage (18.0), while there is no real difference in median age at first marriage for the other three categories. Median ages at marriage by tribal groups are also shown in Table 2.5. These ethnic variations sug- gest that the Mano, Gin, Gola, Krahn, Mandingo, and Bassa marry earlier than women of other tribes. Distribution by age group seems to suggest a pattem of higher age at first marriage among younger women in some of the tribes. The remaining tribes show an irregular pattem, particularly among the older women who report relatively high ages at first marriage. This kind of inconsisten- cy may be due to memory lapse on the part of older women who could not remember their ages at first marriage. It is also possible that older women are reporting their age of entry into legal mar- riage as opposed to when they first started living with a man. 2.4 Exposure to the Risk of Pregnancy Although marriage pattems have an effect on fertility and family planning use, the concept of "exposure" to the risk of pregnancy is important in further refining the population of women who are of concem to family planning program administrators. Women are defined as potentially ex- posed to the risk of pregnancy if they are sexually active, ovulating, and fecund. In terms of the LDHS, sexual activity is measured directly by questions on when the respondent last had sexual intercourse. Since determining whether or not a woman is ovulating requires clinical testing, ovulation is measured indirectly in the LDHS, by questions on menstruation. Thus, women who are pregnant or whose periods have not returned after the birth of their last child are considered to be not ovulating. Similarly, infecundity, or the inability to conceive, is measured indirectly, by the absence of a birth for at least five years despite the non-use of family planning. Table 2.6 gives the distribution of currently married women by exposure status and age group. It should be noted that, due to the hierarchical nature of the exposure categories, they are not mutually exclusive (e.g., a woman may be both sexually inactive and infecund), and a somewhat different distribution would result if the order of the categories were changed. Fifteen percent of currently married women in Liberia are pregnant. As one would expect, the proportion currently pregnant declines with increasing age, ranging from over 22 percent among women in age group 15-19 to only five percent among those in the 45--49 age group. Amenor- 20 TABLE 2.5 MEDIAN AGE ATFIRSTUNIONAMONGWOMENAGED20-49 BYCURRENT AGE, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Women __ Current Aqe Aged Characteristic 20-24 25-2g 30-34 35-3g 40-44 45-4g 20-49 Urban-RuRal Urban 19.7 18,6 17.6 17.7 16.9 18.2 18,5 Rural 17.2 17.3 16.9 16.8 15.9 16.2 16.8 Reqion Sinoe 17.6 17.4 17.1 19.6 16.8 17.2 17.6 Grand Gedeh 17,3 17.4 16.5 15.7 15.6 16.3 16.4 Montserrado 20.0 1g.0 17.8 18.1 16.8 18.8 18.8 Rest of country 17.5 17.3 17.0 16.9 15.9 16.0 17.0 Education No education 17.4 17.3 16.9 16.8 15.9 16.3 16.8 Primary 17.2 17.2 17.7 18.3 17.8 17.5 17.3 Secondary or more 22.2 20.1 19.7 18.8 ~ 20.5 Reliqion Christian 1g.4 18.5 17.3 17.7 16.0 17.0 18.0 Muslim 17.7 17.4 16.2 16.1 15.6 16.5 16,8 Trad'I/Other 16.8 17.1 17.2 16.8 16.1 16.6 16.g None 17.1 16.8 17.6 16.g 16.1 16.0 16.9 Tribe Bassa 18.2 17.2 16.6 17.8 16.3 17.8 17.3 Gio 16,8 16,g 15,4 15.g ~ 15.1 16.0 Gola 19.5 18.0 " 16.0 " 16.7 Grebo 19.2 19.I 17.3 15.8 15.8 15.7 17.6 Kpelle 18.5 18.2 18.3 18.6 17.6 18.9 18.3 Krahn 17.5 16.9 17.9 17.2 15.6 16.5 16.9 Kru/Sapo 19.3 18.9 18.0 18.0 16.3 17.3 18.2 Lorma 19.3 17.5 17.9 18.0 * " 17.9 Mandingo 17.1 17.3 16.3 18.6 " " 17.0 Mano 16.6 15.7 15.6 " 14.8 14.9 15.7 Other/None 19.2 18.6 17.5 17.g 16.9 17.1 18.3 All Women 18.2 17.9 17.2 17.2 16.0 16.6 17.5 " Fewer than 20 unweighted cases 21 rheic women are defined as those who have not menstruated since their last birth. About 24 per- cent of married respondents said they were amenorrheic. The higher percentages reported among the younger age groups reflect their higher fertility rates. TABLE 2.6 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN BY EXPOSURE STATUS, ACCORDING TO AGE, LIBERIA 1986 Aqe Exposure A11 Status 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Ages Pregnant 22.3 18.9 19 .2 13 .3 12.8 7.0 4.8 15.4 Amenorrheic 21.7 28.2 28. I 26.7 22. I 13.5 7.5 23.5 Infecund 0.9 4.6 1O.2 17 .8 23 .7 47.5 56.2 18.1 Sexually inactive 14.2 12.5 10 .0 I0.4 10.0 7.5 6.9 I0.5 Exposed 41.0 35.7 32.5 32 .0 31.3 24.4 24.6 32.5 Total Percent 100.0 100.0 100.0 I00.0 100.0 I00.0 100.0 I00.0 Wtd. Number 360 675 857 539 535 261 311 3.538 Women are considered infecund if they did not have a birth in the five years preceding the sur- vey and had never used any form of birth control. This definition is not very precise, as it may in- clude some fecund women who have not had a birth because they are not sexually active or have had miscarriages, etc., and may exclude some menopausal women who had a birth during the five years before the survey or ever used family planning. In any case, according to this definition, about 18 percent of women are not exposed due to infecundity. As expected, the proportion in- fecund is very low in the younger age groups and increases in the older ages, particularly in age groups 40 4,1 (48 percent) and 45-49 (56 percent). Approximately ten percent of women are not exposed because they are not sexually active. The proportion of women who have not had sexual intercourse in the last four weeks seems to decrease with increasing age of respondents; however, this pattern is most likely due to the fact that many of the older women who are not sexually active have already been classified as in- fecund, if they have not had a birth in the last five years. After extracting the preceding categories, the residual gives the proportion of currently mar- ried women exposed to conception which, in Liberia, is 33 percent. Although the proportion is higher among younger women, even older women seem to have relatively high proportions sus- ceptible to becoming pregnant (about one-quarter of women aged 40-49). This implies that the provision of family planning services should not be limited to young women, but should also be extended to older women, although it should be noted that some of these women might want to have another child or might already be using some method of family planning. Family planning need is discussed in Chapter 5. 22 Table 2.7 shows the percent of all women currently exposed to the risk of pregnancy by mari- tal status. According to this table, 35 percent of all women 15-49 are exposed to the risk of preg- nancy. An almost equal proportion of women who are either currently or formerly married are ex- posed. The high proportions of never-married women who are at risk are disturbing, since most never-married women fall into the first two age groups. These data imply that efforts to reduce teenaged pregnancies should include single women. TABLE 2.7 PERCENTAGE OF WOMEN WHO ARE CURRENTLY EXPOSED TO THE RISK OF PREGNANCY, BY MARITAL STATUS, ACCORDING TO AGE, LIBERIA, 1986 Marital Status Age Currently Formerly Never Total 15-19 41.0 49.1 40.2 40.7 20-24 35.7 49.7 52,1 41.I 25-29 32.5 46.8 39,7 34.9 30-34 32.0 23,3 58.2 32.5 35-39 31.3 27.7 * 30.5 40-44 24.4 19.8 * 23.6 45-49 24.6 6.2 21.2 All Ages 32.5 34.0 43.1 34.9 "Fewer than 20 unweighted cases 2.5 Breastfeeding, Postpartum Amenorrhea, and Abstinence The LDHS collected data on several factors other than contraception that affect the length of pregnancy intervals; namely, breastfeeding, amenorrhea, and sexual abstinence. The information was obtained for the open interval (the interval since the last birth), for all women who had a live birth during the five years prior to the survey and was analyzed for all births within the 36 months prior to the survey. There were 3,249 weighted births occurring 0-35 months prior to the survey. The median durations of breasffeeding, amenorrhea and abstinence were calculated directly from the data given in Table 2.8. Mean durations of these variables, however, were calculated using the "current status" method, by dividing the total number of women breasffeeding, amenorrheic, or abstaining) by the average number of births per month over the past 36 months. Table 2.8 gives the proportion of women still breastfeeding, amenorrheic, and abstaining, by months since the birth. The results show that the practice of breasffeeding is very common among 23 Liberian women. Most women breasffeed their children for long periods, although the proportion still breastfeeding diminishes significantly after 17 months. The mean and median duration of breastfeeding are almost identical at 17 months. The average breastfeeding duration in Liberia is similar to that of other sub-Saharan countries, such as Benin (19 months), Cameroon, Ghana, Ivory Coast, and Senegal (18 months), and Kenya (17 months) (Singh and Ferry, 1984: 21). TABLE 2.8 PROPORTIONS OF WOMEN STILL BREAST~,EEDING, POSTPARTUM AMENORRHEIC, AND ABSTAINING, BY MONTHS SINCE BIRTH, AND MEDIAN AND MEAN DURATIONS OF BREASTI-EEDING, AMENORRHEA AND ABSTINENCE, LIBERIA, 1986 Months St i l l St i l l St i l l Weighted Since Breast- Amen- Abstain- Number of Birth feeding orrheic ing Births 0-I 93.4 86.0 98.1 189 2-3 87.1 72.7 88.2 234 4-5 88.3 66.5 80.1 208 6-7 75.1 52.9 68.3 243 8-9 74.g 48.1 53.9 250 I 0 - I I 65.4 38.0 51.8 209 12-13 60.9 34.5 34.3 204 14-15 55.8 30.1 31.9 183 16-17 51.9 16.6 21.7 157 18-19 27.1 8.3 I0 .7 133 20-21 22.4 4.4 7,2 155 22-23 16.5 6.0 5.9 141 24-25 12.5 4.5 7.3 167 26-27 8.4 3.2 3.9 184 28-29 5.8 3.0 1.2 167 30-31 4.3 2.1 4.4 143 32-33 4.6 2.0 2.0 138 34-35 2.0 l .2 0.0 145 Total 47.2 30.9 36.7 Median 16.7 7.7 I0.7 Mean 17.0 11.2 13.2 3,249 The period following a birth before the return of the menstrual cycle of a woman is usually referred to as the period of postpartum amenorrhea. In most societies, this period lasts about two to three months, during which time the woman is usually infecund. However, the length of time during which a woman is amenorrheic depends to a large extent on her physiological condition and such factors as nutrition, and the length of breastfeeding. The results in Table 2.8 indicate that the median duration of postpartum amenorrhea for Liberian women is about eight months and the 24 mean is eleven months. The mean is similar to that of other African countries such as Benin, Cameroon, and Ghana (12 months), and Ivory Coast and Kenya (10 months). Table 2.8 shows that among respondents who gave birth within three years prior to the survey, almost 30 percent had resumed their menstrual cycles within a period of 2-3 months since birth, while about 48 per- cent had resumed by 8-9 months. Fewer than 5 percent reported that they were still, amenorrheic two years after the birth of the child. In many areas of sub-Saharan Africa, postpartum sexual abstinence is widely practiced. The duration is usually tied to ongoing breastfeeding which is considered essential to the health and normal development of the child. The results in Table 2.8 indicate that postpartum sexual abstinence is generally long in Liberia with a median duration of 11 months. This is three months longer than the median duration of postpartum amenorrhea, and indicates that the period of postpartum protection from pregnancy is determined more by abstinence than by amenorrhea. Differentials in the mean duration of breasffeeding, amenorrhea, and abstinence by back- ground characteristics of the mother are presented in Table 2.9. Women under age 30 breasffeed their children an average of two and one-half months less than women aged 30 and over. To the extent that this represents a trend toward shorter breastfeeding durations, it is a disturbing finding. Although breastfeeding durations are relatively long in any case, a trend toward shorter durations can have adverse effects on child health, since breastmilk provides children with protection against certain illnesses. Urban women breastfeed their children an average of five months less than rural women (14 vs. 19 months). This urban-rural differential also underlies the much shorter average duration of breastfeeding among women in Montserrado County (13 months), than for women in other regions. Montserrado County contains Monrovia, where many women are engaged in economic activities and/or school, which tend to reduce the length of time that they can afford to breastfeed their children. Studies have shown that educated women are unlikely to breastfeed for long durations, primarily due to their greater participation in the labor force. This is confirmed in the case of Liberia, where women with secondary or higher education breastfeed their children for an average of only 10 months, as compared to 17 months for women with primary education and 19 months for women with no education. Differentials by religion show that Christian women breastfeed for slightly shorter durations than women of other religious affiliations, which reflects their higher educational status. Among the tribal groups, the average durations of breastfeeding range from a low of 15 months for Kru/Sapo, Gio, and Grebo women to a high of 19 months for Kpelle, Lorma, and Mandingo women. Breastfeeding tends to suppress the return of menstruation following a birth, thereby lengthen- ing the period of amenorrhea. This relationship is clearly seen from the data in Table 2.9, where the average durations of postpartum amenorrhea follow a similar pattern to those of breasffeeding. Thus, as with breastfeeding durations, durations of amenorrhea are shorter among younger women, urban women, women who live in Montserrado County, those who are better educated, Christian, or from the Bassa, KngSapo, or Gmbo tribes. Differentials in duration of postpartum abstinence generally are not as large as those for breastfeeding and amenorrhea, although they tend to follow a similar pattern. Most women report durations of abstinence of about 11 or 12 months. The major exceptions to this pattern are women 25 TABLE 2.9 MEAN DURATION OF POSTPARTUM BREASTr, V.,EDING, AMENORRHEA, AND SEXUAL ABSTINENCE (CURRENT STATUS ESTIMATE BASED ON B/RTHS WITHIN 36 MONTHS OF THE INTERVIEW DATE), ACCORDING TO BACKGROUND CHARACTERISTICS OF MOTHER, LIBERIA, 1986 Wtd. No. Characteristic Breastfeeding Amenorrhea Abstinence of Births Aae < 30 16.2 I0.6 13.1 2,212 30+ 18.7 12.4 13.4 1,055 Urban-Rural Urban 14.1 9.2 12.1 1,334 Rural 18.9 12.6 13.9 1,933 Reqion Sinoe 17.3 9,8 12.6 98 Grand Gedeh 21.5 13.2 14,0 198 Montserrado 12.9 8.3 l l .3 823 Rest of country 18.1 12.2 13.9 2.148 Edwca~iQn No education 18.6 12.4 14.1 2,150 Primary 17.1 l l . l 13.7 598 Secondary or more lO.O 6.4 9.0 519 Reliqion Christian 15,8 I0.5 I I .7 1,627 Muslim 18.4 I I .4 15.8 521 Trad'I/Other 17.3 11.3 15.0 661 None 18.6 12.8 12.9 459 Tribe Bassa 16.1 9.4 10.8 436 Gio 14.7 I I .3 l l .7 247 Gola 18.2 14.4 15.2 131 Grebo 15.4 lO.O 8.7 232 Kpelle 19.1 12.1 14.1 554 Krahn 18.1 I I .5 13.7 153 Kru/Sapo 14.6 9.9 10.4 335 Lorma 18.5 13.4 17.5 159 Mandingo 18.9 l l .5 15.8 213 Mano 17.2 12.0 13.2 248 Other/None 16.7 I I .0 15.6 560 Total 17.0 I I .2 13.2 3,267 26 with secondary or higher education, who abstain for only 9 months, and, surprisingly, Muslim women who abstain for almost 16 months on average. There is also some variation by tribe, with Grebo, Kru/Sapo, and Bassa women reporting shorter than average durations of abstinence, and Lorma and Mandingo women reporting !onger durations. 27 3. FERTILITY 3.1 Fertility Data Information about fertility levels, trends and differentials was one of the most important topics to be covered in the LDHS. Several questions were asked to derive the total number of live births for each respondent. First, each woman was asked the number of sons and daughters living with her, the number living elsewhere, and the number who had died. Then she was asked for a history of all her births, including the month and year each was bom, the sex, the name, and, if dead, the age at death, and, if alive, whether be/sbe was living with the mother. Although the birth history approach is an accepted and widely used method of collecting fer- tility data, it has limitations and is susceptible to data collection errors. One problem is that events to non-surviving women are not collected. A second problem is that, for earlier time periods, infor- marion on births to women in the older ages of childbearing are not collected. Neither of these problems is considered serious for time periods close to the survey date. Moreover, in order to minimize the effect of the latter problem, total fertility rates presented in this report have been cal- culated for women aged 15--44, instead of the customary 15-49 age group. Defects in data collection take the form of underreporting of births (especially those that die in early infancy) and misreporling of date of birth. Fertility levels can be affected by underreporting, while misreporting of dates of births can seriously distort estimates of trends in fertility over time. There is some indication that hi the LDHS, births occurring four and five years prior to the survey may have been shifted to six years before the survey, presumably so as to avoid the necessity of filling in the health section for those children. 3.2 Levels and Differentials in Fertility LDHS data indicate that the total fertility rate in Liberia is 6.5 children per woman. This means that if current age-specific fertility rates were to stay constant, Liberian women would have an average of six and one-half live births by the time they reached the age of 45. This level is slightly higher than the rate of 6.1 that was found in the 1970-71 Liberian Population Growth Survey (Republic of Liberia, 1971: 9) and of 6.2 that was estimated from 1974 Census data (University of Liberia and MPEA, 1981: T-63) for women 15-44. Thus, LDHS data imply that fertility in Liberia has been more or less constant in the recent past, and may be increasing slightly. Table 3.1 presents the total fertility rates (TFR) for two calendar year periods (1983-1986 and 1980-1982) and for the five years preceding the survey, as well as the mean number of children ever born (CEB) to women 40-49 years old, according to background characteristics of women. Caution should be exercised in comparing these rates, as the average number of births to women 40~9 refers to past or completed fertility, while total fertility in the preceding five years refers to a more current measure of fertility. Comparing the last two columns of the table confirms that fertility has not changed significant- ly in the recent past. The average number of births to women 40-49 is 6.4 while the total fertility rate five years before the survey is almost the same at 6.3 children. A comparison of children ever bom and fertility 0-4 years before the survey, according to background characteristics, suggests 29 TABLE 3.1 TOTAL FERTILITY RATES FOR 1983-86, 1980-82, AND THE FIVE-YEAR PERIOD IMMEDIATELY PRIOR TO THE SURVEY AND MEAN NUMBER OF CHILDREN EVER BORN TO WOMEN 40--49, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Characteristic Total Fer t i l i ty Rate a Mean No. Children 0-4 years Born to 1983-86 b 1980-82 before survey Women 40-49 Urban-Rural Urban 6.1 6.3 5,9 6.6 Rural 6.8 6.9 6.6 6.3 Reqion Sinoe 7.1 7.8 7.1 7.2 Grand Gedeh 7.4 7,7 7,3 7.0 Montserrado 5.6 6.1 5.5 6.4 Rest of country 6.7 6,7 6.5 6.2 Education No Education 6.7 6.8 6,4 6.3 Primary 7.0 7.7 7.0 7.1 Secon'y or more 4.8 4.9 4.7 6.3 Reliqion Christian 6.0 6,7 6.0 6.3 Muslim 6.8 6.6 6.3 6.9 Trad'I/Other 7.1 7.0 6.8 6.8 None 7.1 6.5 6.6 5.9 Tribe Bassa 6.7 6.7 6.5 5.6 Gio 6.3 6.3 6.1 5.9 Gola 5.9 7,2 5.7 7.3 Grebo 6,7 7.0 6.5 6.8 Kpel le 6.9 7.0 6.7 7,1 Krahn 7.2 7.5 7.1 6.4 Kru/Sapo 6.1 7.2 6.4 6.9 Lorma 5.6 5.9 5.2 Mandingo 6.0 6.7 6.0 * Mano 6.4 6.5 6.2 5.0 Other/None 6.7 6.4 6.3 6.7 Total 6.5 6.7 6.3 6,4 * Fewer than 20 unweighted cases a Based on births to women aged 15 to 44 b Includes births occurring in 1986, up to the time of the survey 30 that fertility is declining among urban women, women in Montserrado County, women with secondary education, Muslims, and women from the Gola and Kru/Sapo tribes. Total fertility rates for all three time periods indicate that urban fertility is lower than rural fer- tility. Based on births in the five years before the survey, women in urban areas have a total fer- tility rate of approximately six births while rural women have about six and one-half births. The fact that the mean number of children ever bom to women 40--49 is higher for urban women may be due to poor recall among older, rural women. Regional differences in total fertility rates indi- cate that women in Grand Gedeh and Sinoe Counties have substantially higher fertility--well above seven births per woman--than women living either in Montserrado or in the rest of the country. This is true for the measures of current fertility as well as the average number of children born to women 40-49. The lowest fertility rates are reported in Montserrado County, with an average of fewer than six births per woman. In terms of education, it is interesting to note that the fertility of women with some primary education is higher than for those with no education and substantially higher than the fertility of women with secondary education. Although this pattern is quite common in African countries, its determinants are not readily explainable, since LDHS data on age at marriage, duration of postpar- tum amenorrhea and abstinence, and use of contraception imply that the fertility of women with primary education should be intermediate to that of women with no education and women with secondary education. In fact, however, it appears that more educated women initiate childbearing at an earlier age than do those with no education (see Table 3.6). It is also possible that the relationship is due to greater pregnancy wastage and/or underreporting of births among women with no education. A more detailed study of the survey data might provide an explanation for this phenomenon. Christian women generally have somewhat lower fertility than Muslim women or women in traditional or other sects. Comparison of fertility levels among the tribes is not straightforward due to the varying sample sizes. In terms of current fertility, the tribe that has the highest level is the Krahn, with a total fertility rate of about seven children per woman. Other tribes with high fer- tility are the Kpelle, Grebe and Bassa. Those with the lowest fertility are the Lorma. Looking at measures of completed fertility, some of the tribes that currently have high levels of fertility have had much lower levels in the past. This is true for the Mano and Bassa tribes. The Gola indicated the highest completed mean fertility of 7.3 children; the Liberia Population Growth Survey and various researchers have also reported this phenomenon (Chieh-Joimson, 1987). 3.3 Fertility Trends Table 3.2 presents age specific fertility rates for five-year periods prior to the survey, based on data from the birth histories. To compute the numerator for these rates, births were classified by the segment of time preceding the survey (e.g., 0-4 years, 5-9 years prior, etc.), and by age of the mother at the time of birth. The denominator is the number of woman-years lived in the specified five-year age interval for each time segment. Because the LDHS only interviewed women up to age 50, the data in Table 3.2 become "truncated" the farther back in time, because information for women in older age groups is either incomplete or missing. 31 TABLE 3.2 AGE-SPECIFIC FERTILITY RATES (PER THOUSAND WOMEN) FOR FIVE-YEAR PERIODS PRIOR TO THE SURVEY, LIBERIA, 1986 Years Prior to the Survey Age of Women 0-4 5-9 lO-14 15-19 20-24 25-29 30-34 15-19 184 192 173 173 171 188 20-24 285 293 268 260 249 (263) 25-29 272 281 274 257 (299) 30-34 223 266 232 (291) - 35-39 181 198 (210) - 40-44 I14 (180) - 45-49 (63) - Cumulated Fer- t i l i t y 15-29 3.7 3.8 3.6 3.5 (118) Note: Numbers in parentheses are part ia l ly truncated rates. The data show a peak in fertility 5-9 years prior to the survey for all age groups. While it is possible that fertility rose and then fell, it is more likely that the peak is artificially created by mis- reporting of dates of births, namely, a shifting of births from the period 10-14 and/or 0-4 years prior to the survey to the period 5-9 years prior to the survey. Ignoring this peak, the cumulated rates for ages 15-29 show evidence of a slight increase in fertility over time, which is most pronounced at age group 20-24. This increase might be due in part to a decline in breastfeeding duration among younger women, The data in Table 3.2 also indicate the extreme youthfulness of childbearing in Liberia. Almost 20 percent of teenage girls and over 25 percent of women 20-24 give birth in a given year. Such early childbearing has serious implications for both maternal and child health. 3.4 Children Ever Born Completed family size in Liberia is quite high (see Table 3.3). By the time a Liberian woman reaches the end of her childbearing period (usually between ages 45-49), she would have given birth to about seven children (6.8). Just as marriage occurs relatively early, childbearing also oc- curs early, with teenage girls reporting an average of 0.5 births. The average number of live births increases with the woman's age. Women in their late twenties report just over three live births and women in their late thirties report an average of more than five births. The distribution of women by number of births reveals that almost 40 percent of teenagers and 80 percent of women 20-24 have had at least one child. By the time women reach the end of childbearing, one-quarter have had ten or more live births. Primary infertility--the proportion of married women aged 45-49 who never have children---is quite low at about two percent. 32 TABLE 3.3 PERCENT DISTRIBUTION OF ALL WOMEN AND CURRENTLY MARRIED WOMEN BY NUMBER OF CHILDREN EVER BORN, ACCORDING TO AGE, LIBERIA, 1986 Age Wtd. Mean Number of Children Ever Born No. No.of Total of Chil- 0 1 2 3 4 5 6 7 8 9 I0+ Percent Women dren All Women 15-19 62.8 28.9 7.5 0.7 0.1 0.0 0.0 0.0 0.0 0.0 0.0 100.0 1,137 0.5 20-24 19.3 26.9 25.8 18.2 6.6 2.5 0.7 0.0 0.0 0.0 0.0 100.0 1,030 1.8 25-29 6.5 14 .7 17 .2 18 .2 20.0 13.4 5.3 3.2 1.2 0.3 0.0 I00.0 1,081 3.2 30-34 5.0 B.3 11 .6 16 .2 15 .2 15 .6 12.2 6.9 4.2 3.0 l .8 100.0 658 4.2 35-39 3.2 6.1 8.7 I I .9 8.9 12.2 13.1 13.5 I0.7 5.1 6.6 100.0 626 5.3 40-44 3.4 8.3 8.3 8.5 9.6 9.1 9.2 10.4 6.3 8.4 18 .4 I00.0 327 5.9 45-49 2.6 4.8 8.0 5.I 8.7 5.6 I I .5 10.8 9.6 8.1 25.0 I00.0 380 6.8 All Ages 20.2 17 .2 13 .9 I I .8 9.7 7.6 5.7 4.6 3.2 2.1 4.0 I00.0 5,239 3.1 Currently Married WQm@n 15-19 40.0 43.4 15.8 0.5 0.4 0.0 0.0 0.0 0.0 0.0 0.0 I00.0 360 0.8 20-24 15.3 24 .1 28.6 21.2 7.5 2.9 0.4 0.0 0.0 0.0 0.0 100.0 675 l .9 25-29 5.0 14 .0 16 .5 17 .9 20.6 15.0 6.0 3.6 0.9 0.4 0.0 100.0 857 3.3 30-34 5.0 6.2 8.9 17 .0 15 .0 17 .7 12.6 7.6 4.6 3.4 2.1 100.0 539 4.4 35-39 3.2 5.3 8.5 11.8 8.0 12 .9 14 .2 14 .0 10.6 5.2 6.4 I00.0 535 5.4 40-44 3.7 7.7 7.9 5.8 10.7 9.8 9.0 10.3 7.1 8.8 19 .3 I00.0 261 6.0 45-49 2.1 4.0 8.0 5.8 7.8 5.5 l l .0 10 .3 I0.6 7.7 27.0 100.0 311 6.9 All Ages 9.9 15.1 15.0 13.7 11.4 10.0 7.2 5.8 4.0 2.7 5.1 100.0 3,538 3.8 The mean number of children ever born by age at first marriage and duration of marriage is given in Table 3.4. As expected, the mean number of children born rises with increasing marital duration. The results indicate that irrespective of the age at first marriage, a Liberian woman would have given birth to an average of almost three children during the first 5-9 years ofber marriage. At shorter marriage durations, the mean number of children ever born increases with age at marriage. This could be due to the fact that late-marrying women have had pre-marital births, or that they have shorter birth intervals due to shorter breasffeeding durations. At longer durations of marriage, the relationship between children born and age at marriage is erratic. One clear pattern in Table 3.4 is the lower mean number of children ever born for women who married below age 15, which probably reflects adolescent subfecundity. 33 TABLE 3.4 MEAN NUMBER OF CHILDREN EVER BORN TO EVER-MARRIED WOMEN AGED 15-49, ACCORDING TO AGE AT FIRST MARRIAGE AND YEARS SINCE FIRST MARRIAGE, LIBERIA, 1986 Years Age At First Marriage Since First All Marriage <15 15-17 18-19 20-21 22-24 25+ Ages 0-4 0.9 1.0 1.3 1.4 1.4 2.3 1.2 5-9 2.1 2.6 2.8 2.9 2.9 3.5 2.7 I0-14 3.5 4.0 4.2 4.1 3.9 3.5 3.9 15-19 4.5 5.1 4.9 5.5 6.2 4.8 5.0 20-24 5.6 6.1 5.7 6.0 6.2 5.0 5.9 25-29 5.8 6.4 6.0 6.5 4.5 - 6.1 30 or more 7.4 7.5 6.0 - 7.4 All Years 4.1 4.0 3.3 3.5 3.3 3.3 3.8 3.5 Age At First Birth The onset of fertility is an important demographic indicator. In many countries, the postpone- ment of first births has had a large impact on overall fertility decline. Also, the proportion of women who become mothers in their teenage years is a basic indicator of maternal and child health. Table 3.5 shows the distribution of women by age at first birth and current age. The data show that over one-half of Liberian women become mothers before they reach age 20. This find- ing has serious health implications, since young mothers suffer more health problems than older mothers, and their children have higher mortality rates. The data imply that age at first birth has been declining over time, with younger women having lower median ages at first birth. Although it is possible that childbearing is starting at younger ages than in the past, it should be noted that the data in Table 3.5 are heavily dependent on correct reporting of dates of birth of beth the woman and her first birth. It is perfectly possible that older women had difficulties in remembering the dates of their first births and pushed them closer to the date of the survey, thereby making themselves older at first birth than they actually were. Another plausible explanation is that some women who were actually aged 40 A~ at the time of the survey were recorded as being in the age group 45-49. This would have the effect of falsely increasing their age at first birth. For example, if a 43-year old woman whose first-born child is 23 years old, is erroneously recorded as being 48, her age at first birth would be calcu- lated as 25, instead of 20. This latter hypothesis would also explain why there are more women 45-49 than 40-44. Given these likely defects in the data, it is uncertain whether age at first birth is in fact declining in Liberia; however, it is clear that childbearing starts at very young ages. 34 TABLE 3.5 PERCENT DISTRIBUTION OF ALL WOMEN BY AGE AT FIRST BIRTH, ACCORDING TO CURRENT AGE, LIBERIA, 1986 Wtd. Median Age At First Birth Number Age at Current No Total of First Age <15 15-17 18-19 20-21 22-24 25+ Births Percent Women Birth 15-19 5.4 25.5 6.3 0.0 0.0 0.0 62.8 100 .0 1,137 20-24 I0.5 33.9 19.8 12.3 4.3 0.0 19.3 100 .0 1,030 18.5 25-29 7.9 29.9 24.1 15.3 I I .2 5.1 6.5 100 .0 1,081 19.0 30-34 10.2 29.5 15.7 12.8 17.6 9.2 5.0 100.0 658 19.4 35-39 8.9 25.8 17.4 12.4 13.6 18.7 3.2 I00.0 626 19.8 40-44 I I .4 35.4 12.1 6.7 15.9 15.O 3.4 I00.0 327 18.6 45-49 8.1 22.2 13.4 ]4.7 12.3 26.8 2.6 100.0 380 21.0 All Ages 8.5 29.0 16.0 10.2 8.9 7.3 20.2 100 .0 5,239 - Omi t ted due to censoring Table 3.6 presents data on differentials in age at first birth by background characteristics of women. Surprisingly, urban women generally start childbearing earlier than rural women. Also somewhat surprising is the fact that women with no education report the highest ages at first birth. While these findings may be due to data defects, it is also possible that better educated, urban women are less restricted by traditional restraints and initiate sexual activity sooner than their less educated, rural counterparts. Better educated, urban women presumably have better nutrition and medical care than less educated rural women, which may lead to earlier onset of menarche and a lower incidence of miscarriage and stillbirths. The differential in age at first birth by education may account for the higher fertility evidenced by women with primary education as compared to those with no education or secondary schooling (see Table 3.1). 35 TABLE 3.6 MEDIAN AGE AT FIRST BIRTH AMONG WOMEN AGED 20--49, ACCORDING TO CURRENT AGE AND BACKGROUND CHARACTERISTICS OF WOMEN, LIBERIA, 1986 Current Age Ages Characteristic 20-24 25-29 30-34 35-39 40-44 45-49 20-49 Urban-Rvral Urban 18.7 18.9 18.9 19.1 17.8 20.7 18.9 Rural 18,3 19.1 19.9 20,2 19.1 21.I 19.3 Reqion Sinoe 17.6 18,0 18.2 21.6 17.9 18.6 18.3 Grand Gedeh 18,1 18,5 17.3 18.8 18.4 20.2 18.4 Montserrado 18.6 19,0 18.9 18.8 18.9 20.8 18.9 Rest of country 18.5 19,1 19.9 20.1 18.5 21.1 19.4 Education No education 18,8 19.3 19.7 20.1 19.1 21.0 19.5 Primary 18.1 17,9 18.5 18.3 16.5 19.4 18.1 Secon'y or more 18.2 18,8 19.1 18.7 * * 18.7 R~liqion Christian 18.2 19,0 18.8 19.1 17.9 20.4 18.8 Muslim 20.1 19.4 19.9 20.5 18.5 21.5 19.9 Trad'I/Other 18.1 18,4 19.6 20.4 18.4 19.0 18.8 None 18,6 19.1 19.9 21,2 22.2 21.9 19.8 Tribe Bassa 18.6 18.2 17.9 19.0 19.3 19.5 18.6 Gio 17.1 18.4 18.0 24.0 * 21.9 18.9 Gola 17.5 19.2 * 19.5 * * 19.3 Grebo 18.2 18.8 17.3 18.2 17.0 21.5 18.3 Kpelle 19.2 19.0 21.6 19.6 18.9 18.8 19.4 Krahn 17.8 18.1 18.? 18.8 18.3 19.8 18.4 Kru/Sapo 18.0 18.6 18.1 18.0 17.7 20.8 18.2 Lorma 19.5 20.I 19.3 20.0 * 19.9 Mandingo 20.2 20.1 20.6 24.7 * 20.8 Mano 17.9 18.8 22.1 21.0 20.7 25.0 19.4 Other/None 18.8 19.7 20,1 20.5 17.3 20.5 19.5 All Women 18.5 19.0 19.4 19.8 18,6 21.0 19.2 * Fewer than 20 unweighted cases 36 4. CONTRACEPTIVE KNOWLEDGE AND USE 4.1 Contraceptive Knowledge 13ecause knowledge of contraceptive methods and of places where methods can be obtained are preconditions for their use, determining the level of knowledge of methods and service providers was a basic objective of the LDHS. The survey collected knowledge data first by asking the respondent to name the ways that a man or woman could keep a woman from getting preg- nant. If a respondent did not spontaneously mention a particular method, the method was described by the interviewer and the respondent was asked if she recognized the method. Descrip- tions were included in the questionnaire for seven modem methods (the pill, IUD, injection, con- dom, vaginal methods (diaphragm, foam and jelly), female sterilization, and male sterilization) and two traditional methods (periodic abstinence (rhythm) and withdrawal). In addition, other methods mentioned by the respondent, e.g., herbs, were recorded. Finally, for any modem method that she recognized, the respondent was asked if she knew about a place or a person from which she could obtain the method. If she reported knowing about rhythm, she was also asked if she knew a place or person from which she could get information about the method. The DHS results indicate that 72 percent of Liberian women know at least one contraceptive method (Table 4.1). The]¢ are more likely to report having heard about modem methods (70 per- cent) than traditional methods (30 percent). The pill-----which 64 percent recognize--is the most widely known method. Considering other methods, more than 40 percent have heard about injec- tion and female sterilization, while around 30 percent are familiar with the IUD and the condom. The percentages knowing about rhythm and withdrawal are the same--16 percent---compared to 13 percent recognizing folk methods and 12 percent knowing about vaginal methods. Only 6 per- cent say that they have heard about male sterilization. Surprisingly, knowledge of all methods ex- cept female sterilization is slightly lower among currently married women than among all women. If women are to adopt family planning, they must not only know about methods but they must also be aware of a source from which they can obtain contraceptive services. In the LDHS, less than half of all respondents were familiar with a source from which modem methods or informa- tion about the use of the rhythm method could be obtained. Table 4.1 shows that the percentage knowing a source was highest in the case of injection (29 percent), female sterilization (28 per- cent) and the pill (27 percent) and lowest in the case of vaginal methods (9 percent), periodic abstinence (7 percent) and male sterilization (4 percent). Table 4.2 indicates how the percentage of currently married women knowing any modem method and the percentage knowing a source vary among subgroups within the Liberian popula- tion. Both age and number of living children are related to contraceptive knowledge. The relation- ship with age is U-shaped, with knowledge levels highest in the 20-34 age groups. The percentage knowing a method or a source generally increases directly with the number of living children. Urban women are more likely than rural residents to know about a method and to be able to name a source. By region, knowledge levels are highest in Sinoe County followed by Montser- rado County. The fact that the proportions knowing any method and knowing a source are higher in Sinoe than in Montserrado County (which includes Greater Monrovia) likely reflects the fact 37 that prior to the LDHS, Sinoe County was the target of a special campaign by the Liberian Family Planning Association to increase contraceptive knowledge and use. TABLE 4. l PERCENTAGE KNOWING ANY CONTRACEPTIVE METHOD AND KNOWING A SOURCE (FOR INFORMATION OR SERVICES) FOR A METHOD AMONG ALL WOMEN AND CURRENTLY MARRIED WOMEN BY METHOD, LIBERIA, 1986 Method Knowinq Method Knowlnq Source AW CMW AW CI~ Any Method 71.8 69.8 47.7 44.3 Any Modern Method 70.4 68.0 47.3 44,0 Pill 64.0 61.1 26.6 24.8 IUD 34.5 31.3 21.0 17.6 Injection 44.4 42.4 28.5 25.1 Vaginal methods I I .6 9.8 8.6 6.9 Condom 30.5 26.2 17.0 14,2 Female sterilization 41.0 41.7 27.7 27.4 Male sterilization 6.4 5.9 4.4 4.0 Any Traditional Method 30.0 27.8 Periodic abstinence 15.5 12.4 Withdrawal 15.5 13.4 Other methods 13.2 14.0 7.1 5.7 AW = a l l women CFt4 = cur rent ly marr ied women Educational status differentials in contraceptive knowledge are substantial. For example, only 60 percent of currently married women who had never attended school knew any modem method, compared to 85 percent of those with some primary education and 95 percent of those with some secondary education. The religion to which a woman belongs is also associated with the level of contraceptive knowledge. Knowledge of modem methods is greatest among Christian women and least among Muslim women. However, those Muslim women knowing a method are more likely than women who profess no religion to know a source. Table 4.2 also shows that contraceptive knowledge varies with a woman's tribe. The highest percentages knowing a method and knowing a source are found among women belonging to the Kru/Sapo and Grebo tribes. Other tribes for which knowledge levels exceed those reported for the country as a whole include the Bassa, Lorma, Gio, and Krahn. 38 TABLE 4.2 PERCENTAGE KNOWING ANY MODERN CONq~ACEPTIVE METHOD AND PERCENTAGE KNOWING A SOURCE FOR MODERN METHODS AMONG CURRENTLY MARRIED WOMEN, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Knows Knows Knows Knows Characteristic Method Source Characteristic Method Source Aoe Education 15-19 53,2 27,0 No schooling 60.4 35.6 20-24 71.2 46.0 Primary 85.1 55.8 25-29 71.2 47.9 Secondary or more 94.9 80.9 30-34 73,8 50.1 35-39 67.1 45.8 Reliqion 40-44 64.3 43.5 Christian 78.7 56,4 45-49 64.0 35.4 Muslim 49.4 32.7 Traditional/Other 72.9 40.0 Livinq Children None 56.6 28,0 None 59.1 32,5 l 63.4 39.1 Tribe 2 65.6 45.5 Bassa 75.1 41.4 3 71.4 45.4 Gio 69.8 30.8 4 70.6 45.6 Gola 53.2 38.5 5 72.3 50.5 Grebo 82.7 68.0 6 or more 78.8 54.4 Kpel le 68,1 44.8 Krahn 68.3 51.3 Urban-Rural Kru/Sapo 86.4 66.5 Urban 76.8 56.5 Lorma 75.1 50.1 Rural 62.8 36.6 Mand~ngo 41.2 26.8 Mano 63.5 30.2 Other/None 62.6 43.2 Reqion Sinoe 87.2 65.4 Grand Gedeh 64.1 49.2 Montserrado 77,3 58.2 Rest of country 64,4 37.7 Total 68.0 44.0 4.2 Ever Use of Contraception The LDHS asked whether women had ever used each method that they knew and then ob- tained information on whether they were currently using a method, and, if so, the method that they were using. A total of 22 percent of all women reported that they had ever used a contracep- tive method (Table 4.3). The level of ever use among currently married women (19 percent) is slightly lower than the level for all women. Among ever users, the majority have had experience with modem methods. The pill is by far the most frequently adopted modem method. Sixteen per- cent of women have used the pill, while the percentages reporting ever use of other modem methods do not exceed three percent. 39 TABLE 4.3 PERCENTAGE WHO HAVE EVER USED CONTRACEPTIVE METHODS AMONG ALL WOMEN AND CURRENTLY MARR/ED WOMEN BY METHOD, LIBERIA, 1986 Currently Method All Women Married Women Any Method 21.7 18.8 Any Modern Method 18.6 15.9 Pill 15.7 13.4 IUD 2.9 2.8 Injection 1.6 1.8 Vaginal methods 1.0 1.0 Condom 2.7 1.7 Female sterilization 1.0 l . l Male sterilization 0.0 0.0 Any Traditional Method 8.3 6.7 Periodic abstinence 4.4 3.3 Withdrawal 4.2 3,4 Other methods 1.3 1.2 The percentage of women in Liberia who reported ever using a traditional contraceptive method--8 percent--is fairly low in comparison to the level of use of these methods in many other African countries. Looking at specific traditional methods, 4 percent of the LDHS respon- dents said they had employed periodic abstinence, an identical percentage said they had relied on withdrawal and less than 2 percent indicated that they had tried folk methods. Successful practice of periodic abstinence is dependent upon a correct understanding of when in the ovulatory cycle a woman is most likely to become pregnant. Table 4.4 presents the distribu- tion of all respondents and the small number of respondents who had ever used periodic abstinence by knowledge of the period in the ovulatory cycle when a woman is fertile. Almost two-thirds of the women said that they did not know when a woman was most likely to become pregnant. Ever-users of periodic abstinence seem to be much more knowledgeable about the ovulatory cycle, but, even in that group, only about one in five identified the fertile period as oc- curring in the middle of the cycle ("between two periods"). It should be noted that methods of dividing the ovulatory cycle are culture-specific and many of the women who were coded in the category "soon after period ends" may actually have a fairly accurate understanding of their fer- tile period. 4.3 Current Use of Contraception Eight percent of the LDHS respondents indicated that they were currently using a contracep- tive method (Table 4.5). As with ever-use, the comraceptive prevalence rate among currently mar- tied women is somewhat lower (6 percent) than among all women. Figure 4.1 gives an overview 40 TABLE 4.4 PERCENT DISTRIBUTION OF ALL WOMEN AND WOMEN WHO HAVE EVER USED PERIODIC ABSTINENCE BY KNOWLEDGE OF THE FERTII .~. PERIOD DURING THE OVULATORY CYCLE, LIBERIA, 1986 Ever Users of Fertile Period All Women Periodic Abstinence During menstrual period 1.7 1.2 Soon after period ends 18.5 30.6 Between two periods 6.6 21.7 3ust before period begins 4.1 5.7 At any time 6.7 3.6 Other 0.2 Don't know 62.1 37.1 Total Percent lO0.O lO0.O Number 5,239 229 TABLE 4.5 PERCENT DISTRIBUTION OF ALL WOMEN AND CURRENTLY MARRIED WOMEN, ACCORDING TO CONTRACEPTIVE METHODS CURRENTLY USED, LIBERIA, 1986 Currently Method All Women Married Women Using Any Method Using Any Modern Method 8.4 6.4 7.0 5.5 Pill 4.7 3.3 IUD 0.7 0.6 Injection 0.3 0.3 Vaginal methods 0.1 0.2 Condom 0.2 O.O Female sterilization l.O l . l Male sterilization O.O O.O Using Any Traditional Method l .4 O .9 Periodic abstinence 0.9 0.6 Withdrawal 0.2 0.I Other methods 0.3 0.2 Not Using 91.6 93.6 Total Percent lO0.O lO0.O Number 5,239 3,538 41 of family planning knowledge and use and shows the rapid drop-off in the proportion of women who know about methods, know a source, have ever used, and am currently using family planning. Figure 4.1 Family Planning Knowledge and Use Currently Married Women 15-49 8O 70 / 6O 50 40 3O 20 10 0 Knows Me[hod 70 / 44 / / / /" 19 i. / nows Source Ever Used Family Planning 6 13-- ~aUrren Iv Usina mi/y P1 anni n~i Most users rely on modem methods. The pill is the predominant method; more than half of all current users rely on the pill. Female sterilization, periodic abstinence and the IUD account for the majority of the rest of the users (see Figure 4.2). Table 4.6 presents the relationship between the level of use of contraceptive methods and back- ground characteristics of currently married women. With regard to age patterns, the percentage of currently married women using a method increases directly with age, from 2 percent in the 15-19 age group to 8 percent in the 25-29 cohort, after which, the use rate declines to 5 percent among women 35-39 before peaking again at 8 percent for the 40-49 cohorts. Family planning use also increases with the number of living children that a woman has, ranging from 3 percent among women with no children to 12 percent among those with 6 or more children. The percentage using family planning among urban women (12 percent) is almost four times the rate observed for rural women. The urban character of Montserrado County likely explains its relatively high use rate. Reflecting their predominantly rural population, the percentages using family planning in both Sinoe and Grand Gedeh Counties are slightly smaller than that reported for Liberia as a whole. Very large differentials in contraceptive use are apparent when women are classified by educa- tion level. The use rate among women with some secondary schooling is roughly three times the rate among women who attended only the primary grades and nearly ten times the rate among women who never went to school (see Figure 4.3). Christians are more likely to be using than are 42 women of other religious groups. The Grebo, KngSapo and Lorma have the highest contraceptive use rates among the various tribes in Liberia. Figure 4.2 Current Use of Family Planning by Method Currently Married Women 15-49 Not Using 94% Tradit ional 1% Other Modern 2% Pil l 3% 4.4 Source For Methods Information on the source for contraceptive methods was obtained by asking women using modem methods about where they had obtained their methods the last time and by asking women relying on periodic abstinence where they had received advice about the method. The most frc- quentiy mentioned source was the Family Planning Association of Liberia (FPAL) which provided 40 percent of the users with methods or advice (Table 4.7 and Figure 4.4). Government hospitals or clinics provide contraceptive services for 29 percent of users while 23 percent rely on private sector sources, including church hospitals or clinics, private doctors or pharmacies or shops. The sources reliod on by users vary with the method used. Women taking the pin--who com- prise the majority of users--arc much more likely to name FPAL as their service provider than are users of other methods. Government clinics or hospitals are the primary source for the compara- tively small number of women using modem methods other than the pill. Among periodic abstinence users, relatives or friends were the most frequently cited source of advice. 43 TABLE 4.6 PERCENTAGE CURRENTLY USING ANY CONTRACEPTIVE METHOD AND CURRENTLY USING ANY MODERN CONTRACEPTIVE METHOD AMONG CURRENTLY MARRIED WOMEN BY BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Any Any Modern Any Any Modern Characteristic Method Method Characteristic Method Method Aqe 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Livinq Children None 2.5 1 3.8 2 5.7 3 6.4 4 9.7 5 6.5 6 or more 12.1 Urban-Rural Urban II .6 Rural 3.4 Region Sinoe 4.4 3.9 Grand Gedeh 3.0 2.9 Montserrado 12.0 9.7 Rest of country 4.9 4.4 Education 2.1 2.0 No schooling 2.8 2.5 5.4 4.9 Primary 7.3 6.6 7.7 6.7 Secondary or more 26.8 22.1 8.1 6.3 5.2 5.1 R~liqion 8.3 5.7 Christian 9.6 8.2 8.0 7.1 Muslim 3.5 2.8 Traditlonal/Other 3.7 3.1 None 3.5 3.3 2.5 3.6 Tribe 4,8 Bassa 4.0 3,3 5.8 Gio 3.6 3.1 7,3 Gola 2.8 2.8 5.5 Grebe 12.2 I I .4 10.6 Kpelle 3.7 3.5 Krahn 4.0 4.0 Kru/Sapo II.3 10.7 9.7 Lorma 12.1 9.8 3.1 Mandingo 1.5 1.5 Mane 6.3 5.4 Other/None 9.6 7.2 Total 6,4 5.5 44 P e r c e R t 30- 26- 20- 15 10 5 0 No Eduoation Figure 4.8 Current Use of Family Planning by Education and Residence Currently Married Women 15-49 73 3,4 26.8 / :::: : :':::::: : : : : : : : : : : : : : : . : : :: :i : : :: ::.:': : : : : : : [ : : : : : : : : : : : : : : : : : : : : Secondar and Higher 11,8 / Urban Primary Rural TABLE 4.7 PERCENT DISTRIBUTION OF CURRENT CONTRACEPTIVE USERS BY THE MOST RECENT SOURCE OF SUPPLY (INFORMATION) ACCORDING TO THE METHOD CURRENTLY USED, LIBERIA, 1986 Female Other Periodic All Source Pill Sterilization Modern I Abstinence Methods 2 Government hospital or clinic 20.7 62.3 42.1 12.0 28,7 Family Planning Assoc. of Liberia (FPAL) 53.9 2.6 31.5 25.5 40.4 Church hospital or clinic 2.6 24.0 6.2 5.7 Private doctor/clinlc 3.7 8.5 8.2 9.4 5.7 Pharmacy/shop 17.9 6.0 3.0 II .9 Fieldworker/other 1.2 2.6 4.0 47.0 6.9 Don't know 2.0 3,0 0.7 Total Percent I00.0 I00.0 lO0.O I00.0 lO0.O Number 244 52 69 46 411 1 Includes IUD, in jec t ion , vaginal methods, and condom 2 Excludes withdrawal and other t rad i t iona l methods 45 Figure 4.4 Source of Family Planning Supply Current Users Family Planning Association of Liberia (FPAL) 40% er 7% Pharmacy/Shop 12% Gov't Hospital ~rivate Doctor or Qlinic " Qlinic 6% 29% Church Hospits[ or Clinic 6% 4.5 Attitude About Pregnancy and Reasons for Non-use Among respondents in the LDHS, 28 percent could be classified as exposed non-users; women falling into this category were not using contraception and were immediately exposed to the risk of pregnancy, i.e., they were not pregnant, they had had sexual intercourse within four weeks of the interview, and they were presumed to be ovulating and fecund. 1 Table 4.8 presents information on the attitude toward becoming pregnant among women in this group. Overall, only about one in four non-users immediately exposed to the risk of pregnancy reported that she would be upset if she became pregnant within the next few weeks. However, the percentage who said that they would be upset increases with the number of living children, ranging from 21 percent among women with no children to 52 percent among those with 6 or more children. Table 4.9 examines the reasons for not using family planning given by exposed non-users who said that they would be unhappy if they became pregnant right away. Approximately one out of four of these women cited factors relating to the availability--high cost or difficulty in obtaining methods--as the main reasons for non-use. Another quarter of the women pointed to fear of side effects or a lack of information about methods as the primary reason that they were not contracept- ing. An additional 13 percent said that either their husbands or they themselves disapproved of the use of contraceptive methods. Thus, almost two-thirds of the exposed non-users gave reasons for non-use which relate to variables which the family planning program in Liberia can address-- either through expanded IE&C efforts and/or improved access to services. 1 For a more complete definition of"exposed to risk," see Section 2.5 46 TABLE 4.8 PERCENT DISTRIBUTION OF NON-USERS EXPOSED TO RISK OF PREGNANCY BY ATrlTUDE TOWARD BECOMING PREGNANT IN THE NEXT FEW WEEKS, ACCORDING TO THE NUMBER OF LIVING CHILDREN, LIBERIA, 1986 Number of Attitude Total Wtd. Living Children Upset Not Upset Percent Number None 21.2 78.7 I00 0 506 1 20.8 79.2 I00.0 241 2 23.3 76.4 I00.0 208 3 29.2 70.4 lO0.O 149 4 33.2 66.8 lO0.O 148 5 40.5 59.5 I00.0 76 6 or more 51.8 48.1 I00.0 136 Total 27.4 72.6 lO0.O 1,463 Note: Exposed women are those who are not pregnant, not postpartum amenorrhelc, fecund, and sexually active. For a more detailed definition, see Section 2.5. TABLE 4.9 PERCENT DISTRIBUTION OF NON-USERS EXPOSED TO RISK OF PREGNANCY WHO WOULD BE UPSET IF THEY BECAME PREGNANT, BY MAIN REASON FOR NON-USE, ACCORDING TO AGE, LIBERIA, 1986 Reason Under 30 Years for Non-use 30 Years or More All Ages Oppose family planning 6.3 3.2 5.3 Husband objects 6.5 lO.l 7.7 Religious concerns 1.7 3.4 2.2 Fears side effects 19.I 13.7 17.4 Methods costly 13.0 I I .6 12.6 Methods d i f f i cu l t to get 12.3 6.7 I0.5 Lack knowledge I0.8 8.1 g.g Breastfeedlng 6.3 18.3 10.3 Menopausallsubfecu~d 0.0 2.7 o,g Infrequent sex 4.2 8.4 5.6 Other 19.7 13.8 17.6 Total Percent I00.0 lO0.O lO0.O Wtd. Number 266 131 398 47 4.6 Intention To Use In the Future Women who were not using a contraceptive method at the time of the LDHS interview were asked if they thought that they would do something to keep from getting pregnant at any time in the future. Table 4.10 shows that around one-third of currently married non-users intend to use in the future, 11 percent are undecided about their intentions, and 57 percent do not plan to do any- thing to avoid a pregnancy in the future. The percentage intending to use increases directly with the number of children (including any current pregnancy) that a woman has, from around 9 per- cent among women with no children to 49 percent among those with 6 or more children. TABLE 4.10 PERCENT DISTRIBUTION OF CURRENTLY MARRIED NON-USERS BY INTENTION TO USE IN THE FUTURE, ACCORDING TO NUMBER OF LIVING CHILDREN, LIBERIA, 1986 Number of Livlnq qhildren" Intention to 6 or Use in Future None 1 2 3 4 5 more Total Intends to use 8.9 22.4 30.7 35.4 38.6 46.1 49.2 32.1 Does not intend to use 79.1 64.1 57.7 55 .1 51.8 43.2 39.4 56.6 Doesn't know I f .8 13.5 l l .6 9.5 9.5 I0.7 11.4 I I .3 Total Percent ]00.0 100.0 ]00.0 100.0 ]00.0 }00.0 100.0 100.0 Wtd. Number 382 656 597 498 452 276 449 3311 " Includes current pregnancy Table 4.11 presents information on method preferences for currently married non-users who say that they intend to use in the future. The pill is clearly the most popular method among non- users; 37 percent of those intending to use in the future say that they would use the pill. The second most popular method is injection which 22 percent prefer, followed by female sterilization which 7 percent prefer. The age of the non-user is related to her method preference. Non-users under the age of 30 years are more likely to prefer the pill than older non-users. Among the latter group, injection is preferred almost as often as the pill (27 percent vs. 28 percent) and 10 percent say that they would have a sterilization. 48 TABLE 4.11 PERCENT DISTRIBUTION OF CURRENTLY MARRIED NON-USERS INTENDING TO USE IN THE FUTURE BY PREFERRED METHOD, ACCORDING TO CURRENT AGE, LIBERIA, 1986 Preferred Under 30 Years Method 30 Years and Older All Ages Pill 43.1 28.2 36.7 IUD 3.7 3.0 3.4 Injection 18.0 27.4 22.0 Vaginal methods l . l 0.7 1.0 Condom 0.5 0.3 0.4 Female sterilization 3.9 I0.4 6.7 Rhythm 1.4 0.g 1.2 Withdrawal 0.0 0.4 0.2 Other 7.3 I0.5 8.7 Not Sure 20.9 18.2 19.6 Total Percent I00.0 I00.0 I00.0 Wtd. Number 606 457 1,063 4.7 Approval of the Use of Family Planning Methods To obtain information about attitudes toward family planning, the LDHS respondents were asked whether they approved of couples who used something to avoid a pregnancy. Although all women were asked the question on approval, the analysis presented here is focused on currently married women, and excludes the roughly 30 percent of married women who have never heard of a contraceptive method. Overall, slightly less than half (46 percen0 of currently married women knowing about family planning approve of its use by couples. An almost equal proportion (39 per- cent) disapprove of family plann!ng use and 14 percent say they are not sure. Currently married women were also asked whether they thought that their husbands approved of the use of family planning methods. Table 4.12 shows the distribution of currently married women who know a method by both their own attitude and their belief about their husbands' at- titude toward family planning use. Looking only at the husband's perceived attitude, the results in- dicate ~at 29 percent of women feel that their husbands approve of family planning, while 36 percent believe that their husbands disapprove of the use of contraception. According to the wife, only 26 percent of couples in Liberia jointly approve of family planning. Communication on the subject of family planning is limited among couples in Liberia. Table 4.13 indicates that almost two-thirds of currently married women did not talk about family plan- ning with their husbands in the year preceding the LDHS. Age is a factor with regard to the likelihood of discussing family planning, with the oldest and youngest age groups being the most likely to have not talked about family planning. However, even among women in the peak 49 TABLE 4.12 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN KNOWING A CONTRACEPTI~/E METHOD BY HUSBAND'S AND WIFE'S ,~I'I1TUDE TOWARD THE USE OF FAMILY PLANNING, LIBERIA, 1986 Wife's Attitude Husband's Attitude Not Approves Disapproves Sure Total Approves 25.8 8.8 11.7 46.4 Disapproves 1.4 24.1 13.4 39.2 Not sure 1.4 3.1 9.9 14.4 Total 28.7 36.0 35.1 lO0.O TABLE 4.13 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN KNOWING A METHOD BY NUMBER OF TIMES DISCUSSED FAMILY PLANNING WITH HUSBAND, ACCORDING TO CURRENT AGE, LIBERIA, 1986 Once or More Total Weighted Age Never Twice Often Percent Number 15-19 76.7 16,8 6.5 I00.0 199 20-24 64.4 19.4 16.2 I00.0 492 25-29 61.8 21.4 16.8 100.0 624 30-34 61.9 20.3 17.8 100.0 406 35-39 61.7 18.5 19.8 100.0 374 40-44 70.5 10.6 18.8 100.0 173 45-49 72.6 14.5 12.8 100.0 202 All Ages 64.8 18.6 16.3 100.0 2,470 50 childbearing years, fewer than four in ten have discussed family planning with their husbands recently. Table 4.14 examines variations in both the percentage of currently married women knowing a method who say that they approve of family planning and the percentage who report that their hus- bands approve, according to characteristics of the woman. Attitudes toward family planning vary according to the subgroup of the population to which a woman belongs. Approval is higher among women in the middle age groups, women with more living children, women living in urban areas or in Montserrado County, women with more education and Christian women. By tribe, the highest approval level is found among the Grebo and the lowest among the Mandingo. Variations in the pattern of perceived approval by husbands generally follows those of their wives. TABLE 4.14 PERCENTAGE OF CURRENTLY MARRIED WOMEN KNOWING A METHOD WHO APPROVE OF FAMILY PLANNING AND WHO SAY HUSBAND APPROVES OF FAMILY PLANNING, ACCORDING TO BACKGROUND CHARAC'I.~RISTICS, LIBERIA, 1986 Woman Husband Woman Husband Characteristic Approves Approves Characteristic Approves Approves Aae Education 15-19 37.9 23.2 No schooling 38.0 21.4 20-24 48.0 29.7 Primary 52.1 35.3 25-29 48.2 29.5 Secondary or more 73.5 50.4 30-34 51.4 32.2 35-39 47.2 29.6 Reliqion 40-44 43.1 22.0 Christian 52.5 33.8 45-49 36.9 26.0 Muslim 41.4 24.3 Tradltional/Other 37.0 21.4 Livina Children None 39.2 22.3 None 22.9 7.8 1 36.6 20.8 Tribe 2 46.4 28.2 Bassa 33.3 18.1 3 48.8 30.0 Gio 36.6 19.0 4 52.9 32.3 Gola 55.7 27.8 5 57.4 33.0 Grebo 64.1 40.I 6 or more 63.6 49.0 Kpelle 51.0 31.6 Krahn 42.7 28.5 Urban-Rural Kru/Sapo 55.3 34.0 Urban 53.9 35.9 Lorma 53.9 34.5 Rural 41.I 23.4 Mandingo 24.1 15.6 Mano 31.4 22.4 Other/None 53.3 35.5 R~gion Sinoe 36.4 21.5 Grand Gedeh 46.6 28.9 Montserrado 54.7 35.7 Rest of country 43.6 26.2 Total 46.4 28.7 51 5. FERT IL ITYPREFERENCES 5.1 Introduction Since the main objective of most family planning programs is to allow couples to have the number of children they want, when they want them, information on fertility preferences is of con- siderable importance to program planners in gauging the need for family planning services. Women who either do not want any more children or want to delay having their next child, can be considered as potentially in need of contraception. Similarly, the proportion of births that are either unwanted or mistimed is an important indicator of family planning need. Even vigorous family planning service programs will make little headway until there is some desire to regulate fertility and changes in fertility preferences over time often predict future changes in contracep- tive practice. The LDHS questionnaire included a number of questions about fertility preferences. All cur- rently married women were asked if they wanted to have another child (after the current pregnan- cy, if the woman was pregnant) and if so, they were asked how long they wanted to walt before having their next child. All women, regardless of marital status, were asked how many children they would like to have altogether, assuming they could go back to the time when they didn't have any children. This latter variable is referred to in this report as the "ideal" number of children. Also, women who had a birth in the five years before the survey were asked if their last birth was either unwanted or mistimed. Responses for each of these sets of questions will be presented in tum. 5.2 Future Fertility Preferences Figure 5.1 shows the distribution of currently married women by whether and when they want another child. Overall, 17 percent of married Liberian women do not want any more children and an additional 33 percent want to wait at least two years before they have their next child. Adding these two figures together indicates that one out of every two women is potentially in need of family planning services either to delay or to limit births. Thirty-one percent of women want another child soon (within two years) and 19 percent either want another child, but do not know when, or are undecided as to whether they want another child. Fertility preferences differ according to the number of living children a woman has (see Table 5.1 and Figure 5.2). The proportion who want no more children rises dramatically from only two percent of childless women to 57 percent of those with six or more children. These results indi- cate substantial interest in limiting fertility; still, the fact that one-third of women with six or more children want to have another child is evidence of a strongly pronatalist culture. The data in Table 5.1 also indicate the high level of interest in spacing births that exists even among women with no children, but especially among women with 1-3 living children Table 5.2 presents data on fertility preferences by age of woman. As expected, the proportion wanting no more children rises with age, from only one percent of women 15-19 to almost half of women aged 45-49. The data also show the considerable interest in spacing births in Liberia. The proportion who want to delay their next child generally falls with age, as does the proportion of women who want their next child within two years. 53 Figure 5.1 Fertility Preferences Current ly Mar r ied Women 15-49 Want to SI (2 or more int No More 33% 17% Undec ided if Want More 8% Want Soon ~ant Another , (w i th in 2 yrs ~dec JdedWhen 31% 11% TABLE 5.1 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN AGED 15-49 BY FERTILITY PREFERENCES, ACCORDING TO NUMBER OF LIVING CHILDREN, LIBERIA, 1986 All Number of Living Children I Currently Fert i l i ty Married Preference 0 I 2 3 4 5 6+ Women Want no more children 2 2.4 1.9 5.5 9.8 23.0 33.0 56.5 17.2 Want next child in 2 or more yrs 16 .6 41.8 40.8 40.0 35.9 29.2 16 .3 33.0 Want next child within 2 years 56.2 36.4 33 .2 31.8 23.6 20.3 13 .3 30.9 Want another, undecided when 20.6 14 .2 13.6 9.9 6.8 8.3 5.6 I I .4 Undecided i f want another 4.2 5.7 6.9 8.6 I0.6 9.2 8.3 7.5 Total Percent I00.0 I00.0 lO0.O I00.0 I00.0 lO0.O lO0.O lO0.O Wtd. Number of Women 395 682 633 530 497 294 507 3,538 1 Includes current pregnancy 2 Includes women who have been s ter i l i zed 54 80% Figure 5.2 Fertility Preferences by Number of Living Children Currently Married Women 15-49 60% 40% 100% 20% 0% 0 1 2 3 4 5 6* Number of Living Chi ldren mWant No More ~Want to Space ~Want Soon ~Undec ided TABLE 5.2 PERCENT DISTRIBUTION OF CURRENTLY MARRIED WOMEN AGED 15--49 BY FERTILITY PREFERENCES, ACCORDING TO AGE GROUP, LIBERIA, 1986 Fert i l i ty Preference All Age Group Currently Married 15-19 20-24 25-29 30-34 35-39 40-44 45-49 Women Want no mere children* 1.0 6.9 Want next child in 2 or more yrs 38 .2 44.3 Want next child within 2 years 38.2 33.6 Want another, undecided when 13.0 lO.O Undecided i f want another 9.6 5.2 9.1 19 .5 25.2 35.0 47.6 17.2 40.2 32.7 24.4 16 .5 l l .6 33.0 32.2 29 .1 27 .1 30.5 23.2 30.9 l l .6 9.6 13 .2 I0.4 12 .6 l l .4 6.8 9.2 lO.l 7.5 4.9 7.5 Total Percent I00.0 lO0.O lO0.O lO0.O I00.0 lO0.O lO0.O lO0.O Wtd. Number of Women 360 675 857 539 535 261 311 3,538 * Includes women who have been s ter i l i zed 55 In terms of fertility preference measures, the proportion of women who want no more children is perhaps the single most important figure. For this reason, it has been chosen as an indicator for studying differentials in fertility preferences by background characteristics of women, shown in Table 5.3. Since fertility preferences are so highly correlated with number of living children, and since number of living children probably is related to the background characteristics, the data in Table 5.3 are tabulated by number of children as well. Urban women are somewhat more likely than rural women to want to stop childbearing, espe- cially those women who have more than three children. Looking at the data by region of residence, the proportion of women who want to have no more children is generally higher in Montserrado County and lower in Sinoe County than in the other areas. With regard to education, women with some secondary schooling are more likely than their less educated counterparts to want no more children. Regardless of the number of children they have, Christian women are more likely to want to stop childbearing than Muslim women. There is no clear pattem with regard to women with no religion or with traditional religious beliefs. Perhaps due in part to their predominantly Muslim culture, the Mandingo show the smallest proportion of women who want no more children. Krahn and Bassa women also have small proportions wanting to stop childbear- ing, while the proportions are highest among the Kru/Sapo, Kpelle, and Lorma. The proportion of women who want to stop childbearing or who want to space their next birth is a crude measure of the extent of need for family planning, since not all these women are ex- posed to the risk of pregnancy and some of them may already be using contraception. Table 5.4 presents more refined measures of need for family planning, namely, the percent of currently mar- ried women who are fecund, not contracepting and who either want no more children or want to postpone their next child. The data, along with the proportion of women in need who intend to use, are presented according to background characteristics of the woman. Overall, about 40 percent of currently married Liberian women are in need of family planning. That is, they are fecund and not using contraception, despite the fact that they do not want another child in the near future. Of these 40 percent, almost three-quarters are in need because they want to space their next births, while one-quarter are in need because they do not want to have any more children. Furthermore, 22 percent of married women (half of the women in need) say that they intend to use contraception. In other words, not only is there a substantial need for family planning services in Liberia--mostly for temporary methods--but also, many women intend to use them. The proportions in need, and the proportions in need who intend to use, are remarkably similar across background characteristics of women. Both need for and intention to use contracep- tion are slightly higher among urban women and women with primary or secondary education. Also, although Muslim and Mandingo women have a higher than average proportion in need of family planning, they have a smaller than average proportion reporting an intention to use in the future. To summarize, the data in Table 5.4 do not indicate that there are particular groups of women with greater need for family planning, but rather that need is more or less uniformly spread in the country. 56 TABLE 5.3 PERCENTAGE OF CLrRRENTLY MARRIED WOMEN AGED 15--49 WHO WANT NO MORE CHILDREN, ACCORDING TO NUMBER OF LWING CHILDREN AND BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Number of Living Children** Characteristic 0 l 2 3 4 5 All Currently Married 6+ Women Urban-Rural Urban 2.8 1.6 4.2 I I .0 25.0 39.3 64.1 Ig.4 Rural 2.2 2.1 6.2 g,o 21,5 29.1 52.0 15.8 Reqion Sinoe 0.0 1.3 3.3 4.3 8.9 18.5 44.4 12,5 Grand Gedeh 0.0 o.g 6.5 I0.8 9.0 19.7 43.4 14,g Montserrado 1.4 0,0 5.1 13.3 25.0 43.1 60.7 18.9 Rest of country 3.1 2.6 5.7 8,g 24.2 31.7 57.6 17.0 Education No education 3.2 1.8 5.1 7.2 20.6 31.8 53,1 16,6 Primary 0.0 2,4 4.8 12.1 22.8 26.8 65.0 14.8 Secondary or more 0.0 1.6 8.1 18.6 35.4 51.9 74.0 22.8 Reliqion Christian 3.1 2.6 5.4 13.8 25.8 35.5 58,9 20.5 Muslim 1.7 2.0 2.1 5.0 13.1 24.5 45.0 I0.6 Trad'I/Other 0.0 2.5 7.3 5.9 26.1 43.9 61.3 18,5 None 3.0 0.0 8.0 6.4 21.4 2g.l 54.g 14.3 Tribe Bassa 5.0 1.3 6.8 7.0 21.7 22,1 55.9 13.9 Gio 0.0 3.7 lO.O 0.0 24.3 * 75.1 18.3 Gola * * 3 .8 17 .4 * * " 15 .6 Grebo * 0.9 6.3 9.3 19.7 7.6 52.6 17.3 Kpelle 2.4 2.6 4.1 10.8 25.4 35.4 61.8 20.6 Krahn 0.0 0.0 6.2 4,8 5.8 36.2 37.7 12.0 Kru/Sapo 0.0 3,4 9.5 ll.O 15,8 36.9 56.3 21.8 Lorma * 4.2 0.0 * 12.5 * 20.4 Mandingo 3.9 0.0 2.4 6.1 4.6 * 6,7 Mano * 3.8 8.9 12.5 28.2 * 40.9 15.7 Other/None 0.0 l.O 4.0 16.2 38.2 36,7 59.3 lg, l Total 2.4 l .g 5.5 9.8 23.0 33.0 56.5 17.2 Note : Women who have been s ter i l i zed are inc luded among women who want no more ch i ld ren . * Fewer than 20 unwelghted cases ** Includes current pregnancy 57 TABLE 5.4 AMONG CURRENTLY MARRIED WOMEN, THE PERCENTAGE WHO ARE IN NEED OF FAMILY PLANNING AND THE PERCENTAGE WHO ARE IN NEED AND INTEND TO USE FAMILY PLANNING IN THE FUTURE, ACCORDING TO BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Characteristic In Need and In Need Intend to Use Contraceotlon Want No Want to Want No Want to More Postpone Total More Postpone Total Urban-Rural Urban I f .3 32.3 43.6 8.8 13.3 22.1 Rural lO.O 25.6 35.6 7.7 8.9 16.6 Reqion Slnoe 9.3 24.7 34.0 7.7 I0.9 18.6 Grand Gedeh 9.9 31,5 41.3 8.4 I I .7 20.2 Montserrado ll.O 31.8 42.8 8.2 12.4 20.6 Rest of country I0.4 26.7 37.0 8.1 9.8 17.8 Education No education I0.3 26.2 36.5 7.8 7.2 14.9 Primary 9,4 35,7 45,1 8,1 18,3 26,4 Secondary or more 12.4 31.7 44.1 I0.2 22.5 32.7 Rellqion Christian l l .4 24.9 36.2 8.8 12.4 21.1 Muslim 7.1 34.5 41.6 5.6 8.3 13.9 Trad'I/Other 13.9 29.6 43.5 I0.3 lO.l 20,3 None 9.0 29.0 38.0 7.4 8.6 16.0 Trib~ Bassa 9.2 25.3 34.5 6.2 7.3 13.5 Gio I f .2 28.9 40.I 8.9 I0.3 19.2 Gola ll.O 22.0 33.0 9.2 7.3 16.5 Grebo 9.9 26.7 36.5 8.8 16.2 25.0 Kpelle 13.5 26.2 39.7 I0.9 8.9 19.8 Krahn 7.5 29.6 37,0 7.0 13.5 20.6 Kru/Sapo 12.2 24.1 36,3 9.7 12.6 22.3 Lorma I0.6 27.5 38.0 9.8 I0.8 20.6 Mandingo 5.2 42.8 48.0 3.2 I0.2 13.3 Mano 7.2 27.0 34.2 5.4 15.3 20.7 Other/None l l .8 29.4 41.2 8.1 9.2 17.3 Total I0.5 28.1 38,6 8.1 I0.5 18.6 Note: Women in need of family planning are those who are fecund, not contracepting and who either want no more children or want to postpone their next birth for two or more years. 58 5.3 Ideal Number of Children In order to asses fertility preferences in Liberia, all LDHS respondents, regardless of marital status, were asked: "(If you could go back to the time when you didn't have any children, and) if you could choose the number of children to have in your whole life, how many would that be?" Women with children were asked the entire question, while those with no children were asked the part excluding the phrase in parenthesis. The question measures two things--first, among women who have just started childbearing, the data give an idea of the total number of children these women hope to have in future; secondly, among older, higher parity women, the data can provide some idea of the level of unwanted fertility. It is important to note that some women have difficulty in answering such a hypothetical ques- tion, presumably more so in cultures in which control over fertility is still a new concept. The fact that one out of every four women in Liberia gave a non-numeric answer ("As many as God gives me," "Don't know," etc.) is evidence of this. There is also an indication that in Liberia, some women may have interpreted the question to mean "before you had any children, how many did you want to have?", which tums it from a hypothetical into a factual question. Finally, it is usually assumed with this question that some women report their actual number of children as their ideal number, since they find it difficult to admit that they would not want some of their children if they could choose again. The distribution of women by ideal number of children and actual number of children (Table 5.5) shows that more than one out of every three women states an ideal of six or more children and fewer than one out of twenty women would prefer two or fewer children. Women who al- ready have several children state higher ideal family sizes than women with fewer children in fact, the mean ideal number of children rises steadily from 5.3 for women with no living children to 7.4 among women with six or more living children. This may be due either to the fact that women who want more children actually end up having them, or to the phenomenon mentioned above, that women rationalize the number of children they already have. Despite the generally pronatalist attitude of Liberian women, there is some evidence of unwanted fertility in the fact that over 20 percent of women with six or more living children report lower ideal numbers of children. Table 5.6 indicates that there is considerable variation in mean ideal numbers of children by age and background characteristics of women. The data point out the fact that younger women have considerably smaller ideal family sizes than older women, which implies that, if young women can succeed in having only those children they want, fertility rates may fall in the future. Regarding other differentials, rural women report higher ideal numbers of children than urban women at all age groups. Overall, the difference is large, amounting to almost one and a half children (5.2 for urban vs. 6.6 for rural women). Differences by region are even more pronounced, with women in Grand Gedeh County report- ing a mean ideal of 7.8 children, compared to 7.4 for women in Sinoe County, 5.1 for women in Montserrado County, and 6.2 for women in the rest of the country. The pattern within age groups is almost uniformly the same as the overall pattern. Women with no education also have a high ideal number of children (6.8), compared to women with primary education (5.3), and women with secondary education or more (4.5). At each age group, Christian women have lower mean 59 ideal numbers of children than women of other faiths. Differences by tribal affiliation are general- ly minimal, with women in most tribes favoring about six children. Lorma women report the smal- lest ideal number of children (5.1), while Krahn and Mandingo women report the highest, at 6.9 each. TABLE 5.5 PERCENT DISTRIBUTION OF WOMEN AGED 15--49 BY IDEAL NUMBER OF CHILDREN, ACCORDING TO ACTUAL NUMBER OF LIVING CHILDREN, LIBERIA, 1986 Ideal Actual Number of Living Children" Number of All Children 0 1 2 3 4 5 6+ Women 0 0.I 0.0 0.2 0.4 0.0 0.0 0.0 0.I 1 l.O 0.6 0.3 0.2 0.0 0.9 0.3 0.5 2 6.1 5.3 4.0 2.1 1.2 2.1 2.1 3.8 3 lO.l 9.4 5.8 6.8 2.0 2.7 3.6 6.7 4 21.0 18.8 20.5 13.8 15.7 10.6 I I .0 17.1 5 12.2 14.9 I I .5 16.8 9.3 12.6 4.2 12.0 6+ 27.8 28.4 35.3 37.2 41.9 45.7 46.3 35.2 Non-numeri c responses 21.7 22.8 22.4 22.8 29.8 25.5 32.5 24.5 Total Percent I00.0 I00.0 I00.0 lO0.O I00.0 I00.0 lO0.O I00.0 Wtd. No. of women 1,096 1,121 838 664 590 342 588 5,239 Mean, All Women 5.3 5.5 5.8 6.0 6.6 7.0 7.4 6.0 Mean, Currently Married Women 6.1 6.1 6.1 6.2 6.7 7.2 7.6 6.5 Includes current pregnancy 5.4 Unwanted Fertility Although the comparison of ideal and actual family size in Table 5.5 provides some insight into the extent of unwanted fertility, a more direct way to measure it is to ask women if their last child was wanted. The LDHS included two questions for women who had a birth in the five years before the survey: "Before you got pregnant with (NAME OF LAST BIRTH), did you want to have more children?" and (if the answer was either "yes" or"don't know"), "Were you glad that you were pregnant then, or did you prefer to wait?" The objective of the first question was to identify births that were unwanted, whereas the second question was designed to investigate mis- timing of births. The results from these questions are given in Table 5.7 for women whose last birth occurred in the 12 months before interview. 60 TABLE 5.6 MEAN IDEAL NUMBER OF CHILDREN FOR WOMEN AGED 15--49 ACCORDING TO AGE AND BACKGROUND CHARACTERISTICS, LIBERIA, 1986 Characteristic Aqe 15- ~'20- 25- 30- 35- 40- 45- All Ig 24 29 34 39 44 49 Women Urban-Rural Urban Rural ReaSon Sinoe Grand Gedeh Montserrado Rest of country 4.5 4.9 5.2 5.6 6.4 6.2 6.8 5.2 5.9 6.2 6.4 6.7 7.3 7,8 7.7 6.6 6.4 6.9 7.3 7.8 8.2 8.8 9.9 7.4 5.6 6.9 7.3 8.2 9.4 9.9 lO.O 7.8 4.4 4.8 5.1 5.4 6.1 6.2 6,5 5.1 5.5 5.8 6.1 6.3 6.9 7.3 7.3 6.2 Education No education 6.6 6.4 6.3 6.6 7.3 7.6 7.7 6.8 Primary 4,8 5.5 5.6 6.3 6.1 5.9 7.2 5.3 Secondary or more 4.2 4.4 4.8 4.4 4.6 * * 4.5 R~liqien Christian Muslim Trad'I/Other None Tribe Bassa Gio Gola Grebo Kpelle Krahn Kru/Sapo Lorma Mandingo Mano Other/None Total 4.8 5,2 5.5 6.0 6.7 7.2 7.3 5.7 5.8 6.2 6.1 6.5 7.7 8.8 8.2 6.5 6.0 5.7 6.3 6.4 6.9 6.3 7.2 6.3 6.0 6.0 6.5 6.3 7.2 7.7 7.9 6.6 5.4 5,6 6.2 6.7 7.5 6,2 6.5 6.1 5.8 6.3 6.4 6.0 8.0 " 7.2 6.5 5.1 5.9 5.8 * 6.6 " * 6.0 4.8 5.1 6.4 7.1 7.5 7.3 8.9 6.0 5.2 5,4 5.8 6.2 6.3 7.5 7.7 6.0 5.5 7.2 6.4 6.8 8.1 ]0.3 9,4 6.9 5.2 5.5 5.3 6.5 6.8 8.1 8,3 6.0 4.5 4.6 5.0 5.1 6.3 * * 5.1 6.0 6.3 6.0 7.6 8,9 * 6.9 5.7 6.0 6.3 6.9 6.3 6.9 7.7 6.3 4.9 5.0 5.7 5.0 6.1 6.6 5.7 5.4 5,2 5.5 5.9 6.2 6.9 7.4 7.5 6.0 * Fewer than 20 unweighted cases 61 TABLE 5.7 PERCENT DISTRIBUTION OF WOMEN AGED 15-49 WHO HAD A BIRTH IN THE 12 MONTHS PRIOR TO THE SURVEY BY WHETHER THEY WANTED THE CH/LD THEN, LATER, OR NOT AT ALL, ACCORDING TO BIRTH ORDER, LIBERIA, 1986 Birth Order Preference I - 2 3 or More Total Wanted child then Wanted child later Did not want child 72.9 66.1 68.7 5.9 5.4 5.6 21.2 28.4 25.6 Total Percent 100.0 1O0.0 100.0 Wtd. No. of Births 509 808 1317 The data indicate that one-quarter of women did not want their last birth, and an additional six percent say that their last birth was mistimed. Although it is possible that these rather high figures are accurate, there is some evidence from field observation that women may have misinterpreted the first question as asking whether they wanted to get pregnant at that time, as opposed to whether they wanted to have any more children ever, thus confusing the two categories. Even if some of the births reported as unwanted were actually mistimed instead, the fact that over 30 per- cent of the births in the 12-month period were either unwanted or mistimed is alarming. 62 6. MORTALITY AND HEALTH 6.1 Mortality In the Liberia Demographic and Health Survey, data on mortality were collected primarily for the purpose of estimating infant and childhood mortality rates. This focus is a result of the fact that data appropriate for adult mortality estimation require very large samples and are difficult to collect by the retrospective household survey approach. In this section mortality rates are presented for three age intervals: Infant mortality - the probability of dying between birth and exact age one; Childhood mortality - the probability of dying between age one and age five; Under five mortality - the probability of dying between birth and exact age five. Mortality rates are calculated on a period basis (i.e., utiliizing information on deaths and ex- posure to mortality by age during a specific time period) rather than on a birth cohort basis. The period approach is preferred for two reasons: first, period-specific rates are more appropriate for program evaluation and second, the data necessary for the calculation of cohort-based childhood mortality rates are only partially available for the five-year period immediately preceding the sur- vey. A complete description of the methodology for computing period-specific mortality prob- abilities is given elsewhere (Rutstein, 1984). Birth History Survivorship Data The data for the estimation of mortality rates were collected in the reproduction section of the Individual Woman's Questionnaire. The section began with questions about the aggregate childbearing experience of respondents (i.e., the number of sons and daughters who live in the household, who live elsewhere and who have died). These questions were followed by a retrospec- tive birth history in which data were obtained on the sex, date of birth, survivorship status and cur- rent age or age at death of each ofa respondent's live births. The data obtained from these questions are used to calculate infant and childhood mortality rates. A retrospective birth history, in which data are collected from respondents aged 15-49 as of the survey date, is susceptible to truncation bias and other kinds of data collection errors. Trunca- tion bias refers to the fact that for any time period prior to the year of the survey, data are not available for women at the oldest ages of childbearing (e.g., for the period 10 to 15 years prior to the survey, there is no information about births to women aged 40 to 49). Other data collection er- rors involve underreporting of events, misreporting of age at death and misreporting of date of birth. In general, all of these data problems are less serious for time periods close to the survey date. Mortality Levels and Trends, 1971-1986 Table 6.1 displays infant and childhood mortality rates for the five-year period preceding the survey (1981-86) and for two earlier five-year thne periods (1971-75 and 1976-80). Rates are presented by sex and by urban-rural residence as well as for all Liberia. 63 TABLE 6.1 INFANT AND CHILDHOOD MORTALITY ESTIMATES BY TIME PERIOD, SEX AND URBAN-RURAL RESIDENCE, LIBERIA, 1986 Time Period Percent Decline 1971-75 1976-80 1981-86" 1971-75 to 1981-86 Males Infant mortality 216 180 160 26 Child mortality 104 90 88 15 Under five mortality 298 254 234 21 Females Infant mortality 167 147 128 23 Child mortality I01 99 go I I Under five mortality 251 232 206 18 Urban Infant mortality 170 153 130 24 Child mortality 86 90 89 -3 Under five mortality 242 229 207 14 Rural Infant mortality 207 171 154 26 Child mortality I14 98 89 22 Under five mortality 298 253 230 23 Total Infant mortality 192 164 144 25 Child mortality I03 95 89 14 Under five mortality 275 243 220 20 * Includes exposure during 1986 up to the calendar month preceding the survey 64 The infant mortality rate for all Liberia for the period 1981-86 is 144 per 1,000 llve births and the childhood mortality rate is 89 per 1,000. The overall probability of dying between birth and exact age five is 220 per 1,000 (i.e., more than one in every five births dies before reaching five years of age). While these rates indicate high levels of mortality, the rates for earlier time periods are even higher so that, over the ten-year interval between 1971-75 and 1981-86, there is a clear trend of declining mortality (see Figure 6.1). During that interval infant mortality declined by 25 percent, childhood mortality by 14 percent, and the overall probability of dying between birth and age five by 20 percent. The extent to which this mortality decline is distributed by sex and urban-rural residence is also indicated in Table 6.2. The trend in infant mortality appears to be broad based, with the per- cemage decline, between 1971-75 and 1981-86, being about the same for males (26 percent) and females (23 percenO and in urban (24 percent) and rural (26 percent) areas. The decline in childhood mortality is also shared by males (15 percent) and females (11 percent), but is confined to rural areas (22 percent) and not apparent in urban areas (-3 percent). R a t e P e r 1 0 0 0 300 " 250 " 200 " S50 " 100 " 50" F igure 6.1 T rends in In fant and Ch i ld Mor ta l i ty 275 103 0 - Infant Mortality Child Mortality Under 5 Mortality m 1971-75 ~ 1976-80 ~ 1981-86 Mortality Differentials, 1976-86 In order to have a sufficient number of births to calculate reliable rates for the study of mor- tality differentials across population subgroups, period-specific rates are presented for the ten- year period 1976-86. Mortality differentials by geographic area, mother's level of education and urban-rural residence are presented in Table 6.2. 65 TABLE 6.2 SOCIOECONOMIC DII~I~ERENTIALS IN INFANT AND CHILD MORTALITY, 1976-1986, LIBERIA, 1986 Infant Child Under Five Mortality Mortality Mortality (IQo) (4Ql) (5Qo) Urban-Rural Urban 140 89 216 Rural 161 93 239 Reqion Sinoe (178) (lOl) (261) Grand Gedeh 161 (124) (266) Montserrado 150 lOl 237 Rest of country 152 83 222 Mother's Education No education 164 93 242 Primary (202) (llO) (289) Secondary or more (150) (87) (224) Total 153 91 230 Note: The rates presented include exposure during 1986 up to the calendar month preceding the survey. Figures in parenthesis are based on fewer than 500 cases. Regional-specific rates were produced for Sinoe, Grand Gedeh, and Montserrado Counties, and the rest of the country. Rates of infant mortality for Montserrado and the rest of the country are about the same as for all Liberia while the rates for Sinoe and Grand Gedeh are higher. Es- timates of infant mortality by mother's education indicate an erratic pattern which may be the result of the relatively small number of births to women in the higher education categories. The urban-rural rates display the most definitive differentials. In terms of infant mortality, the rural rate (161 per 1,000) exceeds the urban rate (140) by 15 percent. Mortality differentials by sex, mother's age at birth, birth order, and length of the previous birth interval are shown in Table 6.3. Infant mortality estimates by sex differ by about 20 percent, with lower rates for females. Mortality rates by mother's age at birth display the expected differen- tials: infant mortality is highest for births to women under age 20 (177), declines for women aged 20-24 and 25-29 (155 and 136, respectively) and increases somewhat for women aged 30 and above (142) (see Figure 6.2). Similarly, infant mortality estimates by birth order display the ex- pected differentials. 66 TABLE 6.3 DEMOGRAPHIC DIP~ERENTIALSININFANTANDCHILDMORTALITY, 1976-1986, LIBERIA, 1986 Infant Child Under Five Mortality Mortality Mortality (IQo) (4Ql) (5QO) Sex of Child Male 168 89 242 Female 136 93 217 Mother's Aae at Birth Less than 20 177 llO 267 20-24 155 97 237 25-29 136 78 203 30+ 142 78 209 Birth Order l 157 97 239 2-3 147 93 227 4-6 146 85 219 7 or more 172 88 245 Previous Birth Interval" Less than 2 years 203 92 277 2-3 years 124 95 207 4 years or more 72 58 126 Total 153 91 230 Note: The rates presented include exposure during 1986 up to the calendar month preceding the survey. Based on births of order two and higher The most significant differentials are those associated with the length of the preceding birth in- terval. The infant mortality rate estimates are 203 per 1,000 for birth intervals of less than two years, 124 for intervals of 2 to 4 years, and only 72 for intervals of 4 years or more. Overall, the rate for intervals under two years is almost three times the rate for 4 years or more. There are also substantial differentials in childhood mortality by length of the preceding birth interval----essential- ly in the same direction as the infant mortality differentials. These differentials suggest that a change in birth spacing practices would, by itself, have a favorable impact on mortality levels. Data Quality The reliability of the LDHS mortality data has been investigated by the application of a series of internal consistency checks. Although some heaping was found in the reported age at death data at 12 months of age, the evaluation revealed no major data defects. These findings cannot be considered as establishing the quality of the data set, however, because the power of internal con- sistency checks for detecting errors is quite limited. 67 Figure 6.2 Differentials in Infant Mortality 250 200 161 / 150 - 14 ~. Z~ x : x x : xx:::: xxxx: x:: :. :x: x: 1 oo 50 - i ! i: i : 0 - Urban Rur81 Residence 177 / I 165 : : : : : : : / "1 142 :::.::: ;: :: J 13~ / xx xx : . . . . : : h : : [ ] [ ~ x: x : : x x : xxx : :x : :x x x . . . . . . . . . . . . x : xx :x : xxxxx : x :::::~ :: . : xxx : xxx :x . . . . . . . . . . . . . : xxxx x :x xx : x x x x . . . . . . . . . xx xx : x :x : x x : [ : [ [x [ ' : : : : xxxx x : :x :xxx . . . . . . . . . . . . :x:::x x:x: ::R: xxx: : :. : : ::: xx : ::" xx xx : : : x :x : : : x h : : : : xx x x .xx : x x : : : : : : x :x : x :: : : : x:. : : xx : : <20 20- 25 - 30" 24 29 Age of Woman 203 / x : : : xx : :x : : : : : h . " : x : xxx xx :x : : :X3: : :x: ::x::x::: :xxx: ::::x: :: :: x: 83 : : : : : ' : : : : : x : ~2 y rs 2 -e 4* y rs y rs Birth Interval Two tests of the data are presented, both of which focus on the plausibility of the age pattern of reported deaths. In the LDHS, age at death was recorded in one of three units: in days for deaths under one month, in months for deaths under two years, or in years for deaths at age two and above. These data permit testing for gross underreporting of events which is thought to result sometimes from a failure of respondents to report births of children who die in very early infancy. A test for such underreporting consists of forming the ratio of deaths under seven days to those under 30 days of age. While the true value of this ratio is unknown, it is known that mortality rates decline with age throughout infancy and that this ratio should exceed 0.25. Values for this ratio in Liberia arc as follows: 1976-80 1981-86 Males 0.64 0.63 Females 0.66 0.73 68 The ratios conform to expectations for both males and females. The fact that they are consis- tently higher for females than for males suggests that the sex differentials in infant mortality reported above are not due to differential completeness by sex in the reporting of infant deaths. The data on age at death were also tested for digit preference (heaping) at 12 months of age. To the extent that such heaping is the result of misreporting the age of deaths occurring in the late post-neonatal period, infant mortality will be biased negatively and child mortality will be biased positively. The distribution of deaths by age in months is as follows: Age in 1976-80 1981-86 months Males Females Males Females 8 25 10 20 13 9 17 12 20 11 10 6 1 1 6 11 10 3 6 0 12 27 30 39 39 13 3 4 7 8 14 4 1 2 4 15 5 2 3 1 16 1 2 0 3 The data indicate substantial heaping at 12 months of age and a deficiency of events in the im- mediately preceding and succeeding months. The problem can be corrected by reassigning half of the deaths at 12 months to infancy, which would result in an increase in infant mortality of about 5 percent, and a decrease in child mortality of about 8 or 9 percent. Comparative Estimates Mortality estimates are available for Liberia from the 1971 Liberian Population Growth Sur- vey (LPGS), the 1974 and 1984 Censuses, and the 1984 Mortality and Health Utilization Survey (MUHS). The 1971 LPGS employed two systems of data collection and the dual record estima- tion technique to estimate infant and childhood mortality (Republic of Liberia, 1972). The LDHS rates for the period 1971-75 (192 and 103 for infant and child mortality, respectively) exceed by a considerable margin the LPGS rates for 1971 (159 and 82 for infant and child mortality, respec- tively). All that can be concluded from the comparison is that the estimates from the LPGS do not impugn the quality of the LDHS data. Indirect estimation techniques were used to estimate infant mortality rates of 141 for 1974 and 127 for 1984 (see Section 1.6). These rates are substantially lower than the comparable LDHS rates of 192 for 1971-75 and 144 for 1981-86. No explanation for the differences is readily avail- able. 69 The 1984 MUHS employed a truncated pregnancy history to collect data for the period 1977- 84 in three of the thirteen counties of Liberia (GrandCape Mount, Bomi, and Lofa). These three counties are predominantly rural. Thus, despite coverage differences, it seems appropriate to com- pare the 1984 survey with the rural strata of the LDHS. The 1984 survey estimated infant mor- tality at 189 per 1,000 births, while a comparable estimate from the LDHS is 163 (the mean of the rural rates for 1976-80 and 1981-86). Taking into consideration coverage differences and sam- pling variance, the estimates are not inconsistent. However, the 1984 survey was followed by a reinterview check survey which concluded that infant mortality was underestimated by about 20 percent (Becker et. al., 1987). Tile implied infant mortality rate of 225 per 1,000 is less consistent with the rural LDHS rate. Proportion Dead Among Children Ever Born Additional evidence of the high level of childhood mortality in Liberia is the proportion of children ever born who have died, tabulated by age of woman (Table 6.4). One-quarter of all children born to women 15--49 have died. With the exception of age group 30-34, the proportion dead rises with age of woman, as expected. TABLE 6.4 MEAN NUMBER OF CHILDREN EVER BORN, SURVIVING AND DEAD AND PROPORTION OF CHILDREN DEAD BY AGE OF WOMEN, LIBERIA, 1986 Mean Number of Children: Age of Ever Proportion Woman Born Surviving Dead Dead 15-19 0.5 0.4 0.I .184 20-24 ~.8 1.4 0.4 .213 25-29 3.2 2.4 0.7 .236 30-34 4.2 3.2 0.9 .226 35-39 5.3 3.9 1.4 .259 40-44 5.9 4.3 I .6 .264 45-49 6.8 4.7 2. l .308 All Ages 3.1 2.3 0.8 .249 6.2 Maternity Care The health care that a mother receives during pregnancy and at the time of delivery is impor- tant to the survival and well-being of the child as well as the mother. To obtain information on the type of maternity care that Liberian women receive, LDHS respondents who had given birth in the five years preceding the interview were asked if they had seen anyone for a prenatal check before their last birth and if anyone had assisted with the delivery of that child, i f they had had a prenatal checkup or received assistance at delivery, they were asked who had provided the care. In cases where the matemity care was received from more than one provider, the most qualified 70 provider was recorded by the interviewer. With regard to the service provider, it is important to note that Liberia has a well-established program of gaining for traditional midwives. Since neonatal tetanus has been shown to be a major cause of infant deaths in developing countries like Liberia, mothers also were asked if they had received an injection before the last birth to keep the baby from getting tetanus ("jerking"). The responses to this question are, of course, affected by the mother's recall of events during the pregnancy and, particularly, by her ability to distinguish the tetanus toxoid vaccination from other injections she may have received. Moreover, the failure of a respondent to be immunized against tetanus during the pregnancy prior to her last birth does not mean that the mother and child were exposed to the risk of tetanus; protection may have been provided by tetanus toxoid vaccinations before that pregnancy. Despite these drawbacks, the proportion of women receiving a tetanus toxoid vaccination during pregnan- cy provides a basic measure of the success of a country's MCH efforts since routine immuniza- tion of pregnant women is an effective means of preventing neonatal tetanus. The LDHS results suggest that the majority of mothers in Liberia receive at least some mater- nity care (see Table 6.5). Looking at prenatal care indicators, 83 percent of mothers had seen a doctor or gained nurse/midwife to check on the pregnancy, and 71 percent had had a tetanus toxoid injection before the last birth. More than 50 percent of the mothers also reported that they were assisted at the delivery of the last birth by a doctor or trained nurse/midwife, and 33 percent reported that a traditional birth attendant helped with the delivery (see Table 6.6). Although matemity care seems to be widespread among Liberian women, there are noticeable differences between population subgroups in the matemal health care indicators presented in Tables 6.5 and 6.6. Older women, rural women, women living outside Montserrado County, women with no education, and women who practice traditional religion are less likely than their counterparts to have any prenatal care, receive a tetanus toxoid injection during pregnancy or be assisted at the delivery by a doctor or gained nurse/midwife. The relationship between these in- dicators and tribe is not as consistent, but in general, the percentages receiving care are lowest among women from the Gola, Grebo, Bassa, and Krahn tribes. 6.3 Child Health Indicators The LDHS included a series of questions intended to provide information on immunization coverage and on the occurrence and treatment of diarrhea, fever, and respiratory illness among children under the age of five. The children for whom these data were collected do not include all children under five in Liberia, but only children of women who were eligible for interview in the LDHS. Thus, no information was obtained for children of mothers who had died, who were in- stitutionalized, or who, for some other reason, were not interviewed in the survey. Although the immunization status and morbidity experience of the latter children are likely to differ from that of children whose mothers were interviewed, their numbers are not large so that the results presented below can be considered as generally describing the health status of children under five years of age in Liberia. 71 TABLE 6.5 PERCENTAGE DISTRIBUTION OF WOMEN AGED 15-49 WHO HAVE GIVEN BIRTH IN THE FIVE YEARS PRECEDING THE SURVEY BY TYPE OF PRENATAL CARE PRIOR TO THE MOST RECENT BIRTH, AND PERCENTAGE OF SUCH WOMEN WHO RECEIVED A TETANUS TOXOID INJECTION PRIOR TO MOST RECENT BIRTH, ACCORDING TO BACKGROUND CHARACTERISTICS OF MOTHER, LIBERIA, 1986 Characteristic Doctor Tvpe of Prenatal Care Percent Wtd. Trained Received Number Nurse/ No Total Tetanus of Midwife Other Care Percent Injection Women Aae Under 30 18.4 67.6 1.0 12.9 I00.0 74.0 1,9gl 30 and over 16.2 60.6 1.2 21.8 I00.0 66.6 1,083 Urban-Rural Urban 25.2 66.2 0.3 8.2 I00.0 76.8 1,256 Rural 12.4 64.4 1.6 21.4 lO0.O 67.6 1,817 Region Sinoe 22.3 62.9 1.4 13.4 lO0.O 60.4 87 Grand Gedeh 9.2 62.4 0.5 27.4 lO0.O 47.9 193 Montserrado 29.8 61.6 0.2 8.3 lO0.O 76.2 780 Rest of country 13.5 66.9 1.4 18.0 lO0.O 72.2 2,013 Edu¢ation No education 12.4 65.2 1.4 21.0 lO0.O 67.0 2,001 Primary 20.I 69.0 0.9 I0.0 I00.0 75.3 531 Secondary or more 35.1 61.2 0.0 3.7 lO0.O 83.8 543 Reliqion Christian 19.4 65.1 l.O 14,3 lO0.O 72.9 1,592 Muslim 18.2 70.6 0.9 I0.0 lO0.O 75.8 481 Trad'l/Other I I .4 63.4 0.8 24.3 I00.0 64,6 587 None 18.9 6].3 2.0 17.8 100.0 70.0 414 Tribe Bassa 16.1 60.0 0.0 23.9 I00.0 64.8 368 Gio 23.5 66.0 2.9 7.6 I00.0 82.3 234 Gola 9.1 63.6 0.0 27.3 lO0.O 55.7 121 Grebe 17.4 62.1 l.O 19.3 lO0.O 57.7 220 Kpelle 11.8 69.3 1.6 17.3 100.0 72.4 540 Krahn 15.6 60.8 1.2 22,0 100.0 52.7 138 Kru/Sapo 24.3 60.I 0.7 15.0 lO0.O 69.3 311 Lorma 16.7 74.7 0.0 8.7 lO0.O 87.0 159 Mandingo 20.0 70.6 0.0 9.4 lO0.O 76.7 203 Mane 18.8 59.1 2.2 19.9 100.0 69.6 249 Other/None 19.7 67.8 l.O ll.O lO0.O 79.6 530 Total 17.7 65.2 I . I 16.0 I00.0 71.4 3,074 Note: Women giving birth 1-59 months before the survey are included. 72 TABLE 6.6 PERCENT DISTRIBUTION OF WOMEN AGED 15-49 WHO HAVE GIVEN BIRTH IN THE FIVE YEARS PRECEDING THE SURVEY BY TYPE OF ASSISTANCE AT DELIVERY OF THE MOST RECENT BIRTH, ACCORDING TO BACKGROUND CHARACTERISTICS OF MOTHER, LIBERIA, 1986 Characteristic Tvoe of Assistance at Delivery Weighted Trained Trad'l Number Nurse/ Birth Rel'tve/ No Total of Doctor Midwife Att'dnt Friend One Percent Women Aqe Under 30 6.9 54.3 29.g 7.0 1.9 lO0.O 1,991 30 and over 6.0 46.3 38.2 7.3 2.1 lO0.O 1,083 Urb@n-Rural Urban lO.O 66.9 16.6 4.4 2.0 lO0.O 1,256 Rural 4.3 40.8 44.0 9.0 1.9 lO0.O 1,817 Reqlon Sinoe 3.9 40.4 44.3 8.9 2.5 lO0.O 87 Grand Gedeh 2.2 34.8 24.9 31.7 6.1 lO0.O Ig3 Montserrado 12.g 65.6 15.0 4.3 2.1 lO0.O 780 Rest of country 4.7 48.1 39.9 5.8 1.4 lO0.O 2,013 Education No education 4.9 44.5 41.0 7.7 l.g lO0.O 2,001 Primary 5.1 56.5 26.9 9.0 2.3 lO0.O 531 Secondary or more 14.4 72.3 8.3 3.2 1.8 lO0.O 543 Reliqion Christian 7.1 57.0 25.8 7.5 2.5 lO0.O 1,592 Muslim 8.3 54.7 29.2 6.1 1.5 lO0.O 481 Trad'I/Other 4.5 41.9 46.0 6.9 0.7 lO0.O 587 None 5.6 40.0 45.1 7.4 1.8 lO0.O 414 Tribe Bassa 6.4 46.5 40.8 4.5 1.9 100.0 368 Gio 2.3 54.8 42.3 0.0 0.6 lO0.O 234 Gola l . l 31.8 47.7 15.9 3.4 lO0.O 121 Grebe 7.6 49.5 18.7 15.8 8.3 lO0.O 220 Kpelle 5.9 40.7 45.9 6.4 1.0 I00.0 540 Krahn 6.6 43.2 19.0 28.6 2.1 lO0.O 138 Kru/Sapo 7.7 57.0 24.8 6.7 3.7 lO0.O 311 Lorma 7.7 71.4 15.6 4.4 0.9 lO0.O 159 Mandingo 8.8 65.6 22.1 3.3 0.2 lO0.O 203 Mano 5.0 50.8 43.6 0.6 0.0 lO0.O 249 Other/None 9.1 57.5 24.4 7.2 1.4 lO0.O 530 Total 6.6 51.5 32.8 7.1 1.9 I00.0 3,074 Note: Women giving birth 1-59 months before the survey are included. 73 Immunization of Children In the LDHS, women who had children under the age of five were asked if the children had health cards. If a health card was available, the interviewers copied from the card the dates on which the child had received immunizations against the following diseases: tuberculosis (BCG); diphtheria, whooping cough (pertussis) and tetanus (DFF); polio; and measles. If the child had no card or the interviewer was not able to examine the card, the mother was asked if the child bad ever received a vaccination. However, no information was obtained on specific vaccinations for these children because of doubts about the reliability of the mother's recall. The data in Table 6.7 indicate that immunization cards were seen for slightly more than one- third of all children under the age of five. Of those with cards available, almost all had received at least one immunization. This is not surprising, since one of the major reasons for issuing a health card is to record immunizations. For another 32 percent, vaccination cards were unavailable but their mothers reported that they had been immunized. Thus, in Liberia, around two out of every three children under the age of five may be assumed to have received some immunization. The information on specific immunizations collected for children with health cards is also presented in Table 6.7. In examining these data, it should be borne in mind that the World Health Organization recommends that children be fully immunized before they reach the first birthday, according to the schedule outlined below (Sherris, et.al., 1986): Age Immunizations Birth BCG 6 weeks DPT, Polio i0 weeks DPT, Polio 14 weeks DPT, Polio 9 months Measles The LDHS found that, among children age one year and over for whom health cards were available, almost 80 percent had received a BCG vaccination and more than 60 percent had been immunized against measles. DPT and polio coverage was much lower; although 80 percent or so of the children age one year or older had had at least the first immunization against both DPT and polio, less than 30 percent have received either the three DPT or the three polio doses considered necessary for full protection. Overall, only about one out of five children age one year and over with health cards was fully immunized, i.e., had received BCG and measles vaccinations and com- pleted three doses of DPT and polio. Immunization coverage rates do not differ between boys and girls. Both the residence and educational status of the child's mother are, however, associated with immunization coverage, with the highest rates of coverage reported for children of urban women, of women living in Montserrado County, and of women with a secondary education (Table 6.7). Considering differen- tials by religion, immunization coverage is greatest among children whose mothers are Christian. 74 TABLE 6.7 AMONG ALL CHILDREN UNDER 5 YEARS OF AGE, THE PERCENTAGE WITH HEALTH CARDS AND THE PERCENTAGE RECORDED ON HEALTH CARD OR REPORTED BY MOTHER AS HAVING BEEN IMMUNIZED AND, AMONG CHILDREN WITH HEALTH CARDS, THE PERCENTAGE FOR WHOM BCG, DPT, POLIO, AND MEASLES ARE RECORDED AND THE PERCENTAGE CONSIDERED TO BE FULLY IMMUNIZED, ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTHER, LIBERIA, 1986 Percent of Children Under 5: Percent of Children Under 8 Who Have Received: Immunized Wtd.No. With as Rec'rd Immunized All of Health on Health as Rep'td DPT Polio Mea- Immuni- Child'n Cards Cards by Mother BCG 1 2 3 1 2 3 sles zations Under 5 <6 mos. 34.0 31.5 12.5 6-II mos. 44.4 43.4 22.4 12-]? mos. 42.3 40.3 28.3 18-23 mos. 43.8 42.4 30.9 24-35 mos. 36.0 35.2 36.0 36-59 mos. 24.9 24.0 40.7 Sex Boy 34.2 32.6 32.1 Girl 34.4 33.4 31.2 Urban-Rural 85.4 54.4 13.7 4.8 63.9 20.I 8.0 8.9 0.8 510 89.8 78.7 32.4 15.7 79.8 32.9 12.6 23.5 6.9 607 81.9 82.9 42.2 28.5 77.4 39.3 24.1 63.6 21.4 478 82.7 90.2 52.4 34.7 85.8 46.8 30.I 72.6 23.4 338 75.6 84.2 42.6 27.5 71.8 39.3 24.8 71.7 16.8 762 75.1 84.6 45.2 28.2 71.I 42.1 27.6 69.5 21.5 1,529 81.1 80.7 40.6 24.7 74.1 38.4 22.7 53.4 15.8 2,143 80.6 79.5 37.2 22.6 74.6 36.3 20.7 53.9 15.4 2,081 Urban 37.1 36.1 35.6 84.2 78.1 38.5 27.0 74.6 39.6 25.8 49.1 17.6 1,743 Rural 32.3 30.8 28.9 78.2 81.8 39.2 21.0 74.2 35.5 18.4 57.3 14.0 2,481 Reqiqn Sinoe 22.3 22.6 33.4 59.9 91.5 20.4 7.2 69.7 16.5 5.9 47.4 3.3 123 Grand Gedeh 23.0 22.6 29.6 76.3 81.2 16.1 9.7 70.4 12.4 4.8 40.3 3.2 258 Montserrado 31.7 31.0 40.4 84.9 79.2 37.6 27.8 80.4 38.4 26.9 48.6 20.4 1,065 Rest of c'try 36.9 35.2 28.4 80.4 80.1 41.I 23.5 72.7 39.0 21.4 56.2 15.1 2,778 EduCati qn No education 32.0 30.4 29.1 79.6 76.9 36.3 18.9 71.5 33.9 16.7 51.8 11.6 2,783 Primary 36.5 35.7 31.8 81.6 81.6 27.7 15.3 74.1 28.5 16.0 50.4 9.3 731 Secondary+ 41.0 40.6 41.6 84.0 88.8 57.0 45.8 83.4 55.9 42.3 62.0 33.7 710 Reliqiqn Christian 35.6 34.8 32.5 82.6 81.7 41.4 26.8 75.8 40.5 25.8 55.2 19.3 2,140 Muslim 34.0 31.7 34.3 78.7 74.8 41.9 24.0 70.9 36.8 19.6 51.9 14.0 666 Trad'I/Other 31.7 31.0 30.4 80.2 84.2 34.1 16.7 76.4 31.4 12.5 48.9 9.4 834 None 33.3 31.1 27.4 77.6 74.9 32.1 20.1 69.9 33.4 20.5 55.8 II.3 584 Tribe Bassa 32.3 29.2 22.3 75.9 74.1 41.2 30.0 71.6 41.2 30.9 51.0 23.2 497 Gio 19.7 19.7 42.4 70.4 74.6 I I . I 6.6 47.1 13.2 8.7 59.5 2.4 331 Gola 25.2 24.4 32.5 71.0 93.6 48.4 25.8 83.9 45.2 32.3 61.3 19.4 169 Grebo 35.0 34.5 31.9 75.5 91.4 41.4 32.1 77.2 36.8 27.I 41.4 17.4 305 Kpelle 43.5 42.4 24.2 83.5 77.5 39.9 22.5 76.0 40.5 21.0 57.7 14.4 768 Krahn 17.8 18.0 30.5 86.5 85.9 36.2 28.0 83.7 34.2 25.0 43.7 19.7 180 Kru/Sapo 25.9 25.6 37.7 77.5 87.1 25.9 17.3 76.1 27.9 14.8 46.8 I0,8 435 Lorma 48.5 47.8 27.5 90.4 86.3 53.4 38.4 87.7 53.4 37.0 61.6 27.4 207 Mandingo 37.6 35.6 35.5 72.6 73.9 47.2 25.5 69.6 36.2 16.8 52.9 12.1 273 Mano 26.6 25.8 38.1 84.6 76.9 27.7 15.4 64.6 24.6 I0.8 52.3 6.2 336 Other/None 41.7 39.5 34.1 85.5 79.5 41.1 20.7 76.0 38.7 18.3 53.9 16.0 723 Total 34.3 33.0 31.6 80.9 80.2 38.9 23.6 74.4 37.3 21.7 53.6 15.6 4,224 75 Looking at tribe, the highest rates are observed for children of women belonging to the Lorma and Bassa tribes. Child Morbidity and Treatment In addition to the immunization data, information was collected for all children under the age of five on the occurrence of diarrhea, fever, and respiratory illness in the four weeks preceding the interview and the treatment provided for children experiencing these illnesses. Information was also collected on whether the children had ever had measles. The data on diarrhea, fever, and respiratory illness cannot be used to measure incidence of these aliments. However, they provide the basis for a period prevalence estimate for each illness, i.e., the percentage of children under 5 years whose mothers report that they had the illness in question during the four weeks preceding the survey. In considering the morbidity information, it is important to remember that the measures are influenced by the mother's subjective evaluation of whether the child experienced the illness in question. For example, the question on diarrhea simply asked the mother whether the child had "running stomach" during the the last four weeks. The responses to the question are clearly dependent on what the mother understood by the term "running stomach" and, thus, there may be considerable variation in the length and severity of the diarrheal episodes reported in response to the question. The morbidity measures also are affected by the reliability of the mother's recall as to when the episode of the illness in question occurred. Both the failure to report illness occurring within the reference period of four weeks and the reporting of episodes that occurred prior to the four week period affect the accuracy of the prevalence estimate. In interpreting the morbidity data, the period in which the LDHS fieldwork took place (March-July) should be kept in mind since the number of cases of the illnesses in question---diarrhea, fever, and respiratory problems--vary seasonally. Diarrhea Table 6.8 shows the percentage of children under the age of five reported as having had diar- rhea during the four-week period prior to the survey. According to their mothers, four out of ten children in this age group had had at least one episode of diarrhea during the period in question. Diarrhea prevalence rates varied with the age of the child, with the rate exceeding 50 percent among children aged 12 to 24 months. The sex differential was insignif icant~0 percent of boys were reported as having had diarrhea compared to 38 percent of girls. Diarrheal prevalence was lower among children of urban mothers and children whose mothers had at least secondary education. However, the percentage reported as having had diarrhea ex- ceeded 30 percent, even among those children whose mothers were living in urban areas or whose mothers had had a secondary education. By region, diarrhea prevalence was lowest in Montser- rado County. Considering tribe, i't was highest for the Bassa and Gola and lowest among the Man- dingo. For children who had had a diarrheal episode in the four weeks preceding the survey, Table 6.8 also indicates what, if anything, mothers said that they had done to treat the diarrhea. In over 80 percent of the cases, some effort was made to treat the diarrhea. Among mothers who treated the diarrhea, the most frequently reported treatment was antibiotics; 44 percent of the children who had diarrhea during the four weeks before the survey were given antibiotics. Country 76 TABLE 6.8 AMONG ALL CHILDREN UNDER 5 YEARS OF AGE, THE PERCENTAGE REPORTED BY THE MOTHER AS HAVING HAD DIARRHEA IN THE FOUR WEEKS PRECEDING THE SURVEY AND, AMONG CHILDREN WITH DIARRHEA, THE PERCENTAGE RECEIVING VARIOUS TREATMENTS, ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTHER, LIBERIA, 1986 Charac- teristic ~unong Children Among Children with Diarrhea, Percent Treated with: a Under 5, Percent Wtd.No. Reported Having Home Country Other No of Diarrhea in Salt/Sugar Anti- Medicine/ Treat- Treat- Child'n Past 4 Weeks ORS Solution biotics Herbs ment ment Under 5 b < 6 mos. 33.6 3.4 3.2 39.I 32.7 5.1 22.0 510 6-11 mos. 48.4 7.0 3.4 40.6 30.9 4.0 14.9 607 12-17 mos. 57.7 7.6 4 .8 45.9 22.1 8.0 13.0 478 18-23 mos. 52.1 4.9 3.1 56.1 22.2 5.7 8.6 338 24-35 mos. 44.9 7.1 3.4 44.9 21.6 6.2 15.7 762 36-59 mos. 25.8 6.8 1.7 39.8 27.4 3.1 20.0 1,529 Sex Boy 40.0 6.4 3.5 42.0 27.7 5.1 16.1 2,143 Girl 38.3 6.5 2.9 45.5 24.1 5.2 16.0 2,081 Urban-Rural Urban 36.3 9.8 3.1 54.5 13.7 5.6 14.3 1,743 Rural 41.1 4.4 3.3 36.9 33.5 4.9 17.1 2,481 R~qipn Sinoe 39.7 1.5 O.O 57.6 22.9 l . l 17.0 123 Grand Gedeh 37.4 3.0 2.0 33.0 32.0 6.9 23.8 258 Montserrado 33.6 10.4 3.9 57.3 14.2 5.4 12.3 1,065 Pest of c 'try 41.4 5.7 3.2 39.7 29.2 5.1 16.5 2,778 Education No education 40.2 4.5 2.3 42.6 30.4 4.0 16.2 2,783 Primary 42.1 9.2 4.6 44.0 21.9 7.9 14.5 731 Secondary+ 31.9 12.3 5.6 48.3 9.4 7.3 17.6 710 R~liqion Christian 39.5 6.7 3.6 45.0 26.3 4.4 15.3 2,140 Muslim 33.9 5.0 3.7 45.9 20.9 8.1 13.9 666 Trad'I/Other 38.8 5.7 2.6 45.7 23.9 4.1 18.1 834 None 44.3 7.8 2.4 34.9 31.6 6.6 17.8 584 Tribe Bassa 45.6 4.3 4.3 30.7 34.0 5.5 22.5 497 Gio 41.6 13.0 4.0 30.0 33.0 8.0 I0.0 331 Gola 47.2 I0.3 5.2 41.4 39.7 3.5 6.9 189 Grebo 38.8 3.8 7.8 44.3 23.0 7.I 17.8 305 Kpelle 41.7 4.9 1.5 45.7 33.4 2.7 I0.6 768 Krahn 39.5 6.1 0.5 30.9 30.8 8.8 27.6 180 Kru/Sapo 40.0 5.9 5.5 55.9 16.5 5.9 13.3 435 Lorma 32.8 6.1 0.0 61.5 4.1 6.1 22.3 207 Mandingo 28.0 5.8 0.0 44.4 18.4 I0.8 13.4 273 Mano 40.6 9.I 2.0 33.3 34.3 4.0 17.2 336 Other/None 34.0 6.2 2.8 56.6 IO.7 3.4 19.8 723 Total 39.1 6.5 3.2 43.7 25.9 5.2 16.0 4,224 a Mothers were able to spec i fy more than one t reatment , so percents may add to more than 100. b Includes children aged 1-59 months 77 medicine or herbs was the second most commonly employed treatment. The latter treatment may have had some effect on dehydration, the chief cause of death among children with diarrhea. Few mothers relyed on oral rehydration therapy (ORT), a relatively cheap and effective means of preventing or treating dehydration. Less than 10 percent of the children were treated with a solu- tion of oral rehydration salts (ORS) or with a home prepared salt/sugar solution. Differences in treatments generally followed expected pattems. Use of antibiotics was greatest among urban, well-educated mothers, presumably because modem medical products were more available and more affordable for these women than for women living in rural areas or those with less than a secondary education. Although antibiotics remained the predominant treatment given by mothers in the latter categories, they were more likely than those in urban areas or those with secondary education to say they had used country medicine or herbs. The somewhat greater reliance on ORT among urban, well-educated women also likely reflects the fact that women in these categories have greater access to ORS packets or exposure to informational efforts designed to promote the use of homemade salt and sugar solutions. Fever In Table 6.9, information is presented on the percentage of children under the age of five reported to have had fever during the four weeks prior to the LDHS interview. Malaria is endemic in Liberia, and the questions on fever were designed to obtain an estimate of the extent to which children experienced a bout of malaria during the reference period. Overall, mothers reported that one out of two children under age five bad had fever during the month before the survey. The age of the child was related to the reported episode of fever, with the prevalence peaking at 67 percent among children age 12-17 months. There was no evidence of sex differentials in the reporting of fever prevalence. The likelihood that a child had had fever varied somewhat with the residence and education of the mother, but even among children of urban, educated mothers, the percentage suffering from fever exceeded 40 percent. The overwhelming reliance on antimalarial medication to treat fever suggests that malaria was considered to be the cause of the fever in the majority of the reported cases. Table 6.9 shows that roughly three out of four children who experienced fever during the month before the survey were given antimalarial drags to treat the fever. Other treatments reported by mothers included country medicine, which was used in treating 12 percent of the children who had fever and antibiotics, which were used in treating less than 4 percent of the cases. Table 6.9 shows that there is little variation in the percentage of children receiving treatment for fever, with either the age or sex of the child. However, the type of treatment given differs somewhat according to the socioeconomic characteristics of the mother. Although antimalarial drugs were the most commonly employed treatment in all subgroups, the percentages treated with antimalarials are highest among children of urban women, of women riving in Montserrado Coun- ty and of women with secondary education. Access to pharmaceutical products again is likely to be greater for these groups than for others in the population. 78 TABLE 6.9 AMONG ALL CHILDREN UNDER 5 YEARS, THE PERCENTAGE REPORTED BY THE MOTHER AS HAVING FEVER IN THE FOUR WEEKS PRECEDING THE SURVEY AND, AMONG CHILDREN WITH FEVER, THE PERCENTAGE RECEIVING VARIOUS TREATMENTS, ACCORDING TO BACKGROUND CHARACTERISTICS OF CHILD AND MOTHER, LIBERIA, 1968 Charac- teristic Among Children Among Children With Fever, Percent Treated with a Wtd Under 5, Percent No. of Having Fever in Anti- Anti- Country Other No Children Past 4 Weeks biotics malarials Med/Herbs Treatmt Treatmt Under 5 b A~e < 6 mos. 41.9 3.9 66.7 15.8 1.6 II.3 510 8-II mos. 63.2 3.1 74.4 11.3 5.0 9.7 607 12-17 mos. 67.4 5.7 73.9 11.3 4.5 9.5 478 18-23 mos. 61.1 3.6 78.8 7.5 3.8 10.0 338 24-35 mos. 53.8 2.3 76.4 10.3 3.4 9.4 762 36-59 mos. 40.9 3.9 71.2 12.5 2.8 12.0 1,529 Sex Boy 51.1 4.0 72.4 12.1 4.4 10.2 2,143 Girl 51.2 3.4 74.6 I0.9 2.6 10.7 2,081 Urban-Rural Urban 45.5 3.8 87.3 3.8 3.6 4.9 1,743 Rural 55.2 3.6 65.4 16.0 3.5 13.7 2,481 Req~qD Sinoe 49.7 3.5 60.2 19.2 8.3 14.5 123 Grand Gedeh 51.9 1.7 57.6 22.1 3.6 15.5 258 Montserrado 43.8 3.5 87.0 5.3 3.8 5.9 1,065 Rest of c'try 54.0 3,9 71.2 12.2 3.2 l l .3 2,778 Education No education 52.9 2.7 69.4 13.5 2.5 13.2 2,783 Primary 49.7 6.0 77.9 9.2 4.9 6.3 731 Secondary+ 46.0 5.9 86.B 5.4 6.6 2.7 710 Reliqiqn Christian 50.8 3.5 75.5 I0.7 4.0 9.2 2,140 Muslim 54.8 3.8 75.8 10.7 3.9 9.7 666 Trad'I/Other 50.0 4.8 67.0 13.0 2.1 14.6 834 None 49.9 3.1 72.0 13.7 3.3 I0.5 584 Tribe Bassa 46.2 1.2 87.6 15.6 9.0 I0.2 497 Gio 50,9 3,5 73.7 9.8 0.0 10.6 331 Gola 63.4 2.6 68.0 20.5 5.1 6.4 169 Grebo 49.5 5.9 75.9 8.5 4.7 10.6 305 Kpelle 57.0 4.7 70.6 14.4 3.1 I0.9 768 Krahn 47.3 3.1 62.9 23.3 2.2 If.2 180 Kru/Sapo 51.6 4.5 73.4 11.0 3.3 12.0 435 Lorma 46.0 2.9 82.7 11.6 2.9 4.3 207 Mandingo 48.8 l.O 87.1 4.4 3.1 6.4 273 Mano 51.6 3.2 74.6 10.3 0.0 11.9 336 Other/None 49.2 4.7 75.8 5.5 3.6 12.5 723 Total 51.2 3.7 73.5 11.5 3.5 10.5 4,224 a Mothers were able to specify more than one treatment, so percents may add to more than lO0. b Includes chi ldren aged 1-59 months 79 Cough/Difficult Breathing An attempt was made in the survey to obtain information on the prevalence of respiratory ill- ness by asking for each child under the age of five whether the child had had cough or difficulty in breathing in the four weeks before the survey. Table 6.10 indicates that more than one-third of the children under the age of five were reported by the mothers as suffering from a cough or dif- ficult breathing in the four weeks before the survey. The percentage reported as having some respiratory-related problem varied with the age of the child; the highest percentage--53 percent-- was reported for children aged 6-11 months. Again there appeared to be little difference in the likelihood that a mother would report cough or difficult breathing with the sex of the child. Dif- ferences along socioeconomic lines were generally minor. The most evident differences were be- tween Sinoe County and other areas and among tribes, with the lowest prevalence reported for the Mano (28 percent) and the highest prevalence for the Gola (49 percent). Over 80 percent of the children experiencing cough or difficult breathing received some treat- ment for the problem. Again as with diarrhea and fever, mothers are most likely to use a phar- maceutical product in treating respiratory illness. Table 6.10 shows that 62 percent of the children who had a cough or diffic

View the publication

You are currently offline. Some pages or content may fail to load.